The Joint Annual Meeting of

The Bioelectromagnetics Society and the

European BioElectromagnetics Association

Abstract Collection

www.slovenia.info/en
Table of Contents
Disclaimer...................................................................................................................................................... 3
Plenary Sessions ............................................................................................................................................ 4
Plenary 1: Effects of Static and Oscillatory Magnetic Fields on Cryptochromes: A Mechanistic
Perspective ................................................................................................................................................ 4
Plenary 2: The International Commission on Non-Ionising Radiation Protection (ICNIRP) Draft High
Frequency (100 kHz – 300 GHz) Guidelines .............................................................................................. 7
Plenary 3: Exploring Dielectrophoresis and its Applications in the Biomedical Sciences ......................... 8
Plenary 4: Frontiers of Electroporation, from Mechanisms to Applications ............................................ 9
Plenary 5: Hot Topic ................................................................................................................................ 11
Platform Sessions ........................................................................................................................................ 12
Session 1: Human studies I / brain .......................................................................................................... 12
Session 2: Dosimetry I - Computational studies ..................................................................................... 33
Session 3: In vitro I .................................................................................................................................. 58
Session 4: MRI studies ............................................................................................................................ 78
Session 5: Epidemiology.......................................................................................................................... 99
Session 6: In vitro II ............................................................................................................................... 118
Session 7: Electroporation .................................................................................................................... 133
Session 8: In vivo ................................................................................................................................... 161
Session 9: Mechanisms ......................................................................................................................... 168
Session 10: Human studies II ................................................................................................................ 182
Session 11: Dosimetry II - Measurements ............................................................................................ 204
Session 12: Nano pulses ........................................................................................................................ 230
Session 13: Public Policy I ..................................................................................................................... 251
Session 14: Dosimetry III - Measurements ........................................................................................... 263
Session 15: Dosimetry IV - Computational studies ............................................................................... 290
Session 16: Public Policy II .................................................................................................................... 315
Workshops – Tutorials .............................................................................................................................. 337
Session PW1 - Pre Conference Workshop: EMF exposure from 5G equipment: the state of art of
research and standardization ............................................................................................................... 337
Session PW2 - Pre Conference Workshop: MURI/AFOSR Workshop on Nanoelectropulse-induced
electromechanical signaling and control of biological systems ............................................................ 339
Session T1 - Tutorial 1: Systematic Review Approaches in Environmental Health Sciences ................ 341

1
 Session W1 - Workshop 1: Transferring Evidence from Research and Case Reports into Policy and
Society ................................................................................................................................................... 342
Session T2 - Tutorial 2: Magnetic Nanoparticles as Therapeutic Agents: Focusing on the Role of a
Scientist and Engineer ........................................................................................................................... 343
Session T3 - Tutorial 3: Evaluation of Animal Cancer Studies (overview) ............................................. 345
Session W2 - Workshop 2: EMF and Cancer: Interaction Mechanisms Leading to Hazardous or
Beneficial effects ................................................................................................................................... 345
Session W3 - Workshop 3: Advanced Approaches for Analysis of Biological Effects of Pulsed Electric
Fields ..................................................................................................................................................... 347
Student Flash Poster Sessions ................................................................................................................... 362
Session FA - Student Flash Poster Session A ......................................................................................... 362
Session FB - Student Flash Poster Session B ......................................................................................... 368
Poster Sessions ......................................................................................................................................... 374
Session PA - Poster Session A................................................................................................................ 374
Session PB - Poster Session B ................................................................................................................ 562
Sponsors .................................................................................................................................................... 790
Author Index ............................................................................................................................................. 797

2
These abstracts were generated electronically by their authors. The Bioelectromagnetics
Society and the European Bioelectromagnetics Association make no representations or
warranties as to the contents of this abstract compilation and assume no responsibilities for any
errors or omissions.

3
 Session: P1
Plenary 1: Effects of Static and Oscillatory Magnetic Fields on Cryptochromes: A
Mechanistic Perspective
Monday June 25, 2018 • 09:30 - 10:30
Emerald
Chairs: Meike Mevissen & Azadeh Peyman

P1-1 [09:30]
Cryptochrome-based magnetoreception and the effects of RF-fields on animal orientation
Rachel Muheim1
1Department of Biology, Lund University, Lund, Sweden

Biographical sketch
Rachel Muheim is a researcher at Lund University, Sweden, and part
of the Lund Vision Group and the Center for Animal Movement
Research, CAnMove. Her research focuses on the behavioural and
physiological characterization of the light-dependent magnetic
compass and polarized light sensitivity in birds, and on stopover
behaviour and the use of different orientation cues in migratory birds.
Rachel Muheim received a Master’s degree in Zoology from the
University of Zürich, Switzerland, and a PhD on magnetic orientation
in migratory birds from Lund University. She worked four years as
postdoc in the lab of Prof. John Phillips at Virginia Tech, USA,
investigating magnetic compass orientation in a variety of organisms,
including newts and mice. Upon her return to Lund in 2008, she
developed a novel behavioural training assay to study magnetic
compass orientation in birds in a non-migratory context. The
objectives of her work are to combine the new possibilities that have opened up with this new
behavioural approach with the latest molecular genetic, neuroanatomical, and biophysical tools. The
goal is to unravel the behavioural, physiological and molecular mechanisms of light-dependent
magnetic compass orientation and polarized light sensitivity in birds, identify and describe the
interactions between these two sensory cues, and study how birds use the information for
orientation during migration and in other spatial orientation tasks.
Five relevant publications
1. Pinzon-Rodriguez A, Muheim R (2017): Wavelength and intensity of light determine magnetic
compass orientation in zebra finches. J. Exp. Biol. 220, 1202-1209. DOI 10.1242/jeb.148098
2. Muheim R, Sjöberg S, Pinzon-Rodriguez A (2016): Polarized light modulates light-dependent
magnetic compass orientation in birds. PNAS 113, 1654-1659. DOI 10.1073/pnas.1513391113
3. Phillips JB, Youmans PW, Muheim R, Sloan KA, Landler L, Painter MS, Anderson CR (2013):
Rapid learning of magnetic compass direction by C57BL/6 mice in a 4-armed "plus" water
maze". PLoS ONE 2013, 8:e73112.
4. Phillips JB, Jorge PE, Muheim R (2010a): Light-dependent magnetic compass orientation in
amphibians and insects — candidate receptors and candidate molecular mechanisms. J. R.
Soc. Interface 7, S241-256. DOI 10.1098/rsif.2009.0459.focus
5. Muheim R, Edgar NM, Sloan KS, Phillips JB (2006): Magnetic compass orientation in C57BL/6J
mice. Learn. Behav. 34: 366-373.
Abstract
A wide range of animals can sense the Earth’s magnetic field, and there is growing evidence that
this is an omnipresent capability of the majority of organisms and thought to play a fundamental role
for animals to orient in space in everyday life. Birds and other animals have a light-dependent

4
magnetic compass, which operates well under UV to green wavelengths, but leads to disorientation
or changed orientation at longer wavelengths. This magnetic compass is suggested to be based on
a light-dependent process taking place in specialized photopigments, which upon light excitation
form radical pairs whose spin states, and thereby the photopigments’ response to light, can be
affected by an earth-strength magnetic field. Such magnetosensitive photopigments arranged in an
ordered array, like the avian retina, would show increased or decreased sensitivity to light,
depending on their alignment to the magnetic field. Cryptochromes are the only known animal
photopigments that can form persistent, spin-correlated radical pairs upon light excitation and have
therefore been suggested to act as the receptor molecules in such a radical-pair-based
magnetoreceptor. They have been shown to be expressed in the retinas of various birds, although
their exact location is still a matter of debate.
In my talk I will first give an overview of the behavioural, physiological and anatomical evidence for
such a cryptochrome- and radical-pair-based magnetic compass in various animals. I will then
discuss recent findings from my lab showing that the magnetic compass of zebra finches depends
not only on the wavelengths, but also on the polarization of light. I will further show that magnetic
compass orientation is affected by low-intensity, oscillating radio-frequency (RF fields at 0.1-10 MHz;
<1 μT) electromagnetic fields, which are believed to influence the interconversion between the
excited states of the radical pairs, and thereby alter or eliminate the effects produced by the Earth’s
magnetic field. I will conclude by presenting and discussing the consequences of novel findings on
the temporal expression, distribution, and cellular localization of different cryptochromes in the
retinas of zebra finches.

P1-2 [10:00]
Cryptochrome and mechanisms related to RF and ELF magnetic fields
Daniel R. Kattnig1
1Living Systems Institute and Department of Physics, University of Exeter, Exeter, UK, EX4 4QD

Biographical sketch
Daniel Kattnig is a Senior Lecturer at the recently established cross-
disciplinary Living Systems Institute at the University of Exeter. His
current research activity is in the field of quantum biology or, more
precisely, theoretical and experimental spin chemistry, an
interdisciplinary field dedicated to the effects of weak magnetic fields
on chemical reactions and biological processes. The theoretical
modelling of avian magnetoreception, which has been hypothesised
to rely on quantum phenomena involving transient radical pairs, is at
the heart of his current research endeavours. Daniel completed his
PhD in Physical Chemistry at the University of Technology Graz,
Austria, on studies of photo-induced electron transfer phenomena
(Prof. Grampp). It is during this period that he first delved into the
study of magnetic field effects on charge recombination reactions.
Following a short period in industrial research, Daniel joined the Max
Planck Institute for Polymer Research, Mainz, Germany, where he focused on the investigation of
soft matter by means of pulsed electron paramagnetic resonance (Prof. Spiess). Returning to Graz,
he took on a research position dedicated to magnetic field effects on donor acceptor systems, which
he held until 2013 when he joined the group of Prof. Hore at the University of Oxford. In 2017 Daniel
eventually moved to the Living Systems institute at Exeter, where he is setting up a group dedicated
to the theoretical and experimental investigation of magnetic field effects.
Five relevant publications
1. Kattnig, D. R., Radical-Pair-Based Magnetoreception Amplified by Radical Scavenging:
Resilience to Spin Relaxation. J. Phys. Chem. B 2017, 121 (44), 10215-10227.

5
 2. Kattnig, D. R.; Hore, P. J., The sensitivity of a radical pair compass magnetoreceptor can be
significantly amplified by radical scavengers. Sci. Rep. 2017, 7 (1), 11640.
3. Hiscock, H. G.; Kattnig, D. R.; Manolopoulos, D. E.; Hore, P. J., Floquet theory of radical pairs in
radiofrequency magnetic fields. J. Chem. Phys. 2016, 145 (12), 124117.
4. Kattnig, D. R.; Sowa, J. K.; Solov'yov, I. A.; Hore, P. J., Electron spin relaxation can enhance the
performance of a cryptochrome-based magnetic compass sensor. New J. Phys. 2016, 18,
063007.
5. Kattnig, D. R.; Evans, E. W.; Dejean, V.; Dodson, C. A.; Wallace, M. I.; Mackenzie, S. R.;
Timmel, C. R.; Hore, P. J., Chemical amplification of magnetic field effects relevant to avian
magnetoreception. Nat. Chem. 2016, 8 (4), 384-391.
Abstract
The rates and yields of radical pair reactions are often sensitive to weak magnetic fields comparable
to, or even weaker than, the geomagnetic field (approximately 50 µT). This trait is truly remarkable,
because the interaction energies arising from such weak magnetic fields amount to only a tiny
fraction of the thermal energy per molecule. The underlying mechanism, the so-called radical pair
mechanism (RPM), is of fundamentally quantum-mechanical origin. It has been suggested that
migratory birds, among other species, can sense the Earth's magnetic field by means of the RPM
dictating the fate of photochemically-formed radical pairs in cryptochrome proteins located in their
retinas. Experimental and theoretical support for a radical pair mechanism of compass
magnetoreception is accumulating although it is not yet clear whether cryptochrome is actually the
sensor. The RPM has also been implicated in the context of possibly adverse effects of weak
magnetic fields on human health. The underlying mechanisms are poorly understood, and currently
the subject of extensive research. It is, however, hypothesised that the alteration of radical pair
recombination rates, due to applied magnetic fields, may interfere with circadian rhythms or cause
an increase in free radical concentrations thereby boosting oxidative stress, genotoxic effects and
apoptosis.
In this talk, I will give a short introduction to the foundations of the radical pair mechanism.
Thereafter I will focus on our recent results in the field: I will describe how, under conditions of
continuous excitation, magnetic field effects can be amplified by slow termination reactions of the
radical intermediates. Realistic biological environments are turbulent, and the interactions of such
environments with the spin systems under considerations give rise to dephasing. With this in mind, I
shall discuss the surprising result that certain spin relaxation mechanisms can significantly enhance
the anisotropy of magnetic field effects. I will continue by elaborating on the recent hypothesis that
three-radical systems can exhibit magnetic field effects vastly exceeding those predicted by the
classical RPM. This effect could also provide a new perspective for our understanding of the
putative effects of electromagnetic fields on (human) biology.

6
 Session: P2
Plenary 2: The International Commission on Non-Ionising Radiation Protection (ICNIRP)
Draft High Frequency (100 kHz – 300 GHz) Guidelines
Tuesday June 26, 2018 • 09:30 - 10:30
Emerald
Chairs: Andrew Wood & Carl Blackman

P2-1 [09:30]
The International Commission on Non-Ionizing Radiation Protection (ICNIRP) Draft
Radiofrequency (100 kHz – 300 GHz) Guidelines
Rodney Croft1, 2
1School of Psychology, Illawarra Health and Medical Research Institute, University of Wollongong,
Wollongong, Australia, 2522
2International Commission on Non-Ionizing Radiation Protection, N/A, Australia

Biographical sketch
Rodney Croft is Professor of Health Psychology within the School of
Psychology, University of Wollongong, where he leads a
bioelectromagnetics research group at the Illawarra Health and
Medical Research Institute (IHMRI). His doctoral research was in
neurophysiology methodology, where he assessed and improved
models for accounting for ocular voltage contamination of the
electroencephalogram (EEG), followed by postdoctoral work utilizing
his EEG expertise to explore the relation between neurochemistry
and psychological/psychiatric function. Croft began researching
effects of radiofrequency fields on humans in 2000, with the focus on
potential detrimental effects on humans using the EEG. His
bioelectromagnetics work now includes a range of methodologies,
with as the unifying theme an attempt to understand the effect of
non-ionizing radiation on human health and well-being. Croft has led
Australian government-funded Centres of Research Excellence in this area since 2005, has been on
the Main Commission of the International Commission on Non-Ionizing Radiation Protection
(ICNIRP) since 2014, and chairs the ICNIRP Project Group developing its new radiofrequency
guidelines.
Abstract
ICNIRP’s current guidelines for the radiofrequency (RF; 100 kHz – 300 GHz) portion of the
electromagnetic spectrum were published in 1998. These have been very influential and form the
basis of numerous international radiation Standards. However, given the large body of relevant
research that has been generated since those 1998 guidelines and the associated development of
knowledge, ICNIRP is now revising these guidelines.
The revision of the guidelines has been underway since 2014. ICNIRP takes the RF-health scientific
literature as the basis for these guidelines, and from this is developing a system of limits that will
protect people from all health effects arising from RF exposure (including both thermal and non-
thermal effects). ICNIRP aims to make the rationale behind the guidelines as transparent as
possible, and to this end will release a public consultation version of the guidelines to encourage
input that will help inform the final guidelines. This will ensure that relevant literature is not missed
and that all perspectives are appropriately considered. ICNIRP plans to release its public
consultation document in 2018, and as part of this process it believes that BioEM 2018 would be an
ideal forum for the initial dissemination of the draft document. Accordingly, this presentation will
describe the ICNIRP RF Guidelines setting process, including its rationale, and the draft limits
themselves. It would also outline the consultation process, how interested parties can contribute to
7
this, and provide considerable opportunity for discussion with the floor.

Session: P3
Plenary 3: Exploring Dielectrophoresis and its Applications in the Biomedical Sciences
Wednesday June 27, 2018 • 10:30 - 11:30
Emerald
Chairs: Myles Capstick & Caterina Merla

P3-1 [10:30]
Dielectrophoresis: present and potential biomedical applications
Ronald Pethig1
1School of Engineering, The University of Edinburgh, Edinburgh, UK, EH9 3JF
Biographical sketch
Ronald Pethig is Emeritus Professor of Bioelectronics at the School
of Engineering, University of Edinburgh. He received training in
electrical engineering and physical chemistry and has enjoyed many
years working with cell biologists and biomedical engineers (e.g., as
an adjunct scientist since 1982 at the Marine Biological Laboratory,
Woods Hole; adjunct professor of physiology at the Medical
University of Charleston, 1984-88). His main research interest has
been the dielectric and electrochemical properties of biological
materials, and in particular the development of the electrokinetic
method of dielectrophoresis for characterizing and manipulating cells
and other bioparticles for biomedical applications. In 2001 he
received the Herman P Schwan Award for work in biodielectrics, and
in 2017 the Lifetime Achievement Award of the American
Electrophoresis Society (AES). Amongst other activities he currently
serves as Editor-in-Chief of IET Nanobiotechnology and enjoys beekeeping.
Five relevant publications
1. Muratore, M., Srsen, V., Waterfall, M., Downes, A. and Pethig, R. (2012) Biomarker-free
dielectrophoretic sorting of differentiating myoblast multipotent progenitor cells and their
membrane analysis by Raman spectroscopy, Biomicrofluidics 6, 034113.
2. Velugotla, S., Pells, S., Mjoseng, H. K., Duffy, C. R. E., Smith, S., De Sousa, P. and Pethig, R.
(2012) Dielectrophoresis based discrimination of human embryonic stem cells from
differentiating derivatives, Biomicrofluidics 6, 044113.
3. Pethig, R. (2017) Review - Where is dielectrophoresis (DEP) going? J. Electrochem. Soc., 164
(5): B3049-55.
4. Pethig, R., (2017) Dielectrophoresis: Theory, Methodology and Biological Applications, John
Wiley & Sons, Chichester.
5. Chung, C., Pethig, R., Smith, S. and Waterfall, M. (2017) Intracellular potassium under osmotic
stress determines the dielectrophoresis cross-over frequency of murine myeloma cells in the
MHz range, Electrophoresis, doi: 10.1002/elps.201700433.
Abstract
Dielectrophoresis (DEP) manifests itself as the motion of a particle induced by exposing it to an
electric field gradient. More than 50 years ago Pohl and Hawk (Science, 152: 647-9, 1966)
demonstrated that this purely physical effect can be used to simultaneously distinguish and separate
live and dead cells. Since then DEP manipulations of DNA, proteins, viruses, bacteria, blood cells,
cancer cells and stem cells have been demonstrated. In this talk an outline will be given of the

8
relationship between the DEP behavior of a bioparticle and its physico-chemical properties and,
where appropriate, its physiological state. The present status and possible ongoing trends of
biomedical applications of DEP will also be outlined. At least six commercial products incorporating
DEP can be cited (Panasonic’s bacteria counter; Shimadzu’s nanoparticle analyzer; the
DEPArrayTM system developed by Silicon Biosystems; DEPtech’s 3DEP system; ApoStreamTM
technology for isolating tumor cells from peripheral blood; devices developed by Biological
Dynamics for isolating nanoparticulate biomarkers from blood).
The cells themselves will “do the talking” for most of this presentation - by way of real-time videos -
to reflect the intention of making the talk amenable to a multi-disciplinary audience.

Session: P4
Plenary 4: Frontiers of Electroporation, from Mechanisms to Applications
Friday June 29, 2018 • 08:30 - 09:30
Emerald
Chairs: Micaela Liberti & Damijan Miklavčič

P4-1 [08:30]
Advances in electroporation and activation of cells by nanosecond stimuli
Andrei Pakhomov1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

Biographical sketch
Andrei Pakhomov is a Research Professor and an Interim Director of
the Center for Bioelectrics. He has broad interest in nanosecond
pulsed electric fields (nsPEF), from molecular and cellular
mechanisms to prospective applications in cancer ablation, deep
tissue stimulation, and defibrillation. He leads research on
nanoelectropore opening and life cycle, conduction properties, cell
stimulation and activation, membrane repair, and cell death
mechanisms and pathways. Pakhomov’s lab recently uncovered
bipolar cancellation, a phenomenon that challenges existing
electroporation paradigms and may enable targeted remote
biostimulation. His ongoing research is supported by a multi-
university MURI grant from AFOSR and an RO1 grant from NIH.
Five relevant publications
1. Pakhomov AG, Semenov I, Casciola M, Xiao S: Neuronal excitation
and permeabilization by 200-ns pulsed electric field: An optical membrane potential study with
FluoVolt dye. Biochim Biophys Acta 2017, 1859(7):1273-1281.
2. Gianulis EC, Labib C, Saulis G, Novickij V, Pakhomova ON, Pakhomov AG: Selective
susceptibility to nanosecond pulsed electric field (nsPEF) across different human cell types. Cell
Mol Life Sci 2017, 74(9):1741-1754.
3. Casciola M, Xiao S, Pakhomov AG: Damage-free peripheral nerve stimulation by 12-ns pulsed
electric field. Sci Rep 2017, 7(1):10453.
4. Pakhomov AG, Gianulis E, Vernier PT, Semenov I, Xiao S, Pakhomova ON: Multiple
nanosecond electric pulses increase the number but not the size of long-lived nanopores in the
cell membrane. Biochim Biophys Acta 2015, 1848(4):958-966.
5. Pakhomov AG, Semenov I, Xiao S, Pakhomova ON, Gregory B, Schoenbach KH, Ullery JC,
Beier HT, Rajulapati SR, Ibey BL: Cancellation of cellular responses to nanoelectroporation by
reversing the stimulus polarity. Cell Mol Life Sci 2014, 71(22):4431-4441.

9
Abstract
Early research into nsPEF bioeffects focused primarily on cell killing and cancer ablation. More
recently, interest shifted towards utilizing unique properties of nsPEF for activation and stimulation,
including targeted stimulation remotely from electrodes. Cell activation is accomplished by
nanoelectroporation and Ca2+ mobilization, which mimics opening of Ca2+ channels even in cells
which do not express them. If voltage-gated channels are present, they can be activated by either
membrane depolarization due to nanoelectroporation or directly, without membrane injury. However,
the latter mechanism is challenging since stimuli are much shorter than the time needed for
translocation of the molecular voltage sensor domain which opens the channel. On many occasions,
stimulation thresholds were equal to or above electroporation thresholds. Nonetheless, we
accomplished peripheral nerve stimulation by 10-ns PEF without observable membrane injury, even
after 50,000 stimulation cycles. This stimulation is subject to strong bipolar cancellation, which (a)
can be utilized to reveal ion channel kinetics, (b) explains, at least in part, why emitted microwaves
and RF have low biological efficiency even at high peak power, and (3) underlies a new paradigm of
science-based targeted remote stimulation, with broad medical applications.

P4-2 [09:00]
Unraveling new key molecular level aspects using computational chemistry
Mounir Tarek1
1Centre National de La Recherche Scientifique (CNRS), Université de Lorraine, Nancy, France

Biographical sketch
Mounir Tarek is a Senior Research Director at the CNRS-Université de
Lorraine. His research involves the use of computational chemistry
methods to study membranes, proteins, ion channels and membrane
transport proteins. Over the last few years, he studied many aspects of
electroporation of cell membranes subject to high electric fields. M. Tarek
is a founding member and a member of the Scientific Council of the
Europeen Associated Laboratory EBAM ‘Pulsed Electric Fields
Applications in Biology and Medicine’.
Five relevant publications
1. Casciola, M.; Kasimova, MA.; Rems, L.; Zullino, S.; Apollonio, F.; Tarek,
M. Properties of lipid electropores I: Molecular dynamics simulations of
stabilized pores by constant charge imbalance. Bioelectrochemistry 109,
(2016) 108-116
2. Casciola, M.; Tarek, M. A molecular insight into the electro-transfer of small molecules through
electropores driven by electric fields Biochim. Biophys. Acta. Biomembranes, 1858, 2278-2289
(2016)
3. Polak, A.; Bonhenry, D.; Dehez, F.; Kramar, P.; Miklavčič, D. and Tarek, M. On the
Electroporation Thresholds of Lipid Bilayers: Molecular Dynamics Simulation Investigations. J.
Membr. Biol. 246, 843-850 (2013)
4. M. Breton, L. Delemotte, A. Silve, L.M. Mir, M. Tarek, Transport of siRNA through Lipid
Membranes Driven by Nanosecond Electric Pulses: An Experimental and Computational Study,
J. Am. Chem. Soc. 134 (2012) 13938–13941.
Abstract
The application of short and intense electric pulses enables to transiently alter the properties of cell
membranes, making them permeable to a wide range of chemical species. This phenomenon is
routinely used in medical applications as well in biotechnology and industrial processing. Most
investigations to date of the processes involved have focused on the ability of intense electric fields
to create pores within the lipid bilayers, allowing us to better understand and control the

10
phenomenon termed “electroporation”. Our knowledge on the other hand, about the chemical
processes enhanced as a consequence of the application of electric fields to cells is still sketchy. In
this contribution we harness the capabilities of computational resources and the predictive power of
advanced atomistic and quantum level molecular dynamics techniques to decipher key steps in
several chemical and biophysical processes occurring during and following Electric field stimulations
of cell membranes. We show that under low-voltage conditions, and predict that under sub-
nanosecond pulse electroporation conditions, peroxidation of model cell membranes by potent
reactive oxygen species (OH• and OOH•) is significantly enhanced. We quantify the permeability of
the peroxidized membranes to a host of species including ions and molecules, to demonstrate that
electrically mediated chemical effects may play a significant role in several processes following
exposure of cells to high electric fields. We discuss the relevance of these effects for cells subject to
radio-frequency electromagnetic fields (RF-EMF) as well as for excitable cells subject to electro-
stimulation.

Session: P5
Plenary 5: Hot Topic
Friday June 29, 2018 • 11:30 - 12:30
Emerald
Chairs: Martin Röösli & Isabelle Deltour

11
 Session: S01
Human studies I / brain
Monday June 25, 2018 • 11:00 - 12:30
Europa C
Chairs: Sarah Loughran & Rene De Seze

S01-1 [11:00]
STUDENT PAPER
Determining the functional consequences of the RF-EMF sleep EEG effect
Sheridan Findlay1, 2, Rodney Croft1, 2, 3 & Sarah Loughran1, 2, 3
1Australian Centre for Electromagnetic Bioeffects Research, University of Wollongong, Wollongong,
Australia, 2522
2School of Psychology, Illawarra Health & Medical Research Institute, University of Wollongong, Wollongong,
Australia, 2522
3Population Health Research on Electromagnetic Energy, Monash University, Melbourne, Australia, 3800
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Sheridan Findlay
Although RF-EMF has been shown to affect sleep spindles (as measured by the sleep EEG), it is
not known whether there is a functional consequence of this. This presentation will report on interim
results of a sleep EEG study designed to determine whether RF-EMF affects memory consolidation
during sleep, a cognitive ability related to sleep spindles.
Introduction
To date the scientific consensus is that there are no harmful effects to human health resulting from
exposure to radiofrequency electromagnetic fields (RF-EMF) produced by mobile phones. However,
subtle RF-EMF induced changes to neural function, as measured by the electroencephalograph
(EEG), have been reliably demonstrated, such as changes in the spindle frequency range during
non-rapid eye movement (NREM) sleep (Loughran, et al., 2005; Loughran, et al., 2012; Regal et al.,
2007; Schmid et al., 2012). The functional consequence, if any, of this increase in spindle frequency
activity following exposure is not yet known. However, as sleep spindles are associated with
declarative memory consolidation and learning it raises the possibility that exposure to RF-EMF
prior to sleep could have an impact upon memory consolidation. The current literature illustrates that
an increase in spindle activity during NREM sleep results in significant improvements in recall
performance (Gais & Born, 2004; Gais, Molle, Helms, & Born, 2002; Walker & Stickgold, 2006;
Studte, Bridger & Mecklinger, 2017). Considering that RF-EMF is known to increase spindle
frequency activity, it may be that RF-EMF has a positive influence on memory consolidation.
Memory consolidation is a crucial everyday function and even small changes, positive or negative,
could have significant impacts. Therefore, the purpose of the present study was to determine
whether RF-EMF exposure immediately prior to sleep affects memory consolidation. Furthermore,
repeated testing within individuals was used in order to strengthen the methodology by enabling
internal replications.
Method
Twelve participants took part in a double blind, randomized, partially counterbalanced 5-night
protocol (full counterbalancing could not be achieved given that this was only a subset of the total
participants that will be recruited). Recruitment is ongoing and the total number of participants will
be 36, with equal numbers of female and male participants; data from additional participants will be
presented at BioEM2018. The protocol consisted of an adaptation night (to adapt to the sleeping
environment), followed by four exposure nights, which were either sham (no exposure) or active (2
W/kg 10g average SAR). The two sham and active exposures were grouped into two sets of pairs.
Exposure duration was 30 minutes, which concluded 15 minutes prior to sleep opportunity. Memory

12
tasks conducted included a word-pair association task, digit span forward task, and digit span
backward task. These tasks were completed prior to exposure, during the last 10 minutes of
exposure, and on the following morning 15 minutes after waking (allowing memory consolidation to
be measured). EEG was recorded continuously during sleep and manually scored for sleep stages
and architecture. Quantitative EEG analyses were conducted on the four exposure nights to
examine sleep spindle activity in the first NREM period of sleep.
Results
Results revealed slightly higher memory consolidation scores for the active condition compared to
the sham condition. The spindle frequency activity was also slightly higher in the active condition in
comparison to the sham condition. There were no changes seen between exposure conditions for
sleep efficiency or sleep architecture. The mean values and effect sizes for memory tasks, sleep
architecture, and sleep spindle EEG power for both exposure conditions are shown in Tables 1 and
2. Effect sizes and not inferential statistics are shown as the latter are not appropriate at present
given the small sample size (i.e. it would be underpowered).
Table 1. Comparison of exposure on memory consolidation and sleep spindle activity.
Std. Dev. Std. Dev.
Comparison Mean Sham Mean Active Effect Size
Sham Active

Word Pair
.57 .20 .59 .19 .14
Task

Spindle
Frequency .61 .18 .63 .11 .17
Power
Table 2. Comparison of exposure on sleep quality and sleep architecture
Mean Sham Std. Dev. Mean Active Std. Dev.
Comparison Effect Size
(min) Sham (min) Active

Total Sleep
418.9 132.8 450.5 19.14 .32
Time

Sleep
94.9 3.34 94.0 4.10 .24
Efficiency

Sleep Latency 19.9 12.34 27.6 24.94 .07

REM Latency 82.8 19.22 79.4 28.70 .14

Stage 1 Sleep 14.7 13.11 17.5 11.26 .23

Stage 2 Sleep 230.4 26.06 223.9 25.04 .26

Slow Wave
91.2 23.63 88.5 27.60 .10
Sleep

REM Sleep 120.9 24.15 122.9 25.17 .08

WASO 8.4 12.26 9.3 10.93 .08

Discussion

13
Analysis is ongoing and has currently only been completed for the first set of exposures. These
preliminary results suggest that RF- EMF exposure prior to sleep may have an effect on memory
consolidation based on the small differences in performance on the word pair task in this
experiment. It appears from this data set that participants’ memory consolidation was slightly better
after 30 minutes of active exposure compared to 30 minutes of no exposure prior to sleep.
Additionally, there was a slight increase in spindle frequency activity in NREM sleep following active
exposure in comparison to no exposure. The findings from the first set of exposures are in
agreement with previous research in that they found increases in spindle frequency activity in NREM
sleep following RF-EMF exposure. This result is in the expected direction, and together with the
memory consolidation data suggests that one functional consequence of the RF-EMF effect on
sleep spindle activity could be memory consolidation. The pilot data also show that the word pair
task is an appropriate task for this experiment as memory consolidation was shown to occur.
However, due to a small and incomplete sample size, which is not yet fully counter balanced, no
conclusions can be drawn in relation to the preliminary results, nor can comments regarding
statistical significance be made (as the analyses are currently underpowered). As analyses are
continuing, the above results will be supplemented with data from additional participants, as well as
the repeat testing results, for the BioEM2018 presentation.
Conclusion
The current pilot sample raises the possibility that memory consolidation is affected by RF-EMF
exposure. However, more data is needed before conclusions can be drawn.
References
ARPANSA Radiofrequency Expert Panel Review of Radiofrequency Health Effects Research –
Scientific Literature 2000-2012, ARPANSA Technical Report No. 164
Croft, R. J., Chandler, J. S., Burgess, A. P., Barry, R. J., Williams, J. D., & Clarke, A. R. (2002).
Acute mobile phone operation affects neural function in humans. Clinical Neurophysiology, 113,
1623-1632. doi: 10.1016/s1388-2457(02)00215-8
Croft, R. J., Hamblin, D. L., Spong, J., Wood, A. W., McKenzie, R. J., & Stough, C. (2008). The
effect of mobile phone electromagnetic fields on the alpha rhythm of human electroencephalogram.
Bioelectromagnetics, 29, 1-10. doi: 10.1002/bem.20352
Croft, R.J., Leung, S., McKenzie, R.J., Loughran, S.P., Iskra, S., Hamblin, D.L., Cooper, N.R., 2010.
Effects of 2G and 3G Mobile Phones on Human Alpha Rhythms: Resting EEG in Adolescents,
Young Adults, and the Elderly. Bioelectromagnetics 31, 434-444.
Gais, S., Born, J., 2004. Declarative memory consolidation: mechanisms acting during human sleep.
Learn Mem 11, 679-685.
Gais, S., Molle, M., Helms, K., Born, J., 2002. Learning-dependent increases in sleep spindle
density. J Neurosci 22, 6830-6834.
Loughran, S., Wood, A., Barton, J., Croft, R., Thompson, B., & Stough, C. (2005). The effect of
electromagnetic fields emitted by mobile phones on human sleep. Neuroreport, 16, 1973 - 1976.
Loughran, S.P., McKenzie, R.J., Jackson, M.L., Howard, M.E., Croft, R.J., 2012. Individual
differences in the effects of mobile phone exposure on human sleep: Rethinking the problem.
Bioelectromagnetics 33, 86-93.
Regel, S.J., Tinguely, G., Schuderer, J., Adam, M., Kuster, N., Landolt, H.P., Achermann, P., 2007.
Pulsed radio-frequency electromagnetic fields: dose-dependent effects on sleep, the sleep EEG and
cognitive performance. Journal of Sleep Research 16, 253-258.
Schmid, M.R., Loughran, S.P., Regel, S.J., Murbach, M., Bratic Grunauer, A., Rusterholz, T.,
Bersagliere, A., Kuster, N., Achermann, P., 2012. Sleep EEG alterations: effects of different pulse-
modulated radio frequency electromagnetic fields. J Sleep Res 21, 50-58.
Schmid, M.R., Murbach, M., Lustenberger, C., Maire, M., Kuster, N., Achermann, P., Loughran, S.P.,
2012. Sleep EEG alterations: effects of pulsed magnetic fields versus pulse-modulated radio
14
frequency electromagnetic fields. J Sleep Res 21, 620-629.
Studte, S., Bridger, E., & Mecklinger, A. (2017). Sleep spindles during a nap correlate with post
sleep memory performance for highly rewarded word-pairs. Brain and language, 167, 28-35.
Walker, M.P., Stickgold, R., 2006. Sleep, memory, and plasticity. Annu Rev Psychol 57, 139-166.

S01-2 [11:15]
The effect of RF-EMF exposure on the waking EEG: A comparison of effects across eyes
open and eyes closed resting EEG derivations
Anna Dalecki1, 2, 3, Sarah Loughran1, 2, 4, Adam Verrender1, 2, 4 & Rodney Croft1, 2, 3, 4
1School of Psychology, University of Wollongong, Wollongong, Australia, 2522
2Illawarra Health and Medical Research Institute, University of Wollongong, Wollongong, Australia, 2522
3Population Health Research of Electromagnetic Energy, Monash University, Melbourne, Australia
4Australian Centre for Electromagnetic Bioeffects Research, University of Wollongong, Wollongong,
Australia, 2522
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Anna Dalecki
RF-EMF (radio-frequency electromagnetic field) exposure is known to affect waking
electroencephalogram (EEG) alpha (8-12 Hz) spectral power during exposure. However, there is
variability within the literature with respect to the direction of this effect. Resting EEG can be derived
under eyes open or closed conditions, with alpha power higher when eyes are closed. Thus, if RF-
EMF increases EEG alpha power during exposure, this may be less likely to be detected using the
eyes closed endpoint due to ceiling effects. The present study aims to determine whether the effect
of RF-EMF exposure on waking EEG alpha power importantly differs between eyes open and
closed resting EEG endpoints.
Background
Radio-frequency electromagnetic field (RF-EMF) exposure is known to affect waking
electroencephalogram (EEG) alpha (8-12 Hz) spectral power during exposure. With respect to the
direction of this effect, the majority of studies report increased alpha power (Croft et al., 2008; Croft
et al., 2010; Curcio et al., 2005) however, there have also been reports of reductions (Ghosn et al.,
2015; Perentos, Croft, McKenzie, & Cosic, 2013) or failures to observe alpha power change during
RF-EMF exposure (Regel et al., 2007).
Resting EEG alpha power can be derived under conditions where the eyes are open or closed.
These differ in terms of alpha power: specifically, alpha power is reduced by 40% under eyes open
compared with eyes closed conditions (Barry, Clarke, Johnstone, Magee, & Rushby, 2007). Thus, if
the effect of RF-EMF is to increase EEG alpha power, this effect may be more apparent under eyes
open conditions where EEG alpha power is lower. Conversely, the use of eyes closed resting EEG
endpoint where EEG alpha power is relatively higher may increase the chance of type II error due to
ceiling effects. Whether the effect of RF-EMF exposure on waking EEG alpha power is importantly
affected by the choice of resting EEG endpoint (eyes open versus eyes closed) has not been
demonstrated and is the aim of the present study.
Method
Design
Thirty-six healthy subjects had their resting EEG recorded under eyes open and eyes closed
conditions, during and after exposure to a Global System for Mobile Communications (GSM)-like,
920 MHz signal with 10g peak-spatial specific absorption rate of 0 W/kg (Sham), 1 W/kg (Low RF)
and 2 W/kg (High RF) (Murbach, Christopoulou, Kuster, Crespo-Valero, & Achermann, 2012). A fully
randomised, counterbalanced, double-blind design was used.

15
Procedure
Participants were fitted with a 19-channel EEG cap (Quik-Cap, Compumedics, Neuroscan) and
seated inside a Faraday cage, facing a computer screen, with an RF antenna (inside a box)
positioned 42mm above the left ear canal and 115mm away from the head. A second box containing
no RF antenna was identically positioned on the right side of the head.
Resting EEG (eyes open and eyes closed; 2 minutes each) was recorded during: (a) a 23-minute
‘Baseline’ block (during which participants were not exposed to RF); (b) at the start and end of a 30-
minute RF-ON block (where exposure was according to the exposure conditions determined by the
randomisation and counterbalancing procedure); and (c) at the start and end of a 30-minute RF-
OFF block (where exposure was always off).
EEG Data Acquisition
EEG data were recorded from 19 channels (FP1, FP2, Fz, F3, F4, F7, F8, Cz, C3, C4, T7, T8, Pz,
P3, P4, P7, P8, O1, O2) and M2 according to the international 10/20 system. EEG data were
referenced to the left mastoid (M1) and grounded midway between FPz and Fz. EOG was recorded
from above and below the left eye and from the outer canthi of the left and right eyes. EEG and
EOG data were recorded with an online 0.05-500 Hz analogue band-pass filter, and digitised at a
rate of 2,000 Hz. All electrode impedances were below 5 kΩ at the start of the recording.
EEG Data Analysis
EEG data were down-sampled to 500 Hz, low pass filtered at 125 Hz (12dB/octave), re-referenced
to digitally linked mastoids, and EOG corrected using the revised aligned artefact average
automated ocular artefact reduction procedure (Croft & Barry, 2000).
EOG-corrected data were epoched into 2000 ms intervals, spline fit to 512 samples, and baseline
corrected to the entire interval (DC corrected). An automatic artefact rejection procedure identified
and rejected epochs containing signals exceeding ± 150 µV in amplitude in any EEG channel. A fast
fourier transform with a 20% Hanning window function was used to obtain amplitude frequency
spectra which were averaged in the ‘Alpha’ (8-12 Hz).
Statistical Analysis
Paired-samples t-tests will compare EEG alpha power (for eyes open and eyes closed derivations
separately) to determine whether firstly, there is an effect of Exposure (defined as the average of the
1W/kg and 2W/kg conditions), secondly, whether any effects are dose dependant, and thirdly,
whether the above differences differ as a function of eye condition.
Results
Data collection is complete and analysis is currently underway.
Discussion
This data will clarify whether the choice of eyes open versus eyes closed resting EEG endpoints
importantly affects results when evaluating the effect of RF-EMF exposure on waking EEG alpha
spectral power.
References
Barry, R. J., Clarke, A. R., Johnstone, S. J., Magee, C. A., & Rushby, J. A. (2007). EEG differences
between eyes-closed and eyes-open resting conditions. Clinical Neurophysiology, 118, 2765-2773.
doi: 10.1016/j.clinph.2007.07.028
Croft, R. J., & Barry, R. J. (2000). EOG correction of blinks with saccade coefficients: A test and
revision of the aligned-artefact average solution. Clinical Neurophysiology, 111(3), 444-451. doi:
10.1016/S1388-2457(99)00296-5
Croft, R. J., Hamblin, D. L., Spong, J., Wood, A. W., McKenzie, R. J., & Stough, C. (2008). The
effect of mobile phone electromagnetic fields on the alpha rhythm of human electroencephalogram.
Bioelectromagnetics, 29(1), 1-10. doi: DOI: 10.1002/bem.20352.

16
Croft, R. J., Leung, S., McKenzie, R. J., Loughran, S. P., Iskra, S., Hamblin, D. L., & Cooper, N. R.
(2010). Effects of 2G and 3G mobile phones on human alpha rhythms: Resting EEG in adolescents,
young adults, and the elderly. Bioelectromagnetics, 31(6), 434-444. doi: 10.1002/bem.20583
Curcio, G., Ferrara, M., Moroni, F., D’Inzeo, G., Bertini, M., & De Gennaro, L. (2005). Is the brain
influenced by a phone call? An EEG study of resting wakefulness. Neuroscience Research, 53,
265-270. doi: 10.1016/j.neures.2005.07.003
Ghosn, R., Yahia-Cherif, L., Hugueville, L., Ducorps, A., Lemarechal, J.-D., Thuroczy, G., . . .
Selmaoui, B. (2015). Radiofrequency signal affects alpha band in resting electroencephalogram.
Journal of Neurophysiology, 113, 2753-2759.
Murbach, M., Christopoulou, M., Kuster, N., Crespo-Valero, P., & Achermann, P. (2012). Exposure
system to study hypotheses of ELF and RF electromagnetic field interactions of mobile phones with
the central nervous system. Bioelectromagnetics, 33(6), 527-533. doi: 10.1002/bem.21710
Perentos, N., Croft, R. J., McKenzie, R. J., & Cosic, I. (2013). The alpha band of the resting
electroencephalogram under pulsed and continuous radio frequency exposures. IEEE Transactions
on Biomedical Engineering, 60(6), 1702-1710. doi: 10.1109/TBME.2013.2241059
Regel, S. J., Gottselig, J. M., Schuderer, J., Tinguely, G., Retey, J. V., Kuster, N., . . . Achermann, P.
(2007). Pulsed radio frequency radiation affects cognitive performance and the waking
electroencephalogram. Neuroreport, 18(8), 803-807.

S01-3 [11:30]
Are there differences in possible effects of RF-EMF exposure on the macro- and
microstructure of sleep between healthy young men (18-30 years) and healthy elderly women
(60-80 years)?
Heidi Danker-Hopfe1, Torsten Eggert1, Gernot Schmid2, Cornelia Sauter1 & Hans Dorn1
1Competence Center of Sleep Medicine, Charité - University Medicine Berlin, Berlin, Germany, 12200
2Seibersdorf Laboratories, Seibersdorf, Austria, 2444
Keywords: Human, RF/Microwaves, Completed (unpublished)
Presented by: Heidi Danker-Hopfe
The study aims to analyse possible differences in RF-EMF effects on the macrostructure and the
microstructure of sleep between healthy young men and healthy elderly women. The results indicate
differences in effects of GSM 900 and TETRA exposure in healthy young men (18-30 years) and
healthy elderly women (60-80 years). However, the results observed so far are not indicative of a
disturbed sleep under RF-EMF exposure.
Aim
To analyse possible differences in RF-EMF effects on the macrostructure and the microstructure of
sleep between healthy young men and healthy elderly women.
Methods
In three double-blind, randomized, sham-controlled cross-over studies effects of different RF-EMF
exposures on sleep were investigated in two different samples of young healthy male volunteers
and in one sample of elderly healthy female volunteers (see Table 1).
Table 1: Sample and exposure characteristics.

Age (years)
Sample Sex n mean ± SD Exposure psSAR10g Duration
Range

17
 GSM 900
25.3 ± 2.6 MHz 2.0 W/kg 8h
Sample 1 male 30
18 - 30 WCDMA 2.0 W/kg 8h
/UMTS

8 h + 30
25.4 ± 2.6 TETRA 1.5 W/kg min
Sample 2 male 32
20 - 30 TETRA 6.0 W/kg 8 h + 30
min

7.5 h + 30
GSM 900
67.8 ± 5.7 2.0 W/kg min
Sample 3 female 30 MHz
60 - 80 6.0 W/kg 7.5 h + 30
TETRA
min

The design of the studies was very similar. Exposure was delivered by a head worn antenna,
specially designed for the projects.
An adaptation night, which also served as a screening night for sleep disorders and as an
adjustment night to the laboratory environment, was followed by nine study nights in the laboratory
(three nights per condition), in which subjects were exposed to three different exposure conditions
(SHAM, and two verum exposures – see Table 1) in an individually randomized order.
Polysomnography was performed (assessment and evaluation) according to the standard published
by the American Association of Sleep Medicine (AASM, Iber et al. 2007), which uses information on
brain activity during sleep derived from frontal, central and occipital brain regions.
All nights were scored according to the standard rules. Power spectra were calculated for frequency
bands as defined by the International Pharmaco-EEG society (IPEG, Jobert et al. 2013).
Additionally, power spectra were calculated for a broad (11-16 Hz) and a narrow (12-14 Hz) range of
spindle frequencies as defined by Iber et al. (2007). Results of the analysis of RF-EMF effects on 27
variables characterising the macrostructure of sleep as well as on sleep stage specific power pectra
of the sleep EEG recorded from 19 electrodes will be presented.
Results
The studies allowed a direct comparison of effects of GSM 900 MHz (pSAR10g = 2 W/kg) and
TETRA (pSAR10g = 6 W/kg) on sleep of healthy young men and healthy elderly women. Based on
significant results (p < 0.05) and effect sizes, the sleep of elderly women is affected in more
variables under both verum exposure conditions than the sleep of young men. Furthermore, effects
are different for younger men and elderly women. Elderly women seem to sleep better under RF-
EMF as compared to sham.
Discussion
The results underline that findings obtained from selected samples cannot be generalized to the
overall population. Whether the observed effects are gender and/or age related will be answered
when the data of a study in elderly men, which is still ongoing, are available. In summary, however,
the results observed so far are not indicative of a disturbed sleep under RF-EMF exposure.

S01-4 [11:45]
STUDENT PAPER
Frequency specific magnetophosphene thresholds in humans exposed to ELF magnetic
fields

18
Cadence Baker1, 2, Michael Corbacio1, Sebastien Villard1, 4, Daniel Goulet5, Michel Plante5, Martine
Souques6, François Deschamps7, Genevieve Ostiguy5, Jacques Lambrozo6 & Alexandre Legros1, 2, 3, 4, 8
1Bioelectromagnetics and Human Threshold Research Group, Lawson Health Research Institute, London,
ON, Canada
2School of Kinesiology, Western University, London, ON, Canada
3Department of Medical Imaging, Western University, London, ON, Canada
4Department of Medical Biophysics, Western University, London, ON, Canada
5Hydro Quebec, Montreal, QC, Canada
6Service des études médicales, EDF, Paris, France
7Service des études médicales, EDF, Paris, France
8EuroMov, University of Montpellier, Montpellier, France
Keywords: Human, ELF/LF, Work in Progress
Presented by: Cadence Baker
Although magnetophosphene perception is the most reliably reported effect of exposure to
extremely low frequency (ELF) magnetic fields (MF), its frequency dependence needs to be further
documented as an indirect clue of the underlying mechanism. Ten healthy volunteers (60 planned)
were exposed to MFs from 20-300 Hz, with a flux density up to 80 mT. Phosphene perception
thresholds were reported and electroencephalography (EEG) was recorded. Establishing a reliable
magnetophosphene threshold frequency-response curve will allow further understand of the
mechanisms involved and will be of instrumental importance for ELF exposure recommendations.
Introduction
An alternating electric current generates, at a close proximity of its source, a time-varying magnetic
field (MF). Through magnetic induction, a time varying MF can induce electric fields and circulation
currents in an exposed conductor, such as the human body. There is evidence that induced electric
fields can change the membrane potential of nervous tissue to elicit a neural response (Attwell,
2003). Magnetophosphenes (colourless flickering lights perceived with the eyes closed in the dark)
are thought to be a result of magnetic induction in retinal cells, specifically the photoreceptors or
bipolar cells (Attwell, 2003), however numerous uncertainties still exist. For instance, although the
retina is thought to be responsible for magnetophosphene perception, the exact threshold for
perception is still uncertain and the specific cell type involved is unknown. Although
magnetophosphene perception thresholds are frequency dependent and these thresholds are lower
at lower frequencies (most sensitive at 20-30 Hz) and become higher for increasing frequencies
(Lovsund et al., 1980), this frequency response is light dependent, documented on only small
groups of volunteers and it has not been replicated yet.
Similar to magnetophosphenes, electrophosphenes are a visual perception occurring as a result of
electric stimulation and are also frequency dependent. Using transcranial alternating current
stimulation (tACS), phosphenes were reported to be strongest from 10-12 Hz (Kanai et al., 2008).
Abe (1951) quantitatively measured electrophosphene perception thresholds. Two frequency
response curves were found; one curve representing electrophosphene perception for a light-
adapted retina, and a second curve for a dark-adapted retina. Typically phosphenes are colourless,
however, using electrical stimulation coloured phosphenes have been reported in one study (Adrian,
1977). Achromatic phosphenes required the lowest stimulus strength around 25 Hz; although
requiring a stronger stimulus, coloured phosphenes were most prominent at 30 Hz (Adrian, 1977).
Supporting the retinal hypothesis, the phosphene threshold is lowest when electrodes are placed
closer to the retina, and further from the occipital cortex (Kar & Krekelberg, 2012).
International agencies use magnetophosphene thresholds as a means to generate ELF MF
exposure guidelines. The understanding of the magnetophosphene frequency response is critical
since guidelines are based on perception thresholds at power-line frequencies of 60 Hz (IEEE,
2002; ICNIRP, 2010), however the reference literature used to generate these guidelines is
experimentally limited below 50 Hz, using small group n-sizes and is not yet replicated (Lovsund et

19
al., 1980).
Magnetophosphene perception is a subjective report that could be completed by
electroencephalography (EEG), an objective measurement of brain electrical activity. Alpha activity
(8-12 Hz) has been recognized as the most prominent component of brain oscillations while awake
with the eyes closed, and decreases with visual perceptions (Berger, 1930; M. Teplan, 2002).
Measured using EEG, Zaehle et al. (2010) used alpha frequency tACS to elicit an increased alpha
power, providing evidence for extrinsically modulated cortical activity.
Objective
The primary objective of this study is to provide a frequency response curve in the ELF range (<300
Hz) for magnetophosphene perception in humans.
Hypothesis
We hypothesize that the threshold for magnetophosphene perception in the dark will be the lowest
at 30 Hz and will gradually increase as the stimulus frequency increases.
Methods
Participants
Ten healthy participants (mean age: 23.5 ± 3.01) were tested under this ongoing protocol involving a
self-adjustment of the MF level (final expected n=60). This protocol was approved by the Western
University Health Sciences Research Ethics Board (#108934).
Procedure
A global head exposure system, designed and built at the Lawson Health Research Institute, was
used throughout this experiment. This system consists of two 99-turn coils. A Helmholtz-like
arrangement enabled the creation of a homogenous MF about the participant’s head, up to 80 mT. A
motorized lift was designed to enable vertical movement of the exposure device such that it could
raise and lower centering the participants eyes between the coils.
Each experimental session lasted 2 hours. After EEG set up the participant was able to familiarize
themselves with the protocol (magnetophosphene perception samples delivered to the subjects until
they confirm they can clearly identify them). Data collection lasted 45 minutes.
Volunteers sat in a whole head MF exposure device wearing a 64-channel MRI-compatible EEG cap
(Compumedics, Charlotte, NC, USA). The protocol involved MF frequencies from 0-300 Hz, for a
total of 21 frequency conditions including: sham, every 5 Hz from 20 to 100 Hz, and then every 50
Hz from 100 to 300 Hz. MF frequency conditions were given in a blinded counterbalanced order.
Each frequency was presented twice, making a total of 44 exposures. Randomized throughout the
44 conditions were 6 additional exposures of 20 Hz, totaling 8 exposures with a frequency of 20 Hz
(used as an indication of the test/re-test reliability of our threshold estimation method), and 50
exposures overall. Each exposure lasted for 30 seconds. The participant had 25 seconds to self-
adjust the MF flux density using a handheld potentiometer to find their threshold value, which was
then steadily delivered for another 5 seconds to allow for EEG recording (Figure 1). Thresholds
were defined as the averaged MF value from the 2 repetitions of each condition. Based on classical
psychophysics and the method of self-adjustment, the lowest stimulus strength required for
magnetophosphene perception was deemed the threshold for that condition (Gescheider, 1997).
Results
A first one-way ANOVA for repeated measures (4 conditions – 20 Hz repeated 4 times) was
conducted to test for our test-retest reliability (Figure 2 - left). This analysis showed no significant
main effect (F(3,27)=0.917, p>0.05, η2=0.09, power= 0.06).
A second one-way ANOVA with repeated measures (21 conditions – 20, 25, 30, 35, 40, 45, 50, 55,
60, 65, 70, 75, 80, 85, 90, 95, 100, 150, 200, 250, 300 Hz) showed a main effect in mean thresholds
as a function of frequency (F(15,114)=22.72, p<0.0001, η2=0.75, power= 0.99). Post hoc
comparisons (Tukey adjusted for multiple comparisons) showed that the magnetophosphene

20
threshold is lowest at 35 Hz (24.59 ± 6.82 mT; Figure 2). All significant differences in mean
thresholds are reported in Table 1. Figure 2 illustrates the differences in 25, 30 and 35 Hz as those
frequencies differed from the largest quantity of frequencies (ie. 25, 30 and 35 Hz differed from 55,
60, 65, 70, 75, 80, 85 ,90, and 95 Hz). Magnetophosphenes were not perceived above 95 Hz
(Figure 3).
To date, EEG data has been collected from ten participants. EEG has not been treated in this
abstract, the results will be presented at BioEM 2018.
20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 95

20 * + + + +

25 ** + + + + + + + +

30 * + + + + + + + +

35 ** + + + + + + + +

40 ** + + + + + + +

45 * ** + + + +

50 * + + + +

55 + + + +

60 * + + +

65 + + +

70 + + +

75 ** + *

80 **

85

90

95

Table 1: Differences in mean threshold values across all frequencies (20-95 Hz). Representing
significance, p<.05 is *, p<.01 is **, and p<.001 is +.
Discussion
The test/re-rest procedure conducted at 20 Hz confirms the reliability of our threshold estimation
method.
The lowest threshold expressed as a function of flux density (and not as a function of dB/dt, which
would be proportional to the induced electric field in the retina) was found to be equivalent between
25 and 35 Hz, with the lowest value at 35 Hz (24.59 mTpeak or 17.39 mTrms). This threshold is 9 mT
higher than the lowest threshold previously reported (Lovsund et al., 1980). This difference likely
results from methodological differences. For example, the MF used is homogenous in the head
region, whereas it was heterogeneous in Lövsund’s study. The heterogeneous MF was such that the

21
medial aspect of the eye was exposed to a MF 20% lower than the lateral aspect of the eye,
perhaps underestimating the maximum exposure. Also, the perception threshold was evaluated as
the highest MF flux density without magnetophosphenes, whereas here it is lowest MF flux density
eliciting perception.
In addition confirming the most sensitive frequencies, we report for the first time perception
thresholds resulting from MF exposures at power-frequency (36.56 mTpeak or 25.85 mTrms at 50 Hz
and 43.73 mTpeak or 30.92 mTrms at 60 Hz), useful to further document the existing guidelines and
recommendations (IEEE, 2002; ICNIRP, 2010).
Another important observation is that magnetophosphenes were always reported to be white
(participants were asked to report their perception after each experimental session), which suggest
that pathways involving rod photoreceptor (achromatic) are most likely targeted. Photoreceptors
respond to stimulus with graded changes in membrane potential, which in this case is caused by
magnetic induction (Attwell, 2003). Rods respond slowly through the summation of stimuli, enabling
the detection of weak stimuli and are 100 times more sensitive than cones (Conner & MacLeod,
1977; Kandel, 1981).
In conclusion, the findings of this experiment address the primary objective of this study and support
the hypothesis that magnetophosphene perception will be lowest at 30 Hz. These results provide
confirmation of the frequency response dynamics with experimental data in humans for the first time
above 50 Hz. Although these threshold values are higher than those in previous literature, we can
attribute this to MF characteristics and threshold determination.
Acknowledgements
This work is supported by MITACS, Hydro-Québec, RTE, EDF, NationalGrid, Energy Network
Association, the EPRI and Lawson (IRF). Thanks to Lynn Keenliside for having designed and built
the exposure system.
References
Abe, Z. (1951). Influence of Adaptation on the Strength-Frequency Curve of Human Eyes, as
Determined with Electrically Produced Flickering Prosphenes. The Tohoku Journal of Experimental
Medicine, 54(1).
Adrian, D. J. (1977). Auditory and visual sensations stimulated by low-frequency electric currents.
Radio Science Supplement, 12(6).
Attwell, D. (2003). Interaction of low frequency electric fields with the nervous system: the retina as
a model system. Radiation Protection Dosimetry, 106(4).
Berger, H. (1930). Uber das Elektrenkephalogramm des Menschen. Zweite Mitteilung. J. Psycho.
Neurol., 40(1875).
Conner, J. D., & MacLeod, D. I. (1977). Rod Photoreceptors Detect Rapid Flicker. Science,
195(4279).
IEEE Standards Coordinating Committee 28. (2002). IEEE Standards for Safety Levels with
Respect to Human Exposure to Electromagnetic Fields, 0-3 kHz. IEEE, Inc.
International Commission on Non-Ionizing Radiation Protection. (2010). Guidelines for Limiting
Exposure to Time-Varying Electric and Magnetic Fields (1 Hz to 100 kHz).
Kanai, R., Chaieb, L., Antal, A., Walsh, V., & Paulus, W. (2008). Frequency-Dependent Electrical
Stimulation of the Visual Cortex. Current Biology, 18(23).
Kandel, E. R. (1981). Principles of Neural Science (4th ed.). NY: McGraw-Hill.
Kar, K., & Krekelberg, B. (2012). Transcranial electrical stimulation over visual cortex evokes
phosphenes with a retinal origin. Journal of Neurophysiology, 108(8).
Lovsund, P., Oberg, P. A., & Nilsson, S. E. G. (1980). Magnetophosphenes: a quantitative analysis
of thresholds. Medical & Biological Engineering & Computing, 18(3).

22
M. Teplan. (2002). FUNDAMENTALS OF EEG MEASUREMENT. Measurement Science Review,
2(2).
Zaehle, T., Rach, S., & Herrmann, C. S. (2010). Transcrania Alternating Current Stimulation
Enhances Individual Alpha Activity in Human EEG. PLoS One.
Figures

Figure 1. MF flux density, in mT, controlled with a handheld potentiometer. The participant was
instructed to increase the MF flux density until they clearly perceived phosphenes, and
decrease until phosphenes were barely noticeable (blue). The final 5 seconds of exposure
were kept constant at the participants phosphene threshold and EEG was recorded during this
time (red).

23
Figure 2. Magnetophosphene thresholds were calculated at 20 Hz four times (left). Average
frequency response (20-100 Hz) for magnetophosphene perception thresholds (right). Error
bars represent standard error of the mean (n=10).Thresholds are reported in peak values.

Figure 3. Percentage of exposures eliciting magnetophosphene perception (n=10).

24
S01-5 [12:00]
STUDENT PAPER
Frequency-dependent and montage-based differences in phosphene perception thresholds
via tACS
Ian Evans1, 2, 3, 4, Stephen Palmisano1, Sarah Loughran1, 2, 3, 4, Alexandre Legros5 & Rodney Croft1, 2, 3, 4
1School of Psychology, University of Wollongong, Wollongong, Australia, 2522
2Illawarra Health & Medical Research Institute, University of Wollongong, Wollongong, Australia, 2522
3Australian Centre for Electromagnetic Bioeffects Research, Wollongong, Australia
4Centre for Population Health Research on Electromagnetic Energy, Monash University, Melbourne, Australia
5Lawson Health Research Institute, Western University, Ontario, Canada
Keywords: Human, ELF/LF, Work in Progress
Presented by: Ian Evans
This study compared the thresholds for tACS-induced phosphene perception at fifteen frequencies
(2-30 Hz) when applied to either an FPz-Cz montage or an Oz-Cz montage. All combinations of
montages and frequencies were tested on the same population (n = 14, 24 upon completion of the
study). Repeated measures ANOVAs revealed consistently lower thresholds for the FPz-Cz
montage compared to the Oz-Cz montage, and differences in thresholds across frequencies. The
lowest stimulation levels required to induce phosphenes was found to occur at a frequency of 16 Hz
regardless of montage.
Introduction
Transcranial alternating current stimulation (tACS) is a non-invasive technique that allows
researchers and therapists to induce localised and controlled interference in the brain by generating
electrical fields at the scalp (Antal et al., 2008; Kanai et al., 2008; 2010). tACS applied at the lower
alpha to upper beta (8-30 Hz) range of frequencies near the eye often generates the perception of
flashing or shimmering light known as phosphenes (Kanai et al., 2008; Schutter & Hortensius,
2010). These effects were originally thought to be due to electrical stimulation of the retina (as their
intensity increases as the distance between the retina and the electrodes decreases - Rohracher,
1935; Schwarz, 1947). While Kanai and colleagues (2008) have shown that phosphenes can also
be induced by tACS applied over the occipital-visual cortex (V1), it is still unclear whether these
perceptions have a cortical or a retinal origin (e.g., volume conduction effects could spread through
the tissue from the occipitally-located electrode to the retina) (Schwiedrzick, 2009; via Laasko &
Hirata, 2013). While both interpretations are plausible and have data supporting them, previous
studies have usually examined the phosphenes generated in only one of these conditions. They
also had rather low statistical power as they tested small numbers of observers. Furthermore, past
conclusions were made based on comparisons of studies with varying (or unreported) lighting
conditions and stimulation frequencies. The aim of this study was to examine the relationship
between the thresholds for tACS-induced phosphenes delivered at a frontal FPz-vertex montage
and those at the occipital (V1) Oz-vertex montage. It systematically examined the phosphenes
generated by both types of delivery on the same participants and across a wide range of stimulation
frequencies (2-30Hz).
Method
16 out of a planned 24 participants (even gender split, aged 20-39, mean 27.9 years) have currently
completed this study after passing a modified safety screening checklist (Keel, Smith & Wasserman,
2000). tACS was delivered using a Magstim NeuroConn stimulator (The Magstim Company Ltd,
Carmarthenshire, UK) through conductive-rubber electrodes (dimensions: 3 x 4 cm cathode, 5.6 cm
diameter circle anode) placed on sponges saturated with a saline solution mixed with a
hypoallergenic amphoteric surfactant, and held in place at the appropriate montage with rubber
straps. All stimulations were sinusoidal with no DC offset.
The experiment involved two testing sessions in which stimulation was delivered to either a frontal

25
(centre of the cathode at FPz) or an occipital (cathode at Oz) montage. In both sessions, the centre
of the anode was positioned at the vertex (Cz). Session order (frontal/occipital) was
counterbalanced across participants, with each session taking ~90 min to complete. Participants
were seated on a chair in a dimly lit room with black curtains covering any walls that would appear in
their field of view (including peripheral vision), and were asked to keep their eyes open during
stimulations. Illumination levels in the participants’ field of view were measured at 6±0.05 cd/m2
using a J6523 Tektronix luminance probe (Tektronix Inc., Beaverton, Oregon).
Once the electrodes were in position, participants were told about the nature of phosphenes and
what they might perceive, whilst their skin and hair was saturated from the saline in the sponges.
Once impedance between electrodes was at 15 kΩ or below, participants were familiarized with the
appearance of phosphenes using ten seconds of 1000 μA tACS, firstly at 12 Hz, then at 22 Hz, to
demonstrate both the visual appearance of phosphenes and how they can change based solely on
varying the frequency of the stimulation. Fifteen frequencies were tested (2-30 Hz) in a
predetermined random order. Thresholds for phosphene perception were determined for each
frequency by varying the tACS current using a QUEST-based Bayesian staircasing procedure
(Watson & Pelli, 1983) with PsychToolbox (Kleiner, Brainard & Pelli, 2007) bound between 25 μA
and 1500 μA (starting at 700 μA), with a stepsize of 25 μA between stimulation levels. All
stimulations lasted for five seconds, and participants were informed when stimulation began and
when it ceased. Once the threshold for a frequency was determined, a new frequency was selected
and the same procedure as described above was applied until all frequencies had been completed.
Results
The results are shown in Figure 1. The lowest threshold for phosphene perception was found at
16Hz for both the Oz-Cz montage (mean = 357.35 μA, standard error = 37.97 μA) and the FPz-Cz
montage (mean = 121.43μA, SE = 9.95μA). A repeated measures ANOVA showed a clear
difference in mean threshold levels as a function of montage (F(1,15) = 183.628, p < .001, partial η2
= .924). Planned contrasts showed quadratic differences in mean threshold levels across different
frequencies (F(1,15 = 1247.125, p < .001, partial η2 = .988), and a quadratic interaction between
montage and stimulus frequency (F(1,15) = 12.667, p = .003, partial η2 = .458).
Discussion
Phosphenes generated with a frontal montage for tACS consistently required lower levels of
stimulation compared to the occipital montage. The frequency-dependence of both types of
phosphenes was also confirmed.
Phosphene thresholds reportedly depend on the ambient lighting conditions (Schwarz, 1947).
Previously the reported minima for phosphene threshold distributions was 20 Hz under lit conditions
(Lövsund, 1980; Kanai et al., 2008; Schwiedrzick, 2009; Laasko & Hirata, 2012). However, this
study found the minima to be 16 Hz for both FPz-Cz and Oz-Cz tACS montages. This is still within
the lower beta range of greater phosphene threshold sensitivity reported by Rohracher (1935), and
close to the 14-18 Hz reported peak frequency-dependent response rate of retinal ganglion cells to
electrical stimulation (Bernardete & Kaplan, 1999). This 16-20 Hz difference may seem trivial,
however when examining oscillatory signals in the cortex the functional difference between 16 Hz
and 20 Hz signals can be important.
One possible explanation for the absence of any 20 Hz minima is that light levels were too low
during testing, and that the threshold distribution more closely matches that expected in dark
conditions. Although ambient light levels in this study could be described as low (6 cd/m2), they
were still higher than the 1.2 cd/m2 illumination levels reported by Lövsund (1980) where he found
the 20 Hz minima in the threshold distribution. Given that testing in the present study began at least
10 minutes after entering the testing area, and that Kanai et al. (2008) reported a threshold minima
of 10 Hz under dark conditions, it is difficult to attribute the frequency-based difference to light
adaptation effects.

26
This issue of volume conduction was approached by halving the surface area of the cathode
compared to the anode, ensuring the current source density generated was greater at the intended
area of stimulation (Nitsche et al., 2008). Given the electrical properties of the head, this issue is
impossible to solve completely (Laasko & Hirata, 2013), however the similarities in both the minima
and the shape of the threshold distributions between montages suggests that the same mechanism
might be responsible for producing phosphenes in both conditions.
Conclusion
Our greater sample size and systematic frequency sampling (from 2-30 Hz, in steps of 2 Hz)
allowed us to examine the relationship between phosphenes generated by tACS more closely than
previous research, and with greater confidence in the statistical power of any relationships or
differences found. The difference in current source density required to induce phosphenes clearly
varies as a function of both tACS frequency and montage selection. Whilst the similarities in the
shape and minima of the threshold distributions indicate that phosphenes may be the result of
retinal stimulation regardless of montage, this remains to be confirmed.
References
Antal, A., Boros, K., Poreisz, C., Chaieb, L., Terney, D. & Paulus, W. (2008) Comparatively weak
after-effects of transcranial alternating current stimulation (tACS) on cortical excitability in humans.
Brain Stimulation, 1, 97–105.
Bernardete, E.A. & Kaplan, E. (1999). The dynamics of primate M retinal ganglion cells. Visual
Neuroscience, 16, 355-368.
Kanai, R., Chaieb, L., Antal, A., Walsh, V. & Paulus, W. (2008). Frequency-dependent electrical
stimulation of the visual cortex. Current Biology, 18, 1839–43.
Kanai, R., Paulus, W. & Walsh, V. (2010). Transcranial alternating current stimulation (tACS)
modulates cortical excitability as assessed by TMS-induced phosphene thresholds. Clinical
Neurophysiology, 121, 1551–4.
Kar, K. & Krekelberg, B. (2012). Transcranial electrical stimulation over visual cortex evokes
phosphenes with a retinal origin. Journal of Neurophysiology, 108, 2173–8.
Keel, J.C., Smith, M.J. & Wassermann, E.M. (2000). A safety screening questionnaire for
transcranial magnetic stimulation. Clinical Neurophysiology, 112, 720.
Kleiner, M., Brainard, D. & Pelli, D. (2007). What’s new in Psychtoolbox-3? Perception (ECVP
Abstract Supplement), 36.
Laasko, I. & Hirata, A. (2012) Computational analysis of thresholds for magnetophosphenes.
Physics in Medicine and Biology, 57, 6147-65.
Laasko, I. & Hirata, A. (2013). Computational analysis shows why transcranial alternating current
stimulation induces retinal phosphenes. Journal of Neural Engineering, 10, 046009.
Lövsund, P., Öberg, P.Å., Nilsson, S.E.G. & Reuter, T. (1980) Magnetophosphenes: a quantitative
analysis of thresholds. Medical and Biological Engineering and Computing, 18(3), 326–34.
Nitsche, M. A., Cohen, L. G., Wassermann, E. M., Priori, A., Lang, N., Antal, A., ... & Pascual-Leone,
A. (2008). Transcranial direct current stimulation: state of the art 2008. Brain Stimulation: Basic,
Translational, and Clinical Research in Neuromodulation, 1(3), 206-223.
Rohracher, H. (1935). Über subjective Lichterscheinungen bei Reizung mit Wechselströmen. Z.
Sinnesphysiology, 66, 164–81.
Schwarz, F. (1947) Über die elektrische Reizbarkeit des Auges bei Hell- und Dunkeladaptation.
Pflügers Arch. Eur. J. Physiol., 249, 76–86
Schutter, D.J. & Hortensius, R. (2010) Retinal origin of phosphenes to transcranial alternating
current stimulation. Clinical Neurophysiology, 121, 1080–4.
Schwiedrzick, C. (2009). Retina or visual cortex? The site of phosphene induction by transcranial

27
alternating current stimulation. Frontiers in Integrative Neuroscience, 3(6), 1-2.
Watson, A.B., & Pelli, D.G. (1983) QUEST: a Bayesian adaptive psychometric method. Perception &
Psychophysics, 33 (2), 113-20.
Figures

Figure 1. Mean thresholds for phosphene perception for frontal and occipital montages at each
frequency tested (error bars = standard error). Where no phosphene was perceived at the
maximum stimulation level (1500 μA), the maximum was substituted.

S01-6 [12:15]
Reconstruction of brain networks involved in magnetophosphene perception using dense
electroencephalography
Julien Modolo1, 2, Mahmoud Hassan1 & Alexandre Legros2, 3, 4, 5
1Univ Rennes, INSERM, LTSI – U1099, F-35000 Rennes, Rennes, France
2Human Threshold Research Group, Lawson Health Research Institute, London, ON, Canada
3Departments of Medical Biophysics and Medical Imaging, Western University, London, ON, Canada
4School of Kinesiology, Western University, London, ON, Canada
5EuroMouv, Université de Montpellier, France, France
Keywords: Human, ELF/LF, Completed (unpublished)
Presented by: Julien Modolo
OBJECTIVE: Characterizing functional brain networks in humans during magnetophosphene
perception. METHODS: Dense electroencephalography (EEG, 128 channels) was performed in N=3
volunteers during high-level (50 mT) magnetic field (MF) exposure. Functional brain networks were
reconstructed, at the cortical level from scalp recordings, using the EEG source connectivity

28
method. RESULTS: Magnetophosphene perception appears to consistently activate the right inferior
occipito-temporal pathway. CONCLUSIONS: This study provides the very first neuroimaging results
characterizing magnetophosphene perception in humans. The use of dense-EEG source
connectivity is a promising approach in the field of bioelectromagnetics.
Introduction
The characterization of extremely low-frequency (ELF, < 300 Hz) magnetic fields (MF) effects on
human brain activity classically uses electroencephalography (EEG), which has the advantage to be
non-invasive an relatively simple to set up. A popular analysis of EEG signals consists in computing
at the electrode (scalp) level the spectral power in various frequency bands (delta, 1-4 Hz; theta, 4-7
Hz; alpha, 8-12 Hz; and beta, 13-30 Hz). However, these studies often use a limited number of
electrodes (typically 32), which has limitations in terms of identifying subtle changes. Furthermore, a
growing number of studies have shown that even dense-EEG spectral power at the scalp level is not
a very sensitive indicator of underlying changes in neuronal networks activity. For example, a study
recently demonstrated in Parkinson’s disease patients that dense-EEG (128 channels) spectral
power alone did not discriminate patients in terms of cognitive deficit, while a recent technique
called EEG source connectivity could [Hassan et al, 2017]. This method is used in the current study.
The principle of EEG source connectivity is to move from the scalp (electrode) level to the cortical
(brain source) level and then to estimate the temporal relationships between these sources. This
results in the reconstruction of brain networks at the millisecond scale, which can be compared over
conditions, with the advantage of identifying specific brain structures involved in corresponding
effects [Hassan and Wendling, 2018]. In this proof-of-principle study, we aimed at evaluating if EEG
source connectivity could identify the brain networks associated with magnetophosphene
perception. Magnetophosphenes correspond to flickering lights seen in the absence of light during
exposure to a sufficiently high time-varying MF [Lövsund 1980]. Since magnetophosphene
perception has always relied on subjective reports from volunteers and never on neuroimaging
results, this would represent an insight in understanding the underlying mechanisms.
Magnetophosphene perception is the current base for setting the exposure limits in international
guidelines from the Institute of Electrical and Electronics Engineers (IEEE) and the International
Commission on Non-Ionizing Radiation (ICNIRP). One hypothesis is that magnetophosphenes
occur from a direct effect of the induced electric field on the retinal photoreceptors, which is
consistent with recent experimental results in humans [Legros et al. 2016]. Since
magnetophosphene is the most reliable and reproducible response of ELF exposure in humans, we
used the induction of magnetophosphene perception in humans to evaluate the potential of EEG
source connectivity in the field of bioelectromagnetics.
In this preliminary work, we enrolled 3 healthy volunteers who first underwent anatomical MRI, and
second dense-EEG (128 channels) during exposure to a 50 mT MF to identify the brain networks
involved while perceiving magnetophosphenes.
Materials and Methods
A total of N=3 healthy volunteers were recruited for this experiment, which was approved by the
Health Sciences Research Ethics Board (approval #17816) of Western University (London, ON,
Canada). Prior to the experiment itself, an anatomical MRI was acquired using a 3T PET/MRI
scanner (Biograph, Siemens, Munich, Germany) for further subject-specific EEG source
reconstruction. High-resolution EEG (128 channels) data was acquired using an MRI-compatible
cap (Compumedics-Neuroscan, Charlotte, NC, USA) at a sampling frequency of 10 kHz.
The exposure device used in this experiment (see [Keenliside et al. 2015] for details) was a water-
cooled Helmoltz-like system consisting in two coils (501 mms outer diameter each) powered by MRI
gradient amplifiers. The resulting magnetic flux density in the homogeneity zone, that includes the
eyeballs and the entire brain volume when the subject is in place, was 50 mTRMS.
Subjects were equipped with a 128 channel MRI-compatible EEG cap (Compumedics, Charlotte,
NC, USA) and the impedance was kept at or below 5 kOhms. They were then setup within the

29
exposure system so that the entire brain was located in the homogeneity zone of the exposure
device (± 5%). Lights were turned off just before beginning the experiment. Then, a total of 10
epochs consisting in a succession of 1-minute “real” and 1-minute “sham” exposure periods were
delivered while high-resolution EEG was continuously recorded. The order of “real” and “sham”
periods was counterbalanced for each of 10 epochs to prevent any possible order effect.
MRI data for each subject was segmented using Freesurfer (https://surfer.nmr.mgh.harvard.edu) to
provide a personalized 3-D brain anatomical model on which to perform EEG source reconstruction
later in the processing pipeline. Each 3D brain mesh was composed of 15,000 vertices. EEG data
was first filtered between 3 and 30 Hz, then re-referenced using the average reference in Brainstorm
(http://neuroimage.usc.edu/brainstorm/), a Matlab (The Mathworks, USA) package. Each individual
channel was visually inspected and compromised electrodes (obvious signal deviations or amplitude
> 80 μV) were removed from further analysis. Missing electrodes were interpolated from
neighbouring ones in a radius of 5 cm2 maximum using Brainstorm. Source reconstruction was then
performed in Brainstorm using the Minimum Norm Estimate method, with the constraint that
reconstructed sources were normal to the cortex. This resulted in the activity of each of the 15,000
sources composing the 3D-mesh of each subject’s brain. These sources were then pooled by
anatomical region using a 68-region brain atlas, providing the EEG time course for each of the 68
brain regions over the entire experiment. Using these reconstructed activity of each of the 68 brain
regions, Phase Locking Values (PLV) were computed between each pair of channels[Hassan et al.
2013]. Computing PLV values between each couple of brain regions resulted in a 68x68 matrix
accounting for the functional connectivity between regions. A threshold value was set at 10% of the
highest PLV values and only PLV values above this threshold were considered as significant. Here,
we preferred using a proportional threshold to absolute threshold to ensure equal density between
groups. See Kabbara et al. [2018] for more details about the choice of threshold values.
The identified networks were characterized using graph theory. This enables extracting local and
global topological properties of the networks. Here, we used the clustering coefficient to explore the
network segregation (local information processing). The clustering coefficient of a node represents
how close its neighbors tend to cluster together. It is defined as the proportion of connections among
its neighbors, divided by the number of connections that could possibly exist between them.
Results
Figure 1 reports the brain regions for which, at the group level, the clustering coefficient was
significantly different between the “rest” and “exposure” time periods.The results presents in Figure
1 have several implications: First, the most obvious modulation is at the level of the occipito-
temporal pathway in the right hemisphere. This modulation mostly dominating the right hemisphere
is consistent with known human neurophysiology, since visual stimuli with low spatial resolution (as
in magnetophosphene perception) activate preferentially the right visual pathway [Kauffmann et al.,
2014]. The occipito-temporal pathway is known as the ventral pathway of visual perception, also
known as the “What?” pathway aiming at identifying a visual perception [de Haan and Cowey, 2011].
Second, there is a consistent activation of the left fusiform cortex, a region from the inferior temporal
lobe involved in face recognition (and in discriminating between face vs. not a face). Third, brain
network modulationsduring magnetophosphene perception appear confined to visual pathways,
which is consistent with the hypothesis of a retinal origin, with a subsequent activation of the visual
cortex. Obviously, since this study is limited in terms of sample size, these results need to be
confirmed within a larger sample of human volunteers.
Conclusions
In previous studies, we used the MRI scanner to both deliver ELF exposure and perform functional
imaging [Legros et al. 2015, Modolo et al. 2017]. However, the maximal MF flux density possibly
generated by a standard 3T MRI scanner without compromising image quality is less than 8 mT, not
inducing magnetophosphene perception at 50 or 60 Hz, [Legros et al. 2016, Modolo et al., 2017]).
Functional MRI cannot therefore be used to image brain networks associated with
magnetophosphene perception. Furthermore, MRI only allows to acquire functional images with low

30
temporal resolution, on the order of the second. Conversely, EEG source connectivity tracks subtle
changes in brain networks at the millisecond scale, which makes it a promising tool in the
investigation of functional brain responses to ELF exposures and stimulations.
In summary, in this pilot study, we have provided the first direct neuroimaging results in humans
during magnetophosphene perception. One implication is that EEG source connectivity is a
promising neuroimaging modality that should provide novel and crucial insights into the fundamental
mechanisms by which induced electric fields interact with physiological brain activity in humans.
References
de Haan E.H., Cowey A. On the usefulness of "what" and "where" pathways in vision. Trends in
Cognitive Sciences, 15(10):460-466, 2011.
Hassan M., Chaton L., Benquet P., Delval A., Leroy C., Plomhause L., Moonen A.J.H., Duits A.A.,
Leentjens A.F.G., van Kranen-Mastenbroek V., Defebvre L., Derambure P., Wendling F., Dujardin K.
Functional connectivity disruptions correlate with cognitive phenotypes in Parkinson's disease.
Neuroimage Clinical 14:591-601, 2017.
Hassan M., Wendling F. Electroencephalography source connectivity: toward high time/space
resolution of brain networks. IEEE Signal Processing Magazine, in press, 2018.
Hermann C.S., Rach S., Neuling T., Strüber D. Transcranial alternating current stimulation: a review
of theunderlying mechanisms and modulation of cognitive processes. Frontiers in Human
Neuroscience 7:279, 2013.
Kabbara A., Eid H., El Falou W., Khalil M., Wendling F., Hassan M. Reduced integration and
improved segregation of functional brain networks in Alzheimer's disease. Journal of Neural
Engineering, 15(2):026023, 2018.
Kauffmann L., Ramanoël S., Peyrin C. The neural bases of spatial frequency processing during
scene perception. Frontiers in Integrative Neuroscience, 8:37, 2014.
Keenliside L., Modolo J., Thomas A.W., Legros A. Creating an ELF Magnetic Field Exposure
System for Human Whole Head Threshold Studies Over 50mT. Annual Joint Meeting of the
Bioelectromagnetics Society and the European Bioelectromagnetics association - BioEM 2015,
Pacific Grove, California, USA, June 14 - 19, 2015.
Legros A., Modolo J. Corbacio M., Goulet D., Plante M., Souques M., Deschamps F., Ostiguy G.,
Lambrozo J., Thomas A.W. Magnetophosphenes in humans exposed to ELF MF up to 50 mT, a
threshold study. BioEM2016 conference (meeting of the Bioelectromagnetics Society and the
European BioElectromagnetics Association), June 5-10, Ghent, Belgium.
Legros A., Modolo J., Brown S., Robertson J., Thomas A.W. Effects of a 60 Hz magnetic field
exposure of 1800 or 3000 microtesla on human brain activation as measured by functional magnetic
resonance imaging. PlosOne, 10(7):e0132024.
Lövsund P., Oberg P.A., Nilsson S.E., Reuter T. Magnetophosphenes: a quantitative analysis of
thresholds. Medical and Biological Engineering and Computing, 18(3):326-334, 1980.
Modolo J., Thomas A.W., Legros A. Human exposure of 50 and 60 Hz magnetic fields up to 7.6 mT:
an integrated EEG/fMRI study. Bioelectromagnetics, 38(6):425-435, 2017.
Figures

31
Figure 1. Left column. Brain regions within the left hemisphere with significantly different local
processing in the exposure condition, as quantified through the clustering coefficient. Middle
column. Similar, using an upper view of both hemispheres. Right column. Similar, for the right
hemisphere. Significant modulation of the right occipito-temporal pathway is consistent with
visual perception of a stimulus with low spatial frequency. (Up) Results in the EEG alpha band,
(Bottom) Results in the EEG beta band. Blue nodes denote a decrease in the clustering
coefficient, while red node denotes an increase. The red box higlights that most changes in
brain functional connectivity occurs within the right hemisphere.

32
 Session: S02
Dosimetry I - Computational studies
Monday June 25, 2018 • 11:00 - 12:30
Europa D
Chairs: Teruo Onishi & Akimasa Hirata

S02-1 [11:00]
Effect of vasodilatation on temperature elevation in human brain during localized RF
exposure: A computational study
Sachiko Kodera1, Jose Gomez-Tames1 & Akimasa Hirata1
1Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology, Nagoya, Japan,
466-8555
Keywords: Dosimetry (computational), RF/Microwaves, Completed (published)
Presented by: Jose Gomez-Tames
Temperature elevation is suggested to be a dominant effect for exposure at frequencies higher than
100 kHz. No previous studies have evaluated temperature elevation in the human head for local
exposure considering thermoregulation. This study aims to discuss the temperature elevation in a
human head model considering vasodilation, to discuss the conservativeness of the current limit for
protection against localized radio frequency field exposure.
Introduction
Specific absorption rate (SAR) averaged over 10 g of tissue is a metric used for human protection
for localized exposure to electromagnetic fields with frequencies ranging from 100 kHz to 3 GHz [1]
or 10 GHz [2]. The limit is 10 W/kg for occupational exposure or in restricted areas, and a reduction
factor of 5 is applied to general public general public or in unrestricted areas. Thermal effect is
suggested to be a dominant effect for exposure at frequencies higher than 100 kHz. In the IEEE
standard [1], the temperature elevation in rabbit eyes without anesthesia has been reported to be
smaller [3] than anesthetized rabbit eyes (1–2 ºC at a peak 10-g SAR). This is caused by the effect
of thermoregulation in non-anesthetized rabbits [4]. We investigated temperature elevations in rat
brains [5] due to localized radio-frequency exposure while considering thermoregulation. However,
no previous studies evaluated the temperature elevation in human head models for local exposure.
We investigate brain temperature elevation in a human head model, while accounting
thermoregulation (i.e., vasolidation), for excessive SAR exposure levels to discuss the
conservativeness of the current limit [6].
Method
This study computes the temperature elevations in an anatomical human head model exposed to
radiation from a dipole antenna at 1, 3, and 10 GHz. The specific absorption rates (SAR) in the
human model are first computed using a finite-difference time-domain method. The temperature
elevation is calculated by solving the bioheat transfer equation by considering the thermoregulation
that simulates the vasodilation [7], [8].
Conclusion
The computed temperature elevation under the current guideline for occupational exposure was <
0.93 ºC, which is within the ranges of brain temperature variability for environmental changes in
daily life. The temperature elevation became large and reached around 40 ºC for the peak SAR of
100 W/kg at 1 GHz, as shown in Figure 1. The effect of vasodilation is significant, especially at
higher frequencies where skin temperature elevation is dominant.
References
[1] IEEE, IEEE standard for safety levels with respect to human exposure to electromagnetic fields,
0-3kHz. Institute of Electrical and Electonics Engineers, 2002.

33
[2] ICNIRP, “Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic
fields (up to 300 GHz). International Commission on Non-Ionizing Radiation Protection.,” Health
Phys., vol. 74, no. 4, pp. 494–522, Apr. 1998.
[3] M. Kojima, I. Hata, K. Wake, S. Watanabe, Y. Yamanaka, Y. Kamimura, M. Taki, and K. Sasaki,
“Influence of anesthesia on ocular effects and temperature in rabbit eyes exposed to microwaves,”
Bioelectromagnetics, vol. 25, no. 3, pp. 228–233, Apr. 2004.
[4] A. Hirata, S. Watanabe, M. Kojima, I. Hata, K. Wake, M. Taki, K. Sasaki, O. Fujiwara, and T.
Shiozawa, “Computational verification of anesthesia effect on temperature variations in rabbit eyes
exposed to 2.45 GHz microwave energy,” Bioelectromagnetics, vol. 27, no. 8, pp. 602–612, Dec.
2006.
[5] A. Hirata, H. Masuda, Y. Kanai, R. Asai, O. Fujiwara, T. Arima, H. Kawai, S. Watanabe, I.
Lagroye, and B. Veyret, “Computational modeling of temperature elevation and thermoregulatory
response in the brains of anesthetized rats locally exposed at 1.5 GHz,” Phys. Med. Biol., vol. 56,
no. 23, pp. 7639–7657, Dec. 2011.
[6] S. Kodera, J. Gomez-Tames, and A. Hirata, “Temperature elevation in the human brain and skin
with thermoregulation during exposure to RF energy,” Biomed. Eng. Online, vol. 17, no. 1, p. 1, Dec.
2018.
[7] A. Hirata, T. Nomura, and I. Laakso, “Computational estimation of body temperature and
sweating in the aged during passive heat exposure,” Int. J. Therm. Sci., vol. 89, pp. 154–163, Mar.
2015.
[8] S. Kodera, J. Gomez-Tames, A. Hirata, H. Masuda, T. Arima, and S. Watanabe, “Multiphysics
and Thermal Response Models to Improve Accuracy of Local Temperature Estimation in Rat Cortex
under Microwave Exposure,” Int. J. Environ. Res. Public Health, vol. 14, no. 4, p. 358, Mar. 2017.
Figures

Figure 1. Comparison of temperature elevations with and without thermoregulation. DTW/O

and DTW refer to computed temperature elevation without and with thermoregulation.

S02-2 [11:15]
STUDENT PAPER
Relationship between transmitted power density and skin temperature elevation for different
beam directions from patch antenna arrays at the frequency bands above 10 GHz

34
Daisuke Funahashi1, Takahiro Ito1, Akimasa Hirata1, Takahiro Iyama2 & Teruo Onishi2
1Electrical and Mechanical Engineering, Nagoya, Japan, 466-8555
2Research Laboratories, NTT DOCOMO, INC, Kanagawa, Japan, 239-8536
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Daisuke Funahashi
Incident power density is used as a metric to protect human from excessive heating from
electromagnetic field exposures at the frequency bands assigned to the 5th generation wireless
system (5G). As presented in GLORE meeting (Washington DC, Dec 2017), one of the possibilities
is the transmitted power density averaged over an area as an alternative at higher frequencies to
correlate with peak temperature elevation. In this study, we investigated the beam direction and the
distance dependence of the relationship between transmitted power density and the skin
temperature elevation from patch antenna arrays.
Introduction
The frequency bands assigned to the 5th generation wireless system (5G) is 28 GHz or higher. A
patch antenna array may be implemented on mobile terminals in this system, and radio waves
emitted from the terminals may form a beam whose direction may depend on time. ICNIRP and
IEEE ICES provide safety standard or guidelines to protect human from excessive heating from
electromagnetic field exposures. Incident power density is used as a dosimetric quantity to prevent
excessive temperature elevation at the body surface from local exposure at the frequency above 3
or 10GHz. However, no sufficient rational for the limit and averaging area is given in the
guidelines/standard.
The metric for human protection at this frequency should have correlation with the surface
temperature elevation. In addition to the external field strength, the internal quantity in human body
should be useful. This kind of discussion is given in GLORE meeting.
In this study, we investigated the beam direction or distance dependence of the relationship
between transmitted power density and the skin temperature elevation from patch antenna arrays.
Materials and Methods
A cube homogeneous model composed of human skin was used in this study. As a wave source,
4-elements patch antenna array is considered (see Table 1. and Fig. 1. See also parameters of the
antenna array). The sinusoidal wave with same phase inputted to each feed point. In the
computational analysis, an in-house code of the finite-difference time-domain (FDTD) method is
used. Transmitted power density averaged over a square of 1cm2, 4cm2 at post processing. A
bioheat equation is solved to obtain the steady temperature elevation due to power absorption.
Dielectric properties and thermal constants in the model were the same as in our previous study [1].
Also, the Blood perfusion rate was chosen as used in [2]. One-dimensional analysis was also
considered to simulate plane wave exposure.
Results
Figures 2 shows the heating factor for transmitted power density (TPD) averaged over the square of
(a) 4cm2, (b) 1cm2 from 4-elements patch antenna array at the distance of 15, 30, 45 mm from the
model surface in case of θ = 0. The heating factor is then defined as the steady peak temperature
elevation divided by the transmitted power density averaged over an area. As shown in figure 2, the
heating factor of 4cm2 does not depends on the distance at 10 - 20GHz and that of 1cm2 is
independent of frequency or distance from the antenna. This mean the averaging area should be
small at frequency higher than 30 GHz such as the beam diameter is smaller than the side length of
the averaging area.
Figure 3 shows the heating factor for TPD averaged over the square of (a) 4cm2, (b) 1cm2 from
4-elements patch antenna array at the distance of 15 mm from the model surface in case of θ = 0,
15, 30, 45. It can been seen that the heating factor is independent of the beam direction. In case of

35
the 1cm2, the heating factor is frequency independent and safe compared as plane wave exposure.
Conclusions
This study discussed the heating factors in a cube homogeneous model composed of human skin
for exposure to a 4-elements patch antenna array. We confirmed the heating factor for TPD does not
depends on the frequency, the distance from the antenna and beam direction. It can be said that
TPD is a good metric to estimate human heating at the frequency bands above 10 GHz because
that has correlation with the temperature elevation at human skin.
References
[1] Y. Hashimoto et al., “On the averaging area for incident power density for human exposure limits
at frequencies over 6 GHz,” Phys. Med. Biol., vol. 62, no. 8, pp. 3124–3138, Apr. 2017.
[2] K. R. Foster, M. C. Ziskin, and Q. Balzano, “Thermal Modeling for the Next Generation of
Radiofrequency Exposure Limits,” Health Phys., vol. 113, no. 1, pp. 41–53, Jul. 2017.
Table 1. Parameters of patch antenna arrays at different frequencies.

Parameter

Frequency [GHz]

Parameter 10 20 30 40 60 80 100
of
patch (D) 0.75 0.15
antenna
[mm] (W) 75 50 30

(L) 9.0 4.75 3.0 2.4 1.4 1.1 0.7

Resolution
0.25 0.125 0.05
[mm]

Relative
(εr) 2.2
permittivity

Figures

36
Figure 1. Geometry of the 4-elements patch antenna array on the dielectric substrate.

37
Figure 2. Distance dependency of heating factor for transmitted power density. (a) 4cm2 (b)
1cm2

38
Figure 3. Beam direction dependency of heating factor for transmitted power density. (a) 4cm2
(b) 1cm2

39
S02-3 [11:30]
Human modeling by template based registration for evaluating the RF EMF exposure
Congsheng Li1, Xiaobang Sun2 & Tongning Wu1
1BEMS, CAICT, Beijing, China, 100191
2biomedical engineering, Dalian university of technology, Dalian, China, 100191
Keywords: Dosimetry (computational), RF/Microwaves, Completed (unpublished)
Presented by: Tongning Wu
In the paper, we proposed a method to reconstruct the individualized human model with high
efficiency, which were based on template based registration. The torso template was constructed
using statistical shape model with computerized tomography (CT) images of 79 healthy Chinese
subjects. The numerical studies demonstrated that the model generated by template based
registration had very few difference in terms of whole body averaged specific absorption rate and
the power absorption in the major tissue layers, compared with the manually segmented model.
I. Introduction
In the study, we used the Chinese male adult model [Wu et al., 2011]from precisely manual
segmentation (PS model), tissue simplification (TS model) and template based registration (TBR
model) for dosimetric simulations at the typical wireless communication frequencies. The resultant
whole body averaged specific rate (WBASAR) and stratified SAR using the three kinds of models
were compared. TBR model achieved very few dosimetric difference compared with the PS model.
Meanwhile, the TBR model can be generated with high efficiency and is suitable for evaluating the
individualized plane-wave exposure to EMF at wireless communication band.
II. Human anatomical models
A. Manually segmented model (PS model)
The Chinese adult male model had 87 different tissues. Segmentation work was performed on
2-Dimensional sliced photos by 30 anatomy experts. Each slice has been segmented separately by
at least two experts.
B. Tissue simplification model (TS model)
The profile of this model was the same as the manually segmented model. We used level-set
method to classify the tissues in the grey-scale images into four categories (skin, fat, muscle and
brain). Firstly, the segmentation was performed on each image to identify only the skin, fat, muscle
and the brain (white/grey matter). Tendons, ligaments cartilages, marrows and bones were
considered as muscles. Secondly, all other unidentified tissues were segmented as the muscle. The
profile of TS model was the same as that of the PS model. To present the difference, we display the
slice on the chest level in Fig. 1.
The segmentation and reconstruction cost about an hour with a personal computer.
C. Template based registration (TBR model)
The deformable torso template (Fig. 2a) was constructed based on torso computerized tomography
(CT) images of 79 healthy Chinese subjects. Major trunk organs were segmented from the training
images, and the statistical shape model (SSM) [Heimann et al., 2009] of torso structures was
constructed based on the surface meshes of all segmented organs. For personalized anatomy
modeling, the SSM was deformed to match the body surface mesh of the Chinese adult male model
(Fig. 2b). The deformed phantom was then voxelized to create organ label image for subsequent
electromagnetic simulation (Fig. 2c).The registration/reconstruction cost around 1 min.
III. Numerical simulations
The study focused on the typical wireless communication frequencies, i.e., 900 MHz, 1800 MHz,
2400 MHz and 3500 MHz.Finite difference time domain (FDTD) was applied to evaluate the

40
electromagnetic power absorption in the models subjected to EMF exposures. The models were
voxelised to 2×2×2mm3 for 900, 1800 and 2400 MHz and 1×1×1mm3 for 3500 MHz. Frequency
dependent dielectric properties [Gabriel et al., 1996] were applied. Time update step was defined as
3.7 picoseconds (2×2×2mm3) and 1.8 picoseconds (1×1×1mm3). WBSAR and the stratified SAR
(SAR for the major layers) was normalized to the plane wave incident power density as . Perfectly
matched layer was used as absorption boundary condition. Transverse Electromagnetic Wave
(TEM, electric field orientation parallel to the height of the anatomical model with direction of
propagation to the face of the anatomical models) was used to expose the anatomical models.
IV. Results and Discussions
WBASAR for the three kinds of models are shown in Fig. 3.The maximal difference in terms of
WBASAR between the TBS and PS model was 8% (900 MHz). The difference decreased to about
5% at 3.5 GHz. In comparison, the difference between the TS model and the PS model decreased
from 25% (900 MHz) to 7% (3500 MHz). It is reasonable to see the decreasing tendency because
the skin layer of the three models were the same. At higher frequencies, EMF power was mainly
absorbed by the superficial layers. Comprehensive numerical results using TS model with various
incidences and orientations can be found in the previous study [Wu et al., 2012]. The TBR model
performed better on WBASAR evaluation compared the TS model.
The cadaver of the Chinese male subject has the height close to the average population value but
the body mass index (BMI) was obviously low. It may contribute to the higher difference in fat and
muscle.
We did not list he results for the brain because the two models used the same brain structures and
torso registration did not change this part.
V. Conclusions
In order to evaluate the individual EMF exposure, we used two kinds of modeling methods to
reconstruct the human model. The first method was based on tissue simplification and the second
method was developed on template based registration. We constructed the registration template
from the CT images of 79 healthy male adult. TBR can generate the individualized model with
reasonable time cost and without exhaustive human intervention. The WBASAR was less than 8%
for the surveyed frequencies.
Acknowledgment
The work is supported by the grants from National Natural Science Foundation of China (Grant No.
61371187 and 61671158) and National Science and Technology Major Project (Grant No.
2018ZX10301201).
References
S. Gabriel, R.W. Lau and C.Gabriel, The dielectric properties of biological tissues: II. Measurements
in the frequency range 10 Hz to 20 GHz, Phys. Med. Biol, 1996, pp. 2251-2269.
T. Heimann and H.P. Meinzer, Statistical shape models for 3D medical image segmentation: A
review, Med. Image Anal, 2009, pp. 543-563.
T.N. Wu, L.W. Tan, Q. Shao, C. Zhang, C. Zhao, Y. Li, E. Conil, A. Hadjem, J. Wiart, B. Lu, L. Xiao,
N. Wang, Y. Xie and S.X. Zhang, Chinese adult anatomical models and the application in evaluation
of RF exposures, Phys. Med. Biol, 2011, pp. 2075-2090.
T.N. Wu, Q. Shao, L. Yang, Simplified segmented human models for whole body and localised SAR
evaluation of 20 MHz to 6 GHz electromagnetic field exposures, Radiat Prot Dosim, 2012, pp.
266-272.
Figures

41
Figure 1. Comparison of the slice on the chest level for PS model and TS model.

Figure 2. Procedure for TBR modeling.

42
 Figure 3. WBASAR of the three models (normalized to 10 W/m2 equivalent plan-wave power
density).

S02-4 [11:45]
SAR values in children and adults for real exposures
Blaž Valič1, Bor Kos2 & Peter Gajšek1
1Institute of Non-ionizing radiation, Ljubljana, Slovenia, 1000
2Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia, 1000
Keywords: Dosimetry (computational), RF/Microwaves, Completed (unpublished)
Presented by: Blaž Valič
Based on the personal measurements of radiofrequency electric fields in Slovenia we determined
the mean exposure of the children and adults. Results show that the most important frequency
bands are FM, 900 MHz, 1800 MHz and 2100 MHz. Based on these mean values SAR values were
calculated for child and adult human model. Whole body SAR values are very low for mean
exposures, but the exposure of the child model is significantly higher compared to the adult model.
Introduction
Due to differences in the use of modern wireless technologies and different everyday behaviour
there are some differences in typical daily exposures tor radiofrequency electromagnetic fields (RF
EMF) of children and adults. Further, the same exposure of the adult does not reflect in the same
SAR values as in the children due to differences in sizes, mass and body proportion between adults
and children. To analyse these differences we have conducted 54 personal RF EMF measurements
of children and 52 measurements of parents. Based on the mean values of the measurements we
also conducted numerical calculations of whole body SAR values for measured exposures in child
and adult models. Results of SAR values were later analysed and compared in-between.

43
Material and Method
Personal measurements were taken by RF EMF personal exposimeters ExpoM-RF (Felds at Work,
Switzerland), which measured RF EMF in 15 different frequency bands corresponding to different
technologies: FM radios, DVB-T, mobile networks at 800, 900, 1800, 2100 and 2600 MHZ (both
downlink and uplink), DECT, WiFi at 2.4 GHz and WiMax at 3.5 GHz. However for further analysis
all uplink results (800, 900, 1800, 2100 and 2600 MHZ) were excluded as only contributions from
the sources in the surrounding were taken into account. The measurement interval was 4 seconds
for all the measurements whereas the total duration of each measurement was from 1 to 3 days.
Based on the results of personal measurements the SAR calculations have been done with
SEMCAD X software package (SPEAG, Switzerland) for three most important RF exposures: FM
radio transmitter operating at 100 MHz and the mobile base station operating on 900 and 2170
MHz. The total whole body SAR value was calculated as the sum of contributions at each of the
calculated frequency. The results were compared to the ICNIRP guidelines from the 1998 [ICNIRP,
1998], which defines the whole body SAR value limit of 0.08 W/kg.
Two human models from the Virtual Family set [Christ et al., 2010] have been used: the Duke adult
model of a 34 y.o. male and a Thelonious child model of a 6 y.o. male. For each model three
directions of incident field were analysed: The direction of the incident field was from the front of the
body and from 45° and 90° to the side. The direction front is the anterior to posterior direction, the
side direction is left to right lateral direction and 45° is in-between front and side direction.
Results
In the study 54 children and 52 parents participated. Total measurement duration was more than
117 days for children and 118 days for parents. For each participant the mean value of the electric
field was calculated for all measured frequency bands.
In Table 1 the range (min and max) and the median value of the mean values of the measured
electric field Emean (in V/m) for each participant is given separately for children and for adults. It is
evident, that among all the frequency bands and technologies, the most important contributors for
both children and adults are FM transmitters and 900 MHz base stations, followed by the 1800 MHz
and 2100 MHz base stations. All other median values are much lower. There are some differences
between children and adult exposures. Median value is slightly higher for adults for FM (0.09 vs 0.06
V/m), 900 (0.11 vs 0.09 V/m), 1800 (0.06 vs 0.04 V/m) and 2100 (0.04 vs 0.03 V/m).
Table 1: Measurement results for RF EMF. For each frequency band the minimum, median and the
maximum value of the mean value Emean (in V/m) for each participant are given separately for
children and for adults.
Emean WiFi Wimax
FM DVB-T 800 900 1800 DECT 2100 2600
V/m 2.4 3.5

Min 0.01 0.00 0.00 0.01 0.00 0.00 0.00 0.00 0.00 0.00

children Median 0.06 0.01 0.00 0.09 0.04 0.01 0.03 0.02 0.00 0.00

Max 0.46 0.08 0.02 0.41 0.30 0.10 0.24 0.06 0.01 0.00

Min 0.02 0.00 0.00 0.03 0.01 0.00 0.01 0.01 0.00 0.00

adults Median 0.09 0.02 0.00 0.11 0.06 0.01 0.04 0.02 0.00 0.00

Max 0.41 0.07 0.03 0.42 0.30 0.11 0.18 0.10 0.03 0.04
SAR calculations have been done for simultaneous exposure to mean values of the measured RF
EMF for FM transmitters at 100 MHz, for 900 MHz base stations and for the exposure at 2170 MHz,

44
where the sum of the mean values at 1800 MHz and 2100 MHz was used as the exposure scenario.
These three frequency bands are the most important contributors to the total mean exposure, as is
seen form the Figure 1.
Table 2: Calculated total whole body SAR values for combined exposure at different orientations of
incident field. The results are given for Thelonious child model and Duke adult model for minimum,
median and the maximum value of the mean value Emean.

SARmedian
Orientation Model SARmin [nW/kg] SARmax [nW/kg]
[nW/kg]

child 12 529 23465

1 front
adult 23 334 6227

child 11 467 21138

2 45°
adult 18 273 5024

child 10 412 18769

3 side
adult 17 245 4585
From the results in Table 2 and Figure 2 we can see that the whole body SAR values are extremely
low, with the highest value of 23465 nW/kg, which corresponds to 0.000023465 W/kg. There is great
variability of the SAR value between the minimum and maximum value for each exposure scenario.
The variability is up to 3 orders of magnitude. If we compare the results of child model and adult
model, we can see that the whole body SAR value in child model is significantly higher for all three
orientations of the incident RF EMF. Nevertheless that results of personal measurements showed
the children are less exposed to RF EMF, the SAR values are still significantly higher for the child
model compared to the adult model.
The orientation of the incident RF EMF influences the results. The highest exposures are obtained
for frontal exposure and the lowest are for the side exposure. This is understandable, as the
exposed area is the highest for frontal exposure and the lowest for side exposure.
Conclusions
The results of the study clearly show that exposure of the children is higher compared to the adult.
The whole body average SAR in the child model is higher than for the adult for all important
frequency ranges. The variability between the minimum and maximum SAR value due to variability
of measured results is much higher than variability of the SAR due to orientation of the incident RF
EMF.
Acknowledgement
The research leading to these results has received funding from the European Community’s
Seventh Framework Programme (FP7/2007- 2013) under the grant agreement no. 603794—the
GERONIMO project.
References
[1] ICNIRP. International Commission on Non-Ionizing Radiation Protection guidelines for limiting
exposure to time-varying electric, magnetic and electromagnetic fields (up to 300 GHz). Health
Phys. 74 494–522, 1998.
[2] Valič B, Kos B, Gajšek P. Typical exposure of children to EMF: exposimetry and dosimetry.
Radiat Prot Dosimetry. 163:70-80, 2015. doi: 10.1093/rpd/ncu057
Figures

45
Figure 1. Measurement results for RF EMF. For each frequency band the minimum, median
and the maximum value of the mean value Emean (in V/m) for each participant are given
separately for children and for adults – parents.

Figure 2. The whole body SAR values for different exposure directions for child and adult
model have been calculated for minimum, median and maximum simultaneous exposure to
mean values of the FM transmitters at 100 MHz, 900 MHz base stations and for the exposure
at 2170 MHz, where the sum of the mean values at 1800 MHz and 2100 MHz was used as the
exposure scenario.

46
S02-5 [12:00]
Assessment of RF-EMF exposure in LTE heterogeneous networks for VoLTE and data
services
Yuanyuan Huang1 & Joe Wiart1
1LTCI, Telecom ParisTech, Paris, France, 75013
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Yuanyuan Huang
LTE is deployed worldwide with macro- and small cells to fulfill different requirements motivated by
numerous mobile services: VoLTE and other data applications. However, the actual human
exposure to electromagnetic field from such networks from base station to user equipment is still
unknown, hence, needs to be communicated. To do so, people’s live traffic usage has to be
characterized. This study aims to investigate the realistic LTE-induced exposure, considering both
the uplink and downlink radio emissions and various factors, such as traffic variation over a day,
cells of different sizes referred to macro- and small cells, different services such as VoLTE and data
and different environments respectively, in indoor and outdoor areas.
Introduction
The evolution of cellular networks meets explosive growth of data traffic as well as urgent demand
of high data rates. In 2016, global mobile data traffic has grown 63% [1]. To anticipate such traffic
usage, Long Term Evolution (LTE) networks have been launched worldwide to achieve the high
service quality and system capacity. So far, LTE traffic has already accounted for 69% of mobile
traffic [1]. In addition, the deployment of LTE base stations (a.k.a., eNB) with lower power, e.g., small
cells (SC) that have a range of 10 meters to a few hundred meters, is envisioned to increase the
traffic capacity by providing indoor coverage or at cell edge of large cell. In parallel, along with those
eNB installations, public attention around risk perception of exposure to radiofrequency (RF)
electromagnetic field (EMF) has grown just as much as their usage of wireless technologies.
Relative research dedicated to the realistic LTE-induced EMF exposure is important but scant.
Several studies on in-situ measurement in LTE operating networks have been carried out to date.
On one hand, authors of the study [2] explore the actual radiated power by one user equipment (UE)
connecting to LTE networks in terms of different voice and data applications. Further, based on a
large amount LTE UE samples, the realistic level of transmit power (UL) considering data service in
different zones was presented in [3]. On the other hand, studies such as [4] investigate the RF-EMF
exposure to LTE coming from eNB (DL). But none of these studies has yet combined both UL and
DL radio emissions in LTE network as a whole to assess the global exposure. As shown in previous
studies [5-6], the contribution of UL and DL exposure, respectively from UE and eNB, to global
exposure is differ from one radio access technology to another. Briefly, the contribution of UL
exposure to global exposure to 3G networks has been highlighted, resulting from voice and data
traffic. Meanwhile, in LTE, the global exposure is dominated by DL exposure. However, the study
had limited scope on focusing only the data usage used in front of the chest. Therefore, results
should be revised by considering a mobile phone close to the head usage, i.e., voice over LTE
(VoLTE).
Investigating differences between the EMF exposure to LTE macro- and small cells, experimental
measurements have been done separately to assess received and transmitted powers when
connecting (or non-connecting) macro (or small) cells [7]. However, these measurements have been
done in outdoor environments. As reported in [8], in France, almost 70% of daily time is spent
indoors. Due to the RF attenuation caused by walls, the power received and transmitted by UE
located in indoor area will differ from the one in outdoor area. For this reason, the indoor exposure is
also significant to analyze. Furthermore, ULand DL EM radiations and network performance depend
on the traffic load. Thus, usage traffic is fundamental in the evaluation of EMF exposure to a
wireless network. The authors are aware that the exposure values should be highly linked to the
evolving data traffic. It is therefore important to collect latest usage statistics from live LTE mobile
networks.
47
Hence, this study amis to assess the realistic LTE-induced EMF exposure, considering both the UL
and DL radio emissions and different factors, i.e., traffic variation observed in operating networks,
cells of different sizes referred to as macro- and small cells, different services of data and VoLTE,
different environments (indoors and outdoor areas). Toward this, firstly, usage data traffic collected
from traffic measurements is presented and the network traffic simulator used for the study. Then,
the global exposure results obtained by considering different factors are introduced. Finally, the
conclusion is presented.
Methods
Live traffic measurements
Usage traffic is a key factor when dealing with realistic exposure assessment issues. During a five-
month period, 17 eNB were monitored 24 h a day in an urban area in France. All subscribers
connected to these eNB have been recorded throughout the time by counters installed in LTE
networks.
User’s hourly mobile usage, such as UL and DL traffic volumes are identified statistically. Figures
1a,b give results of the average UL and DL volumes of data exchanged per user per hour. From
hourly traffic distribution, it was observed that people start to use their mobiles around 5 a.m. The
majority of data traffic was occurred from midday to midnight. It was also found that DL versus UL
average ratio has reached up to 8.2.
Furthermore, voice usage was also explored. Figure 1c corresponds to the average voice calls
duration per user per hour in indoor and outdoor environments in terms of a day. As illustrated in
Figure 1c, a strong heterogeneity in voice usage appears in time and space, in other words, usage
habits change in accordance with people’s daily schedule. It was found that average voice calls
durations generated in indoor area is 2.5 times longer than what was generated in outdoor area
during “working time” from 7 a.m. to 5 p.m. Meantime, voice durations occurs indoors is 0.4 shorter
than what was observed outdoors during “breakfast time” from 5 a.m. to 7 a.m. All these hourly
mobile usage variations will be implemented further in a LTE traffic simulator to evaluate the
corresponding global exposure distribution over 24 h.
Network traffic planning tool
In this paper, a LTE network traffic tool was used to simulate the transmitted power by UE and the
received one from eNB. When considering DL transmissions, a load factor L(ΔT) related to the
considered cellular coverage was implemented to assess the emitted power over a time interval
ΔTby eNB. In the UL direction, the open-loop power control algorithm in 3GPP was applied. Data
are transferred by UE through physical uplink shared channel (PUSCH), whose transmission power
can be thus calculated as follows:

where is the maximum transmitted power by UE. ( , ) are cell specific parameters
respectively linked to target received power at eNB and path loss compensation factor. is
the number of Resource Blocks (RB) scheduled to a certain UE in a time slot. PL is the path loss
between the UE and his connecting eNB.
In our analysis, VoLTE and data services were considered. Voice talks can only be transmitted
through voice over Internet Protocol. The voice packet can be transmitted every 20 ms during the
period of active state and every 160 ms during the idle state period. For LTE traffic, the minimum
allocation unit is one Physical Resource Block (PRB), which consists of 12 sub-carriers in one sub-
frame of 1 ms duration. Hence the number of PRB needed for the voice packet of the i-th VoLTE
user can be formulated as:

where L represents the payload voice packet size, which is determined by the codec mode. The

48
AMR-WB 12.65 coder was used in this research. corresponds to 12 sub-carriers and
equals to 14. is the number of Resource Element allocated to the control information. is the
bandwidth efficiency of modulation and coding set for user i, which is based on the signal-to-
interference and noise ratio. It should be noted that a higher priority scheduling is given to transform
VoLTE traffic due to a significant requirement of delay.
Results
57 macro cells is deployed in hexagon with an inter site distance of 500 m. As shown in Figure 2, 12
SC are deployed at cell edge of central macro cells. Statistics were gathered through these sectors.
VoLTE and data usres were uniformly distributed. Meanwhile 67% of UE were selected randomly to
be located in indoor environment. One should note that UE located indoors suffers a 10 dB loss due
to wall signal penetration for both UL and DL directions.
The duration of each simulation was one hour. Average number of simulated users per cell in terms
of different hours is demonstrated in Figure 3. For each UE, mobile traffic was defined in terms of
different services and various periods (Figure 1). An assumption was made that all VoLTE users talk
all the time during voice communications. Hence, idle state is ignored. Other simulation seetings can
be found in Table 1.
Briefly, 24 simulations were executed respectively, when SC is on and off, by varying UE traffic over
a day. Results such as the average UL transmitted power by VoLTE or data users, the received one
from eNB or SC and the UL throughput were obtained. These values will be further used to evaluate
the UE’s global exposure variation to RF-EMF over 24 h.
Table 1 Simulation settings

Band width 10 MHz (50 RB) UL and DL

EIRP 46 dBm (eNB), 30 dBm (SC)

System eNB: 2D pattern, 14 dBi gain;

Antenna
SC: omni-directional with 5 dBi

(-80, 0.9) (eNB), (-59, 0.6)

( , )
dBm (SC)

Max transmit power 23 dBm

Omni-directional with 0 dBi

Antenna
UE gain

Depends on periods and

Active user density/traffic
environments

Macro 128.1+37.6log(d)
PL
SC 140.7+36.7log(d)

Conclusion
Traffic usage statistics were analyzed based on the data collected from live LTE mobile networks.
The differences in usage among mobile service subscribers affected by different time periods and
different environments were investigated. These traffic analyses were used in a LTE network traffic
tool to obtain the average UL transmitted power by VoLTE or data users, the received one from eNB
or SC and the UL throughput. All these data will be further used as inputs to evaluate the UE’s
global exposure to RF-EMF variation map. Detailed statistical comparisons of the RF-EMF exposure
values will be performed for the UE using voice or data services, where located indoors or outdoors,
49
when SC is on or off.
Acknowledgments
This paper reports work undertaken in the context of the ANSES project AMPERE.
References
[1] Cisco Visual Networking Index: Global MobilE Data Traffic Forecast Update, 2016-2021 White
paper.
[2] Paljanos A et al. 2016. Assessment of mobile phone user exposure to UMTS and LTE signals:
comparative near-field radiated power levels for various data and voice application services. Journal
of Electromagnetic Waves and Applications 30.
[3] Joshi P et al. Output power levels of 4G user equipment and implications on realistic RF EMF
exposure assessments. 2017. in IEEE Access 5:4545-4550.
[4] Verloock L et al.Low-cost rxtrapolation method for maximal LTE radio base station exposure
estimation: test and validation. 2013. Radiat. Prot. Dosimetry 155:11-15.
[5]Huang Y et al.. 2016.Comparison of average global exposure of populationinduced by a macro
3G network in different geographical areas in Franceand Serbia. 2016.Bioelectromagnetics
37:382–390.
[6] Huang Y, Wiart J. 2017. Simplified assessment method for population RF exposure induced by a
4G network. 2017. IEEE J-ERM 1:34-40.
[7] Mazloum T et al.. Assessment of RF human exposure to LTE small- and macro-cells: UL
case.2017. EUCAP.
[8] Zeghnoun A, Dor F. 2010. Description du budget espacetemps et estimation de l’exposition de la
populationfrançaise dans son logement,“ Institut de veille sanitaire.

Figures

50
Figure 1. Mean volume of data exchanged for uplink (a) and downlink (b) per user per hour in
LTE networks in an urban area of France. Mean voice calls duration (c) per user per hour in
indoor and outdoor environments

51
 Figure 2. Macro- and small cells deployment

Figure 3. Average number of connected users per cell over 24 h

S02-6 [12:15]
STUDENT PAPER
Small cell EMF exposure assessment using propagation model based on measurement and
statistical method
Amirreza Chobineh1, Yuanyuan Huang1, Taghrid Mazloum1, Emmanuelle Conil2 & Joe Wiart1
1Chaire C2M, Télécom Paristech, Paris, France, 75013
2Agence nationale des fréquences (ANFR), Maisons-Alfort, France, 94704

52
Keywords: Dosimetry (computational), RF/Microwaves, Completed (unpublished)
Presented by: Amirreza Chobineh
Small cells are low-powered base stations installed close to the users. They improve network
performance and service quality by offloading from the macrocells. In this paper, the global
exposure induced in such networks from the user equipment and base stations has been
investigated. As the Small cell is close to the user, the propagation channel becomes highly variable
and vulnerable to environmental changes such as the road traffic. An innovative statistical path loss
model is constructed based on measurements. This statistical path loss model is then used to
assess global exposure of the adult proportion of a population in a Small cell scenario composed of
indoor and outdoor data users.
1. Introduction
As a result of technological advances, wireless communication systems are used by the majority of
people in every aspects of daily life. This led to a strong increase in data traffic during the last
decade [1].To respond to such growth, cellular mobile network operators try to enhance their
network performances. One of the promising techniques is densifying the traditional macrocell
networks by small cells (SC). Small cells are low-powered base stations (BS). The coverage area of
a SC is ranged from a few meters to a few hundreds of meters. Small cells are often installed at a
low height.
Meanwhile, as the role of wireless communication technologies become progressively important, the
public concern about human exposure to radio frequency (RF) electromagnetic fields (EMFs) still
persists, despite existing international safety limits. Consequently, in addition to the decreased
distance between users and SC antennas, since SCs are expected to be deployed massively such
deployment can result in a significant increase in public concerns.
In this respect in the framework of a European project called Low Electromagnetic Field Exposure
Networks (LEXNET), a new metric called Exposure Index (EI) has been proposed in order to deal
with global exposure in which the downlink and uplink exposures are taken into account
simultaneously [2].
Previous studies [2], [3] have computed EI for different scenarios involving urban LTE and UMTS
networks. In these studies, deterministic simulation tools were used to evaluate the path loss
between BS and the UEs, however, such deterministic tools are dependent on the level of details of
environmental data.In addition, the variability of the environment has not taken into
account.Recently, a new statistical approach has been introduced [4] to handle the structural
variability of the propagation environment by investigating the path loss exponent (PLE) observed in
a typical European city. The distribution of PLE was given for a whole city model, constructed by a
stochastic geometry tool.
Since SCs are installed near the users at lower heights, the propagation channels between UE and
BS becomes more susceptible to environmental changes. For instance, a bus can completely block
the SC’s antenna.Hence, it is important to investigate the variability of PLE for each user, located at
a specific distance from the SC instead of a unique PL distribution for an entire city as presented in
the study [4].
Additionally, a recent study [5] describes that in France, people spend about 70% of their time in
indoor areas such as homes and offices. Thus, indoor users form an important part of the exposure
population and they have to be taken into account in exposure assessment.
This paper presents the assessment of the global exposure induced in a LTE SC network based on
a series of measurements performed using hand-held drive test mobile systems. The variability of
the environment has been modeled through statistical PLE models for each individual LTE SC user
located at different places. In the last section, a simplistic scenario has been designed containing
indoor and outdoor users. The EI has been computed by constructed PLE statistical models.
2. Measurement campaign

53
In June 2017 a series of measurement has been performed in a typical urban French city on two
SCs. SCs are located on top of a bus station and an advertisement panel at a height of 3m. The
street width is about 8 m and average building height is 10 m.Drive test solution technologies have
been used during the measurements to monitor the networks characteristics such as emitted and
received powers and power control algorithm parameters.
A series of path loss measurements has been performed by traveling in the line of sight (LoS)
direction of the SC’s directional antenna. The uplink power emission measurements are carried out
while uploading 100 Mbytes data files to an FTP server.
3. Statistical characterization of path loss relative to small cell emission in the urban
environment
The received power by UE can be evaluated by the following equation:

Where PL denotes the channel path loss, is the transmitted power by the SC, is
the UE antenna gain, the SC antenna gain when UE is placed at the (θ,φ) direction.
is the received power by UE. It is assumed that the UE antenna is omnidirectional (
= 0).
In order to model the path loss in the channel between the SC and UE, a commonly used model has
been applied in which the path loss is presented as a function of distance raised to the power of γ.
This model can be presented as [6]:

Where γ is the path loss exponent (PLE), d is the distance between user and SC antenna, A is the
path loss constant considering that the gain of the receiver is 0 dB. It can be evaluated through the
following equation:

λ represents wavelength of the emitted wave. The following equation is used in order to assess the
path loss exponent at each measurement point [4]:

The value of γ has been computed based on about 10000 measurement points on two sites at 1800
MHz.
The results illustrate (figure 1) that the variation of PLE is different for different distances. This
means that it is possible to classify two different PLE models in terms of distance.
According to results, for the distances lower than 60m the variation of PLE follows a generalized
extreme value (GEV, k = -0.31, s = 0.42, m = 2.7 ) distribution and for the distances higher than 60m
the PLE variations follow a Beta(α1 = 3, α2 = 3.4, a = 2.2, b = 3.2).These distributions have been
used to assess the path loss distribution of user and consequently, the received and emitted
powers.
4. LTE uplink power control algorithm
In LTE technology, the emitted power by UE is determined by the uplink power control algorithm. In
this study, we focus on Physical uplink shared channel (PUSCH) which contains the uplink
transmitted data. The PUSCH power algorithm can be expressed by the following equation [7]:

This algorithm is a combination of open and closed loop components.Pmax is the maximum Tx
power of the UE 23dBm (±2dBm),Po is a cell-specific target power, M is the total number of

54
allocated uplink resources block, α is the cell specific pathloss compensation factor, PL is the
downlink pathloss measured at the UE, is a closed loop parameter specified at the UE by the
upper-layers and is a UE-specific closed-loop corrector [7]. In this study we ignore the closed loop
component due to its minor effect on emitted uplink power.Measurements show that the path loss is
fully compensated ( α = 1 ) by the uplink power control algorithm and Po is fixed to -96 dBm.
5. LTE small cell scenario simulation
LTE SC users (in indoor and outdoor areas) exposure, has been evaluated in a simplistic scenario
of a street of 8m width and 400 m length. This street is surrounded by two series of buildings on
each side, with a penetration depth of 6 m as presented in figure 2. Each building has 4 floors. The
UE is situated at the height of 1,5 from the ground. The penetration loss has been taken into
account as a random variable uniformly distributed on the interval [7-13] dB.
Users are distributed uniformly in the indoor and outdoor areas. We suppose that 70% of
occurrences take place indoors. The SC is located in the outdoor area, at a height of 3 m from the
ground at the center of the map.
The distance between the user and the SC is computed for every observation. In order to evaluate
the path loss of each observation, a random sampling of the proper distribution of path loss
exponent (GEV or Beta according to distance), has been realized.With the path loss distribution in
hand, the distribution of received and emitted powers can be evaluated for every observations as
detailed in section 3.
The EI has been assessed for the adult proportion of a population, located in the indoor and outdoor
area, in a dense urban zone, using an LTE SC in standing position during 1h. It is considered that
an average user upload 100 Mbytes of data each day. Hence a moderate user uploads 4.16 MByte
per hour.
The evaluation of average emitted and received powers and uplink exposure time let us to assess
the EI by use of normalized whole-body SAR values. The normalized whole-body SAR value
induced by a mobile phone for an adult using data service at 1800 MHz in standing position is
extracted from the values computed in the framework of LEXNET project. Results show that the
uplink component of exposure is as expected dominating EI and the exposure of indoor users is
higher than outdoor users (figure 3).
Parameters Indoor Outdoor

(W) 0.0114 0.0012

(W/m2)

EI (W/kg)

Global EI (W/kg)

6. Conclusion
In this study, the great variability of local propagation channel in case of SC is taken into account by
using an innovative statistical approach. Innovative statistical path loss models have been built
using the measurements performed with SCs installed in a French urban city. EI has been computed
by using these models in an indoor and outdoor scenario. Results show that the exposure of indoor
SC users is higher than outdoor users. In adition, the computed EI value for SC users is lower than
the EI computd for LTE macrocell users [4].In this paper, the designed scenario is quite simplistic.
More complex configurations will be investigated in future works.

55
References
[1] “Cisco Visual Networking Index: Global Mobile Data Traffic Forecast Update, 2016–2021 White
Paper.”
[2] N. Varsier et al., “A novel method to assess human population exposure induced by a wireless
cellular network,” Bioelectromagnetics, vol. 36, no. 6, pp. 451–463, Sep. 2015.
[3] Y. Huang et al., “Comparison of average global exposure of population induced by a macro 3G
network in different geographical areas in France and Serbia,” Bioelectromagnetics, vol. 37, no. 6,
pp. 382–390, Sep. 2016.
[4] Y. Huang and J. Wiart, “Simplified Assessment Method for Population RF Exposure Induced by a
4G Network,” IEEE J. Electromagn. RF Microw. Med. Biol., vol. 1, no. 1, pp. 34–40, Jun. 2017.
[5] A. Zeghnoun and F. Dor, “Description du budget espace temps et estimation de l’exposition de la
population française dans son logement.” Sep-2009.
[6] J. D. Parsons and J. G. Gardiner, Mobile Communication Systems. Boston, MA: Springer US,
1990.
[7] J. Gora, K. I. Pedersen, A. Szufarska, and S. Strzyz, “Cell-Specific Uplink Power Control for
Heterogeneous Networks in LTE,” in 2010 IEEE 72nd Vehicular Technology Conference - Fall, 2010,
pp. 1–5.

Figures

Figure 1. The variation of PLE in terms of distance. PLE varaitions become stable after about
60 m.

56
Figure 2. Scenario structure (3D). Outdoor users : red circles, Indoor users : blue stars. SC is
located at the center of the map (orange dot).

Figure 3. IE value for Indoor and outdoor users.The results have shown that, in this
configuration, the public global exposure of indoor LTE SC users is about 6 times higher than
the outdoor users. This is due to higher emitted power by UEs in indoor compared to
outdoor.These results, highlight the importance of the uplink exposure in global exposure and
EI assessment.

57
 Session: S03
In vitro I
Monday June 25, 2018 • 16:30 - 18:00
Europa C
Chairs: Olga Zeni & Mats-Olof Mattsson

S03-1 [16:30]
Real-time detection of neuronal activity induced by time-varying magnetic field or electric
field exposure in in vitro
Masateru Ikehata2, Keiji Wada3, Kei Makino3, Yukihisa Suzuki3, Atsushi Saito1, Sachiko Yoshie2 & Satoshi
Nakasono1
1Central Research Institute of Electric Power Industry, Abiko, Japan
2Biotechnology laboratory, Railway Technical Research Institute, Kokubunji, Japan
3Tokyo Metropolitan University, Hachioji, Japan
Keywords: In vitro, All Frequencies, Work in Progress
Presented by: Masateru Ikehata
Rat primary cortical neuron culture and human iPS cell-derived neurons were used to investigate
threshold of magnetic/electric stimulation of nerve cells in in vitro.Originally developed live cell
imaging system for real-time detection of neuronal activity without disturbing by interference of
magnetic field was used. In magnetic field exposure experiment, it seemed that response of neuron
cell that synchronized with magnetic field in some cases but observed neuronal activity was not
necessarily synchronized with magnetic field exposure. Improvement of cell direction, orientation of
neuronal network and other factors may need to observe neuronal cell response certainly. Further
results will be presented in the Congress.
Introduction
International guidelines that aims to protect people from adverse health effect by exposure to
magnetic fields (MF) has established by international organizations such as the International
Commission on Non-Ionizing Radiation Protection (ICNIRP) or the Institute of Electrical and
Electronic Engineers (IEEE). In the guidelines, the target effects below 100 kHz is nerve stimulation
based on a few provocation studies. Because of a few studies, the value of basic restriction of the
guidelines adopt extrapolation widely based on limited scientific evidences with several frequencies.
Therefore, determination of threshold value among extrapolated frequencies is important for
verification and improving of the current basic restriction. Here, we report progress of our study to
determine of threshold value of nerve stimulation by magnetic or electric field in in vitro.
Materials and Methods
Rat primary cortical neuron culture and human iPS cell-derived neurons (ReproNeuro, Reprocell,
Japan) were used in this study. For the rat primary culture, the cell suspension was prepared by
standard method using trypsin 1). Cells resuspended in the Neurobasal medium with B-27,
Glutamax (Thermo Fisher Scientific, USA) and PS and then poured onto PEI coated cover glass
(φ13mm, Matsunami Glass Ind. Ltd., Japan). After 3 weeks, the culture was mature and ready to
use for experiment. For human neuron culture, differentiation and maturation was performed
according to manufacturer’s guidance.
For single cell fluorescence observation under strong magnetic field, we developed original device
insists X175 zooming lens with 5cm working distance with UV excitation/filter unit (Keyence Co.,
Japan), a Cooled CCD camera (ORCA 3.8, Hamamatsu Photonics K.K., Japan), Camera control
and analysis software (Metamorph, Molecular Device Co., USA), non-magnetic cell chamber (Blast
Inc., Japan) and non-magnetic lens and camera holder (original). Experiment of electric field

58
stimulation was also performed using this system. Fluo-4 AM was used as Calcium ion indicator and
loading procedure was based upon manufacture’s protocol (Dojin-do, Japan).
For an intermittent repetitive high density magnetic field exposure such as 10 cycle of 200Hz, up to
650mT was performed by originally developed magnetic field generator (K. Wada et al., 18th
European Conference on Power Electronics and Applications).
Numerical dosimetry was performed for the magnetic field and the electric field exposure
experiment, respectively. Induced electric field in the culture medium was calculated by the
combination of the finite element method, which is provided as the commercial software (COMSOL
Multiphysics® ver. 5.1), and the impedance method, which is originally developed by Tokyo
Metropolitan University, in the case of the exposure to magnetic filed. In the case of the exposure to
the electric field, the finite integration technique, which is also provided as the commercial software
(CST EM STUDIO®), was used to calculate induced electric field between electrodes immersed in
culturemedium.
Results and Discussion
As a preliminary experiment, increase of fluorescence depend on applied electric voltage was
observed by stimulation of 200Hz electric field that was applied by originally developed electrode as
a positive control in both rat and human neuron cell culture. In magnetic field exposure, it seemed
that response of neuron cell that synchronized with exposure of 10 cycle of 200Hz, 650mT magnetic
field in some cases but observed neuronal activity was not necessarily synchronized with magnetic
field exposure. Estimated induced electric voltage by numerical analysis was exceeded threshold
value assumed by the ICNIRP however improvement of cell direction, orientation of neuronal
network and other factors may need to observe neuronal cell response certainly. Further results will
be presented in the Congress.

S03-2 [16:45]
50 Hz 1 mT field exposure does not affect DNA global methylation of an in vitro model for
Parkinson’s Disease
Barbara Benassi1, Stefania Santangeli2, Caterina Merla1, Letizia Tarantini3, Valentina Bollati3, Carmela
Marino1 & Claudia Consales1
1ENEA-Italian National Agency for New Technologies, Energy and Sustainable E, Rome, Italy, 00123
2Department of Life and Environmental Sciences, “Marche Polytechnic” University, Ancona, Italy, 60131
3Department of Clinical Sciences and Community Health, University of Milan, Milan, Italy, 20122
Keywords: In vitro, ELF/LF, Completed (unpublished)
Presented by: Claudia Consales
Epigenetic modifications are dynamic processes reflecting a complex interplay between an
organism and its environment. Over the last years, ELF-MFs exposure have been shown to affect
cell epigenome [1-3]. On this topic, we have shown that 50 Hz 1 mT field induces a decrease of
miR-34 b/c expression in neurons, with consequent increase of one of its gene target, α-synuclein,
whose overexpression is one of the hallmark of Parkinson’s disease [4]. In this work, we analyzed if
50 Hz 1mT field could alter global DNA methylation of neuronal like cells, dopaminergic
neuroblastoma cell line SH-SY5Y, exposed up to 72 hours, or exposed for 24 hours and then
administrated with Parkinson neurotoxin 1-methyl-4-phenylpyridinium (MPP+).
Background
The term epigenetics refers to heritable changes in gene expression that do not involve modification
in the underlying DNA sequence. These modifications are tissue-specific and reversible, but
mitotically stable, so they can also be transmitted through the different generations when found in
germ cells. The major responsible for the epigenetic changes is the environment, which may affect
DNA methylation, histone modifications, or non‐coding RNA signaling, and thus cause persistent

59
perturbations of regulatory pathways and eventually induce an altered susceptibility to disease [5].
DNA methylation consists in the modification of the cytosine in fifth position to 5-methyl cytosine
through the action of DNA methyl transferase. This is a major epigenetic modification in multicellular
organisms and in mammals, that occurs almost exclusively at CpG dinucleotides mainly associated
to gene promoter regions. Cytosine methylation triggers transcriptional repression either by
interfering with transcription factor binding or by inducing a repressive chromatin structure [6].
Epigenetic regulation is fundamental in enabling brain development and plasticity, therefore
alteration of epigenetic assessment is critical for the onset of neurodegenerative disorders, PD
included [7]. In particular, alteration in DNA methylation has been reported in PD patients, which
correlates to modification of gene expression [8].
The last few years have seen increasing public concern about the health effects of both residential
and occupational exposure to ELF-MFs, such as those generated by power lines (50 or 60 Hz), and
typically affecting specific categories of workers. Epidemiological data have also highlighted the
potential impact of ELF-MFs exposure in brain and PD onset [9], although evidence is still poor and
controversial. In this context, we have demonstrated that 50 Hz 1mT magnetic field exposure can
change redox homeostasis of dopaminergic neuroblastoma cell line SH-SY5Y and sensitizes cells
to a PD neurotoxin MPP+, increasing its capacity of inducing oxidative stress and cell death [10].
Considering how relevant are epigenetic changes in the onset and progression of PD, another
possible way by which ELF-MFs exposure could influence this disease is through the modulation of
neurons epigenome. ELF-MFs exposure, indeed, has been demonstrated to affect epigenetics [1-3]
and our group have demonstrated that exposing dopaminergic SH-SY5Y cells and neuronal primary
cultures to 50 Hz 1 mT field decreases miR-34 b/c expression increasing its target gene,
α-synuclein, one of the hallmark of Parkinson’s disease. We also demonstrated that miR-34 b/c
decrease is due to a hyper-methylation of microRNA promoter [4].
On the basis of these results, in this work we aimed at verifying, in the SH-SY5Y cell line
differentiated or not towards a dopaminergic phenotype, if 50 Hz 1 mT magnetic field exposure,
could also influence global DNA methylation. At this purpose, we analyzed global methylation of
DNA from cells chronically exposed up to 72 hours to 50 Hz 1 mT or to SHAM field. Moreover, we
also evaluated if ELF-MFs exposure could affect global DNA methylation when they are applied in
combination to the PD neurotoxin MPP+.
Methods
Exposure system
The exposure system consists of a couple of two coaxial coils of square shape (side 32 cm),
previously characterized and detailed [4, 10].
Cell Culture and Differentiation
Human SH-SY5Y neuroblastoma cells were cultured in complete DMEM/F12 medium (50:50 mix,
Euroclone), supplemented with 10 % heating activated fetal bovine serum, 2 mM L-glutamine, 100
μg/ml penicillin-streptomycin, and kept in culture up to 15 passages. Cell viability was assessed by
Trypan blue dye exclusion.
For experimental points with differentiated SH-SY5Y, cells at 60–70% confluence were sub-cultured
and induced towards a dopaminergic phenotype with 5 μM retinoic acid (RA) for the first 3 days,
followed by 80 nM phorbol myristate acetate (PMA) for further 3 days in reduced serum condition (5
%) [10]. The dose of 5 μM RA was selected in the range between 1 and 10 μM as it allowed a
detectable expression of dopamine transporter and tyrosine hydroxylase without severely affecting
cell viability.
ELF-MFs Exposure and Cell Treatments
For ELF-MF exposure (50 Hz, 1 mT), cells were seeded in 60- mm Petri dishes and placed in either
ELF-MF or sham incubator (under blind condition) 24 h after plating. Cells underwent a continuous
exposure to ELF-MF/sham over a time window of 24–72 h. MPP+ was administrated at a final

60
concentration of 0.5 mM. In both proliferating and differentiated conditions, MPP+ was administered
after 24 hrs of SHAM/ELF-MFs exposure and left in the medium for further 24 hrs until analyses.
Over treatment with MPP+, the exposure system was kept switched-off.
Global DNA Methylation Analysis
After bisulfite treatment of cell DNA, the methylation level of the retrotransposonic sequences Alu,
LINE-1 and Satα, generally used as a proxy of the sperm global methylation level [11, 12], were
assessed by specific PCR-pyrosequencing.
DNA was extracted and converted by a commercial kits (Zymo Research, Irvine, CA, USA)
according to the manufacturers protocol. Bisulfite treated DNA (50 ng) was amplified in a 50 μL
reaction mixture containing 25 μL of PyroMark PCR master mix (New England Biolabs, Ipswich, MA,
USA) and 0.2 μM of the primers specific for the human sequences.
Percentage of 5-methylcytosine were obtained through pyrosequencing analysis, using the
PyroMark ID Q24 system (Qiagen, Hilden, Germany) according to the manufacturer’s protocols.
Quantification of methylation levels was performed using the software provided. The percentage of
methylation (%5-mC) was expressed for each DNA locus as the number of 5-mC divided by the sum
of methylated and unmethylated cytosines. The LINE-1, and SATα methylation levels were
measured at 4 consecutive CpG sites; while for ALU sequence three CpG were considered. From
the quantitative results obtained by pyrosequencing, the average level of methylation for the whole
fragment was calculated.
Statistical Analysis
Statistical analysis of pyrosequencing results was performed using two-way ANOVA followed by
Tukey's multiple comparison tests. Significance was set atp<0.05. Results are presented as mean ±
standard deviation.
All statistical procedures were run using GraphPad Prism 6.
Results
In order to evaluate global DNA methylation, the analysis of the 5-methyl-cytosine percentage,
through bisulfite pyrosequencing assay, of three highly repeated sequences in mammals genome
was evaluated. The sequences analyzed were: the short interspersed nuclear sequence ALU, the
long interspersed nuclear element, LINE-1, and the centromeric satellite DNA, SATα. These
repetitive elements comprise roughly 66-69% of the human genome and consist of interspersed
repeats derived from non-autonomous or autonomous transposable elements [13]. They contain
numerous CpG dinucleotides, so their methylation status is relevant to understand global DNA
methylation levels [11], and have been also investigated in relation to human disease [12].
Experiments were performed in SH-SY5Y grown as proliferating and differentiated cells.
Furthermore, the possible contribute of PD neurotoxin MPP+ was investigated, since in a previous
paper we have demonstrated that 50-Hz 1 mT exposure sensitizes SH-SY5Y cells to the action of
this neurotoxin, by enhancing its ability to induce oxidative stress and cell apoptosis [10].
Proliferating cells
Cells grown in medium with serum were exposed up to 72 hours to 50 Hz 1 mT or SHAM field.
Comparing DNA methylation of all the sequences described before, no significant difference was
observed at any of different time points (24, 48 and 72 hrs). Furthermore, no change in ALU, LINE-1
and SATα DNA methylation in the same group (Sham or 50 Hz 1 mT) at different time points was
detected.
For MPP+ condition, cell were exposed for 24 hours to 50 Hz 1 mT or SHAM field and then treated
with the neurotoxin. Also with this experimental protocol no difference of percentage of
5-methilcytosine was found in ALU, LINE-1 and SATα sequences of DNA from ELF or SHAM
exposed cells in combination with MPP+. Moreover, no influence of PD neurotoxin alone was
demonstrated on DNA methylation of any of the three sequences.

61
Differentiated cells
Dopaminergic phenotype was strongly increased through the administration of RA and PMA as
described in materials and methods.
As for proliferating cells, differentiated ones were exposed up to 72 hours to 50 Hz 1 mT or SHAM
field and also in this condition no effect of magnetic field exposure was visible on global DNA
methylation. The contribute of MPP+ was investigated too, and, as for proliferating cells, no
influence of this neurotoxin was observed on DNA methylation of ALU, LINE-1 and SATα together or
not with 50 Hz 1 mT exposure.
The only strong difference in DNA methylation of the three sequences observed was between DNA
from proliferating cells, compared to DNA from differentiated ones. Indeed, differentiated cells
showed a significant reduction of LINE-1 methylation and a compelling increase of ALU and SATα
methylation.
Conclusion
In our experimental conditions we demonstrated that 50-Hz MF, in combination or not with PD
neurotoxin MPP+, does not affect global DNA methylation of neuronal in vitro model for PD.
It is important to highlight that the lack of influence of ELF-MFs exposure on global DNA
methylation, does not mean that these fields could not affect gene specific DNA methylation. Indeed
in our last paper we have demonstrated that 50 Hz 1 mT exposure decreases miR-34 b/c
expression by inducing a hyper methylation of its promoter [4].
So, even though results reported in this work can be considered as a step toward the
comprehension of action of ELF-MFs on cell DNA methylation, the investigation of the effect of ELF-
MFs on specific gene promoters methylation must be anyhow pursued, in order to dissect the
possible association among ELF-MFs exposure and human diseases.
References
[1] Giorgi G, Pirazzini C, Bacalini MG, et al. Radiat Environ Biophys. 2017 May;56(2):193-200.
[2] Liu Y, Liu WB, et al. PLoS One. 2015 Oct 6;10(10):e0139949.
[3] Liu Y, Liu WB, Liu KJ, et al. Biomed Res Int. 2015;2015:237183.
[4] Consales C, Cirotti C, Filomeni G, et al. Mol Neurobiol. 2017. doi: 10.1007/s12035-017-0791-0.
[5] Aguilera O, Fernandez AF, Munoz A, Fraga MF. J Appl Physiol2010; 109: 243–251
[6] Jaenisch, R. & Bird, A. Nat. Genet. 33 Suppl., 245–254 (2003).
[7]Hwang JY, Aromolaran KA, Zukin RS. Nat Rev Neurosci. 2017 May 18;18(6):347-361.
[8] Miranda-Morales E, Meier K, Sandoval-Carrillo A, et al. Front Mol Neurosci. 2017 Jul 18;10:225.
doi: 10.3389/fnmol.2017.00225
[9] Consales C,Merla C, Marino C, Benassi B. Int J Cell Biol 2012; 683897
[10] Benassi B, Filomeni G, Montagna C, et al. Mol Neurobiol. 2015 Jul 30.
[11] Consales C, Leter G, Bonde JP, et al. Hum Reprod. 2014 Sep;29(9):2065-72.
[12] Neven KY, Piola M, Angelici L, et al. BMC Genet. 2016 Jun 18;17(1):84.
[13] de Koning AP, Gu W, Castoe TA, et al. PLoS Genet. 2011;7(12):e1002384.
Figures

62
Figure 1. Neuroblastoma cell line, SH-SY5Y, grown as proliferating cells and stained with
ematossilin eosin at 40X magnification.

Figure 2. Experimental setting.SH-SY5Y differentiated or not were exposed up to 72 hours to

a 50 Hz 1 mT magnetic or SHAM field. Global DNA methylation was analyzed just after the
end of the exposure.For ELF-MFs and MPP+ combination, cells were exposed for 24 hours to
magnetic or SHAM field and then aministrated with0.5 mM MPP+ for other 24 hours. During
MPP+ treatment the exposure system was kept switched-off. Global DNA methylation
wasanalyzed just after the end of MPP+ treatment.

63
 Figure 3. Example of pyrosequencing analysis. The pyrogram of SAT alpha sequence is
shown. In the rigth upper side SAT alpha sequences to be analyzed is reported and cytosines
which can be methylated are indicated as red Y. The pyrogram displays the percentage of
methylation of each single cytosine, then the software calculates the average percentage of
the seqeunce considered.

S03-3 [17:00]
Biological effect of the 50 Hz Magnetic Field in a three-dimension (3D) in vitro experimental
model of SH-SY5Y human neuroblastoma cells
Claudia Consales1, Alessio Butera2, Mariateresa Mancuso1, Emanuela Pasquali1, Vanni Lopresto1, Rosanna
Pinto1, Carmela Marino1 & Barbara Benassi1
1ENEA Italian National Agency for New Technologies, Energy and Sustainable E, Rome, Italy, 00123
2Experimental Medicine and Surgery, University of Rome Tor Vergata, Rome, Italy, 00133
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: Barbara Benassi
Given the reported limitations of the conventional 2D in vitro cultures of neuronal cells in terms of
response to drugs and radiations, we here aimed at evaluating the effects elicited by the 50-Hz
magnetic field (MF) on the human neuroblastoma SH-SY5Y cells grown in a new three-dimension
(3D) scaffold in vitro. We demonstrate that (i) no change in proliferation, distribution of cell cycle
phases and viability, and no induction of apoptosis is triggered by the MF in the SH-SY5Y cells if
cultured in the 3D scaffolds; (ii) the intracellular GSH content, as well as the expression of both the
gamma-GCS (responsible for the GSH biosynthesis) and SOD1 genes are impaired by the MF
exclusively in cells grown in the 3D culture conditions.
Introduction
The culturing of primary neurons and neuronal-like cancer cells on conventional two-dimensional
(2D) surfaces is the mainly used in vitro experimental model for both oncologic and
neurodegenerative disorders. The limitations of such growing conditions are recently emerging, as

64
2D cultures lack the complex anatomical and functional connectivity of the neuronal network (1-3).
To provide a more physiological structural and biochemical system that more closely resemble the in
vivo environment, different three-dimensional (3D) matrices have been exploited, including
microporous polystyrene scaffolds, fibrin matrices, agarose, matrigel and collagen hydrogels (4-6).
Comparisons of cell growth in standard 2D monolayer cultures and 3D matrix have shown clear
phenotypic differences in cellular surface area, stress fiber distribution, cellular migration and
adhesions, neurite growth and dimensions, protein and gene expression, and epigenetics
(7,8).Moreover, cellular response to drugs and ionizing radiations is significantly affected by culture
conditions (9,10). In this context, the exposure to the electromagnetic fields (EMFs) of cells grown in
3D systems has been gaining growing interest, mainly for tissue regeneration purposes. Exposure
to extremely low frequency (ELF) magnetic fields (MFs) has been tested in primary bovine
chondrocytes seeded in 3D polyurethane scaffolds to stimulate chondrogenesis and cartilage
maturation (11), and in epidermal stem cells seeded in collagen sponge scaffolds for improving
wound healing (12). 3D cultures have been also characterized in response to mobile phone
frequencies, as reported in cardiac myocyte cells obtained from chicken embryos undergoing pulsed
EMFs (900 MHz, 8.3 Hz pulse; 100 and 250 mW/kg SAR) (13).To the best of our knowledge, there
are no experimental data assessing the response of neuronal cells grown in 3D matrices to ELF-
MFs. We thus aimed at comparing the effect triggered by the ELF MF (50 Hz, 1 mT) in the SH-SY5Y
human neuroblastoma cells grown in both conventional 2D monolayer culture and in a novel 3D
scaffold, to verify whether and how the 3D structural environment might affect the biological
response to radiation.
Methods
3D Cell Culture System
For 3D cultures, the commercial 200 μm thick polystyrene scaffolds (Alvetex, ReproCELL, Durham,
UK) were used (figure 1a(i)), and manipulated according to manufacturers’ instructions
(http://reinnervate.com/alvetex). To render the scaffold hydrophilic, inserts were first fully submerged
in 70 % ethanol for 10 min, then washed twice with sterile water and incubated with poly-L-lysine (10
μg/ml, SIGMA-Aldrich, Milan, Italy) for 2 h. After coating, the inserts were washed with PBS and
finally incubated in serum containing media for 2 h at 37°C and 5 % CO2 before plating cells.
2D-3D cell culture systems and exposure to ELF MF
Human neuroblastoma SH-SY5Y cells were cultured as previously described (14,15). Before
exposing cells to the MF, different set-up experiments in the new 3D scaffolds have been carried out
to optimize the cellular density, the visualization of cell by staining with Neutral Red (SIGMA-Aldrich)
(figure 1a(ii)), and the yield of cell amount, proteins and RNA from scaffolds for further analyses.For
exposure experiments, 3x105 cells were seeded in either the 60-mm Petri dish or the Alvetex
scaffold. Twenty-four hours after plating, cells (in both 2D and 3D systems) underwent a continuous
72 hours exposure to either the Sham or the ELF-MF (50 Hz, 1 mT) under blind conditions, as
previously detailed (14,15).
Viability assay
The assessment of viability was carried by the MTT test (SIGMA-Aldrich), according to Alvetex’s
protocol (http://reinnervate.com/alvetex/alvetex-documentation/protocols) (figure 1a(iii)).
Immunohistochemistry (IHC) and immunofluorescence (IF)
The hematoxiline and eosin staining (H&E), as well as the evaluation of the proliferative and
apoptotic index was carried out in both formaldehyde-fixed 2D cultures (by IF) and in formalin-
embedded 3D slices (by IHC, according to specific Alvetex protocols) as previously detailed (16).
Flow Cytometric Analysis
The evaluation of the DNA content for cell cycle analysis was carried as previously described (14).
RNA Extraction, Reverse Transcription, and Gene Expression Analysis
Total RNA was extracted by Trizol® protocol (Invitrogen, Thermo Fisher Scientific, Waltham, MA,

65
USA) from both 2D and 3D cultures. Retro-transcription and amplification by real-time PCR for gene
expression analysis have been performed as previously detailed (15).
Reduced Glutathione (GSH) content assay
It was carried out as previously described (17).
Statistical analysis
The variations of samples values are reported as Mean ± S.D. calculated in N ≥ 3 replicates. The
statistical differences were analyzed trough the KailedaGraph program (Synergy Software, Reading
PA, USA) by applying the two-sided Student’s t test. P values < 0.05 were considered statistically
significant.
Results
When grown under basal (unexposed) conditions, SH-SY5Y cells proliferate and distribute in the 2D
plates and in the 3D scaffolds as reported in figure 1b. The cell cycle distribution and sub-G1
percentage do not differ between the two growing conditions (figure 1c). In terms of genes
controlling the proliferation and the progression through the S-phase, we reported a statistically
significant decrease in the Ki67 mRNA levels in cells grown in the 3D compared to 2D conditions,
with no difference in the levels of MYCN and cyclin D1 (CCND1) genes (figure 1d).
We then exposed both the 2D and 3D cultures to either the ELF-MF (50 Hz, 1mT) or the Sham
system for 72 hours. By MTT test, we first demonstrated that no change in cell proliferation and
viability was triggered by ELF exposure in the SH-SY5Y cells growth in both experimental conditions
(figure 2a); consistently, no difference of cell distribution in the cell cycle and sub-G1 phases was
observed uponELF exposure in both 2D and 3D culture conditions (figure 2b,c). We also verified the
expression levels of MYCN, CCND1 and Ki67 following ELF exposure, and demonstrated that no
change in gene expression was triggered by the 50 Hz-MF in cells grown in the conventional 2D
monolayer nor in the 3D cultures (figure 2d). This is consistent with what demonstrated by IHC/IF, as
no difference in the proliferation biomarkers was observed following ELF exposure (figure 2e).
Moreover, no induction of apoptosis was demonstrated in response to ELF in both experimental
conditions (figure 2f).
We finally analyzed the modulation of redox pathways in both 2D and 3D SH-SY5Y cultures
following either Sham or ELF exposure. We demonstrated that the intracellular GSH content was
significantly reduced in response to the MF exclusively in the 3D conditions (figure 3a), with no
significant variation observed in the 2D cultures. The GSH depletion was accompanied by a
significant impairment in the expression of the gamma-glutamylcysteine synthase (gamma-GCS)
gene, the enzyme catalyzing the first and rate-limiting step in GSH biosynthesis. In particular, the
gamma-GCS heavy chain (gamma-GCSh) mRNA level underwent a significant decrement, but
solely in the 3D scaffold growing conditions (figure 3b). The levels of the gamma-GCS light chain
(gamma-GCSl) were also tested, but no variation was reported in response to ELF
exposure.Moreover, the superoxide dismutase SOD1 gene expression was significantly depleted in
SH-SY5Y cells grown in the 3D scaffold compared to the 2D conventional monolayers (figure 3c).
The evaluation of other antioxidant enzymes, such as catalase and GSH peroxidase (in terms of
expression and enzymatic activity) is in progress, as well as the investigation of possible oxidative
damage to both DNA and RNA.
Conclusions
We here evaluate for the first time the biological response to the 50 Hz MF of the SH-SY5Y
neuronal-like cells grown in a 3D scaffold conditions compared to the conventional 2D monolayer
system. We demonstrate that no change in proliferation and viability, and no induction of apoptosis
is triggered in the SH-SY5Y cells by ELF exposure. This is in line to what previously reported by our
group in this cell line in the 2D conventional cultures (14,15); we here corroborate these data by
demonstrating that, also in a 3D scaffold experimental condition, the exposure to the 50 Hz MF does
not affect cell growth and neuronal death.We further demonstrate that the intracellular GSH content,

66
the expression of its biosynthetic enzyme (gamma-GCSh), and the antioxidant SOD1 levels are
impaired by ELF exposure exclusively in cells grown in the 3D scaffolds, whereas no change is
reported in the same experimental model if cultured in the 2D conventional monolayer. This finding
strengthen the well-established pro-oxidant ability of the ELF MF in neuronal cells, and identify in
the impairment of the gamma-GCSh expression a new key determinant of such oxidative
imbalance.Finally, our findings shed the light on the limitations of the 2D conventional systems in
favor of newly emerging 3D matrixes that, as here evidenced, can allow the detection of biological
variations in response to ELF MF otherwise not detectable by conventional experimental set-ups.
References
1. Bosi S et al. (2015) Sci. Rep. 5:9562.
2. D’Avanzo C et al. (2015) 37(10): 1139–1148.
3. Watson PMD et al. (2017) SLAS Discov 22(5):583-601.
4. Hayman MW et al. (2004) Biochem Biophys Res Commun 314, 483.
5. Yamada KM and Cukierman E (2007) Cell 130, 601–610.
6. O’Connor SM et al. (2001) Neurosci Lett 304, 189.
7. Li G et al. (2007) Tissue Eng 13(5):1035-47.
8. Stevanato L and Sinden JD (2014) Stem Cell Res Ther (2):49.
9. Edmondson R et al. (2016) PLoS One 11(6):e0158116.
10. Gomez-Roman N et al. (2017) Neuro Oncol 19(2):229-241.
11. Hilz FM et al (2014) Bioelectromagnetics 35:116-128.
12. Bai W et al. (2017) Bioelectromagnetics 38:204-212.
13. Daus AW et al. (2011) Bioelectromagnetics 32:351-359.
14. Benassi B et al. (2016) Mol Neurobiol 53(6):4247-4260.
15. Consales C et al. (2017) Mol Neurobiol. 2017 Oct 16.
16. Tanori M et al. (2013) Stem Cells 31(11):2506-16.
17. Benassi B et al. (2012) 2:236–247.
Figures

67
Figure 1. a (i) Representation of the Alvetex Scaffold (from http://reinnervate.com/alvetex);
staining of the SH-SY5Y 3D cell culture by the (ii) non-toxic dye Neutral Red, and (iii) 3-(4,5-
Dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide tetrazole salt (MTT test). (b) Staining of
the 2D and 3D cultures by H&E. (c) Evaluation of the cell cycle percentages (left), and
representative DNA content histograms obtained by FACS analysis (right). (d) Gene
expression analysis of the reported transcripts carried out by real-time PCR. Statistical
analysis: *P < 0.05; **P < 0.01; ***P < 0.001; n.d.: variation not statistically different.

68
Figure 2. (a) Cell viability test performed by MTT assay. (b) Representative DNA content
histograms obtained by FACS analysis. (c) Evaluation of the cell cycle percentages carried out
by FlowJo software. (d) Gene expression analysis of the reported transcripts carried out by
real-time PCR; (e) Evaluation of the proliferative index carried out by IF staining of the
ki67/DAPI markers in 2D cells, and by IHC of the PCNA marker in the 3D cells. (f) Assessment
of the apoptotis percentage performed by active Caspase-3 staining in 2D (IF) and 3D (IHC)
cultures. Statistical analysis: *P < 0.05; **P < 0.01; ***P < 0.001; n.d.: variation not statistically
different.

69
 Figure 3. (a) Evaluation of the intracellular reduced GSH content. Analysis of the (b) gamma-
GCSh and gamma-GCSl, and (c)SOD1 gene expression level carried out by real-time PCR.
Statistical analysis: *P < 0.05; **P < 0.01; ***P < 0.001; n.d.: variation not statistically different.

S03-4 [17:15]
8-oxoG DNA glycosylase-1 inhibition sensitizes Neuro-2a cells to oxidative DNA base
damage induced by 900 MHz radiofrequency electromagnetic radiation
Zhou Zhou1, 3, Lei Zhang2, 3, Chuan Liu2, 3, Qin-Long Ma2, 3 & Zheng-Ping Yu2, 3
1Department of Occupational and Environmental Health, Zhejiang University, Hangzhou, China, 310058
2Department of Occupational Health, Army Medical University (Third Military Medical University), Chongqing,
China, 400038
3Key Laboratory of Medical Protection for Electromagnetic Radiation, Ministry of Education, Army Medical
University (Third Military Medical University), Chongqing, China, 400038
Keywords: In vitro, RF/Microwaves, Completed (published)

70
Presented by: Zhou Zhou
This study provided information about the in vitro putative genotoxicity during cell exposure to
radiofrequency electromagnetic fields (RF-EMFs) with or without 8-oxoG DNA glycosylase-1
(OGG1) siRNA using Neuro-2a cells. Exposure to 900 MHz RF-EMFs with insufficient energy could
induce oxidative DNA base damage in Neuro-2a cells. These increases were concomitant with
similar increases in the generation of reactive oxygen species (ROS). With OGG1 siRNA, RF-EMFs
could cause DNA base damage in Neuro-2a cells as low as 1 W/kg. This suggested that OGG1 is
involved in the process of DNA base repair and may play a pivotal role in protecting DNA bases
from RF-EMF induced oxidative damage.
Background/Aims: The purpose of this study was to provide information about the in vitro putative
genotoxicity during cell exposure to radiofrequency electromagnetic fields (RF-EMFs) with or without
8-oxoG DNA glycosylase-1 (OGG1) siRNA using Neuro-2a cells. Metheds: Neuro-2a cells treated
with or without OGG1 siRNA were exposed to 900 MHz Global System for Mobile Communication
(GSM) Talk signals continuously at a specific absorption rate (SAR) of 0, 0.5, 1 or 2 W/kg for 24 h.
DNA strand breakage and DNA base damage were measured by the alkaline comet assay and a
modified comet assay using formamidopyrimidine DNA glycosylase (FPG), respectively. Reactive
oxygen species (ROS) levels and cell viability were monitored using the non-fluorescent probe 2,7-
dichlorofluorescein diacetate (DCFH-DA) and CCK-8 assay. Results: Exposure to 900 MHz RF-
EMFs with insufficient energy could induce oxidative DNA base damage in Neuro-2a cells. These
increases were concomitant with similar increases in the generation of reactive oxygen species
(ROS). Without OGG1 siRNA, 2 W/kg RF-EMFs induced oxidative DNA base damage in Neuro-2a
cells. Interestingly, with OGG1 siRNA, RF-EMFs could cause DNA base damage in Neuro-2a cells
as low as 1 W/kg. However, neither DNA strand breakage nor altered cell viability was observed.
Conclusion: Even if further studies remain conducted we support the hypothesis that OGG1 is
involved in the process of DNA base repair and may play a pivotal role in protecting DNA bases
from RF-EMF induced oxidative damage.

S03-5 [17:30]
Do GSM-1800 induce DNA damage and genomic instability in nerve cells?
Isabelle Lagroye1, 2, Emmanuelle Poque-Haro2, Rémy Renom2, Florence Poulletier De Gannes2, Corinne El
Khoueiry2, Yann Percherancier2 & Bernard Veyret1, 2
1EPHE, PSL, Paris, France, 75014
2IMS laboratory UMR5218, Bordeaux University, Talence, France, 33405
Keywords: In vitro, RF/Microwaves, Completed (unpublished)
Presented by: Isabelle Lagroye
DNA damage was investigated in rat primary neurons and SH-SY5Y human neuroblastoma cells
exposed to a GSM-1800 signal for 24 hr at SAR levels of 0.375, 1.5 and 6 W/kg, with or without
menadione and methylmethane sulfonate. Both the alkaline comet assay and the micronuclei assay
were used. We found no indication that GSM-1800 could lead to immediate DNA damage nor
delayed DNA damage indicative of genomic instability. Coexposure experiments with chemicals in
neuroblastoma cells showed that a specific combination of GSM-1800 at 6 W/kg and Menadione
could transiently lead to increased micronuclei. Further investigation is needed to determine the
underlying mechanism, although this type of exposure is unlikely in the daily life.
Introduction
In the present work, the potential role of RF in cancer, behavioural effects, and neurodegenerative
diseases is addressed throughin vitroexperiments and the assessment of DNA damage in rat
primary neurons and genomic instability in human neuroblastoma cells exposed to a GSM-1800
signal.

71
Material and Methods
SH-SY5Y human neuroblastoma cells and rat cortical primary neurons were cultured as previously
reported (Poulletier de Gannes et al. 2011). Cells were plated on 35 mm diameter petri dishes and
exposed to GSM-1800 RF signal using the SXC-1800 waveguides (IT’IS-Foundation, Zürich,
Switzerland). This setup used R18 hollow rectangular waveguides operating at a frequency of
1800MHz (Schuderer et al. 2004). Six petri dishes were exposed simultaneously at the H-field
maxima. Two waveguides (exposed and sham-exposed) are placed inside a commercial incubator
to ensure constant environmental conditions (37°C, 95% air/5% CO2, 95% humidity). Temperature
and SAR were monitored at the location of the dishes during exposure and the temperature
difference between the two waveguides never exceeded 0.1°C. To ensure blind exposure, the
computer randomly determined which of the two waveguides was activated. Average SARs of
0.375, 1.5, and 6W/kg were selected and RF exposures lasted 24 hours. To test for potential
interaction with known chemicals, Menadione (MQ) and Methyl Methane Sulfonate (MMS) were
added to the culture medium for 1 hour after RF exposure. The alkaline comet assay was adapted
from Olive (1999). A total of 100 images from duplicate slides were analysed per exposure condition
using the Comet Assay IV software (Perceptive Instrument, UK). Olive tail moment (OTM) and the
percentage of DNA intensity in the tail were used as DNA damage parameters. For the micronuclei
assay, cells were incubated for 5 days after exposures, to allow micronuclei to form during three
subsequent cell divisions. The number of micronuclei was assessed using flow cytometry (BD
FACSCanto IITMflow cytometer, France) equipped with BD FACSDiva software v.8.0.1 and thein
vitroMicroflow®kit (Litron Laboratoires New York, USA). Cells were stained according to the supplier
protocol and 10000 gated events were acquired from each sample. Immediate effects of a 24-hour
exposure to GSM-modulated RF ± chemicals was assayed in both cell types. In human SH-SY5Y
cells, delayed effects were also assessed at 8, 15, 30 and 45 days after the exposures for DNA
damage and 5 and 30 days after GSM-1800 exposure for micronuclei assay. Statistical analysis
used the non-parametric Kruskall-Wallis test for small samples followed, if significant, by the Mann-
Whitney test (Anastats®).
Results
We reported at BioEM2017 that in both cell types, 24 hours of exposure to GSM-1800 RF had no
effects on immediate DNA damage and did not affect chemically-induced DNA damage whatever
the SAR tested. Here we report on data obtained using the micronucleus assay and delayed effects
on DNA. Fig. 1 shows the data obtained in human SH-SY5Y neuroblastoma cells using the comet
assay, at days 0, 8, 15, 30, and 45 after exposure. Up to 45 days after exposure, we found no
evidence of genomic instability induced by GSM-1800 exposure whatever the SAR tested.
Chemicals induced immediate DNA damage but no indication of genomic instability could be found
up to 45 days after treatment.At the doses used, neither MQ nor MMS induced micronuclei at either
5 or 30 days after RF exposure in human SH-SY5Y neuroblastoma cells. GSM-1800 induced no
micronuclei immediately and 30 days after a 24-hour RF exposure. While no interaction could be
detected with MMS, co-exposure to GSM-1800 and MQ significantly increase micronuclei. This
effect was transient since no more detected 30 days after RF exposure.
Conclusion
Exposure to GSM-1800 did not induce immediate DNA damage in primary neurons and SH-SY5Y
neuroblastoma cells nor delayed DNA damage in the neuroblastoma cells. GSM-1800 at 6 W/kg
may interact with MQ but not MMS. This suggests that RF could increase the mitochondrial ROS
production, although this was not tested in the present work.Further investigation is needed to
determine the underlying mechanism, although this type of exposure is unlikely in the daily
life.Finally, we found no indication that either alone or in combination with chemicals, GSM-1800
could lead to delayed DNA damage indicative of genomic instability.
References
F. Poulletier de Gannes, E. Haro, A. Hurtier, M. Taxile, G. Ruffié, B. Billaudel, B. Veyret and I.

72
Lagroye. Effect of Exposure to the Edge Signal on Oxidative Stress in Brain Cell Models. Radiation
Research 175, 225–230 (2011).
J. Schuderer, T. Samaras, W. Oesch, D. Spät and N. Kuster. High peak SAR exposure unit with tight
exposure and environmental control for in vitro experiments at 1800 MHz. IEEE Trans. Microwave
Theory Tech. 52: 2057–2066 (2004).
P.L. Olive. DNA damage and repair in individual cells: applications of the comet assay in
radiobiology. Int J Radiat Biol. 75(4):395-405 (1999).
Acknowledgment
The research leading to these results has received funding from the European Community’s
Seventh Framework Programme (FP7/2007-2013) under grant agreement no. 603794 (the
GERONIMO project)
Figures

Figure 1. Immediate and delayed DNA damage in human SH-SY5Y neuroblastoma cells after
a 24 h-exposure to 1800 MHz, GSM-modulated RF at 6 W/kg or sham exposure followed with
1 h treatment with menadione (15 μM) or MMS (150µg/mL). Olive Tail Moment is an arbitrary
unit. Error bars represent SEM from 5-6 independent experiments with 100 nuclei analyzed in
each sample (thus, a total of 500-600 nuclei were analyzed)

S03-6 [17:45]
Cellular internalization and intratumoral distribution of prospectively therapeutic magnetic
nanoparticles clusters submitted or non-submitted to pulsed electric fields
Jelena Kolosnjaj-Tabi1, Slavko Kralj2, Laure Gibot1, Geraldine Gasc1, Isabelle Fourquaux3, Muriel Golzio1 &
Marie-Pierre Rols1

73
1IPBS, CNRS UMR 5089, Toulouse, France, 31077
2Department for Materials Synthesis - Jožef Stefan Institute, Ljubljana, Slovenia, 1000
3CMEAB, University Toulouse III -Paul Sabatier, Toulouse, France, 31062
Keywords: In vitro, Pulsed, Work in Progress
Presented by: Jelena Kolosnjaj Tabi
The intracellular localization and the (therapeutic) effects of nanoparticles might depend on the
processes by which the nanoparticles had entered the cell. In order to evaluate different
internalization patterns, cultured cells have been incubated with superparamagnetic iron oxide
nanoclusters. Different incubation protocols have been applied, including the application of sub-
permeabilizing and permeabilizing pulsed electric fields, and the efficiency and impact of the applied
cell labeling protocol has been assessed.
Introduction
Inorganic magnetic nanoparticles, especially the superparamagnetic iron oxide ones, are of
particular therapeutic interest because they can be used as diagnostic and therapeutic agents.
While their diagnostic potential for MRI has been extensively reported, and their therapeutic assets
for magnetic hyperthermia are currently exploited in clinics, recent preclinical studies indicate that
iron oxide nanoparticles have also an enormous potential for mechanical tumor destruction (1).
Irrespective of the therapeutic goal, the main challenge remains the selective loading of the cell with
a sufficient (often high) amount of nanoparticles (2). In order to improve the cellular internalization of
superparamagnetic iron oxide nanoclusters (3), which are promising tools for mechanical (and thus
non-thermal) destruction of the tumor, we optimized physical parameters, to improve the cell loading
after different protocols of cell labeling with mentioned nanoparticles.
Methods
Red fluorescent silica coated superparamagnetic iron oxide nanoclusters were provided by Nanos
SCI (Nanos Scientificae Ltd.). Nanoclusters syntheses rely on the assembly of maghemite (γ-
Fe2O3) nanoparticles into clusters, which are coated with a silica shell loaded with rhodamine.
Green fluorescent protein-expressing human colon cancer cells (HCT116 GFP) and murine RAW
264.7 macrophages expressing green fluorescent protein actin (RAW 264.7-Act-GFP) have been
cultured in Dulbecco’s modified Eagle’s medium supplemented and RPMI 1640, respectively. Both
media were supplemented with fetal bovine serum (10 %) and 100 units/mL penicillin/streptomycin.
Cell labelling procedures have been performed in serum-free and penicillin/streptomycin-free RPMI
1640 medium, supplemented or not supplmemented with 5 mM citrate, or in the electropulsation
buffer medium (10 mM phosphate pH 7.4, 1 mM MgCl2, 250 mM sucrose). Different nanoparticle
concentrations (1 to 10 mM [Fe]) and varying incubation times (15 minutes to 2 hours) and chase
times (15 minutes to 7 days) have been applied.
Female nude BALB/c mice were subcutaneously injected with 2 million HCT116 GFP cells
suspended in 50 µL of PBS. When the tumor volume reached 40 to 200 mm3 (calculated by the
formula V=Dxd2/2, where D is the largest diameter and d is the diameter perpendicular to D), the
tumors underwent treatment (nanoparticle injection with 6.5 mmol Fe and/or pulsed electric fields).
Pulsed electric fields have been generated by an ELECTRO cell S20 generator (βTech, France)
delivering square-wave electric pulses. Two stainless-steel plate parallel electrodes (length = 10
mm, inter-electrode distance = 4, 6 or 10 mm), connected to the voltage generator have been used.
Electropulsation of cells has been performed within the electropulsation buffer medium (described in
section “Cell labeling procedures”). The number and duration of applied electric pulses, as well as
their frequency and electric field intensity varied.
Fluorescence microscopy (FM) and Transmission Electron Microscopy (TEM) were used to assess
the intracellular localization of nanoparticles. The FM settings were selected according to the
examined fluorophore and standard preparation procedures applied for TEM samples preparation.

74
Results
Silica coated superparamagnetic iron oxide nanoclusters (Figure 1A) were efficiently introduced into
cells (Figure 1B, C). Different procedures resulted in varying nanoparticles accumulation, regardless
of the fact that the cells were incubated at the same iron concentration (5 mM [Fe]) and same
incubation time (1h). The modulation of the labeling parameters resulted in different nanoparticle
loading within the cells (Figure 2). In the incubation procedure, which was carried in citrate–
supplemented medium, nanoparticles aggregated less and were distributed more homogeneously
within the cells.
In vivo, nanoparticles have a different intratumoral distribution, if the tumor is primed with pulsed
electric fields prior nanoparticle administration (Figure 3). In the tumor, which had been submitted to
pulsed electric fields (Figure 3A), nanoparticles accumulate less in the tumor capsula, and more
throughout the tumor stroma, mainly within the matrix disseminated around tumor cells. Conversely,
when the tumor is not submitted to pulsed electric fields, nanoparticles are mainly located within the
tumor capsula and the central necrotic zone (Figure 3B).
Discussion
In order to exert mechanical destruction of the tumor cells and the tumoral stroma, magnetic iron
oxide nanoparticles should have strong magnetic moments. This was obtained after organizing
magnetic nanoparticles into complex hierarchical structures (clusters). Clusters may additionaly
increase their magnetic moments, when they organize into chains in situ, after exposure to an
external magnetic field. In order to form chains in situ (in tumor cells or within the tumoral matrix),
the nanoparticles formulation and application procedure has to be optimized, to allow the
organization of clusters into chains. Indeed, physical parameters variations (2) can be applied to the
nanoparticle formulation and to cell incubation parameters and can improve nanoparticles
internalization. In addition, nanoparticle internalization can be increased by acting directly on the cell
or the tumor. Physical cues, such as the pulsed electric fields, can prime the tumor prior
nanoparticles administration. When nanoparticles are administered to the tumor, their distribution
follows a different pattern if the pulsed electric fields are applied, and this might strongly affect the
therapeutic outcome.
Conclusion
This study presents different labeling approaches involving silica coated superparamagnetic iron
oxide nanoclusters. In order to allow high nanoparticle loading we tested different labelling
approaches in vitro and applied them in vivo to tumor bearing mice. The application of pulsed
electric fields seems a promising method, which primes the tumor and allows a more homogenous
nanoparticle distribution within the tumor.
References
1. Mansell, R., Vemulkar, T., Petit, D. C., Cheng, Y., Murphy, J., Lesniak, M. S., & Cowburn, R. P.
(2017). Magnetic particles with perpendicular anisotropy for mechanical cancer cell
destruction.Scientific reports,7(1), 4257.
2. Kolosnjaj-Tabi, J., Wilhelm, C., Clément, O., & Gazeau, F. (2013). Cell labeling with magnetic
nanoparticles: opportunity for magnetic cell imaging and cell manipulation.Journal of
nanobiotechnology,11(1), S7.
3. Kralj, S., & Makovec, D. (2015). Magnetic assembly of superparamagnetic iron oxide nanoparticle
clusters into nanochains and nanobundles.ACS nano,9(10), 9700-9707.
Figures

75
Figure 1. A) Representative TEM micrograph of silica coated superparamagnetic iron oxide
nanoclusters. B) FM micrographs showing HCT-116 GFP cells (green color) internalizing
discrete silica coated superparamagnetic iron oxide nanoclusters (red dots) (Magx40) C) FM
micrographs showing HCT-116 GFP cells (green color) internalizing aggregates of silica
coated superparamagnetic iron oxide nanoclusters (red dots) (Magx40).

76
Figure 2. TEM micrographs showing HCT-116 GFP cells, which internalized A) small
aggregates of silica coated superparamagnetic iron oxide nanoclusters, and B and C) large
quantities of nanoclusters. (Red arrows point to nanoparticles and red square indicates the
magnified view on the right side. N denotes the cell’s nucleus).

77
 Figure 3. Macroscopic fluorescence images showing the green fluorescing tumor injected with
red fluorescent of silica coated superparamagnetic iron oxide nanoclusters. A) Nanoparticles
were injected after the application of electric pulses (N= 8 pulses, pulse duration = 100 µs,
frequency 1000 µs, 1300 V/cm) and were preferentially localized within the matrix
disseminated around tumor cells, and B) Nanoparticles were injected into a tumor that was not
exposed to pulsed electric fields, and were mainly localized within the tumor core or within the
tumor capsula.

Session: S04
MRI studies
Monday June 25, 2018 • 16:30 - 18:00
Europa D
Chairs: Manuel Murbach & Marthinus Van Wyk

S04-1 [16:30]
STUDENT PAPER
Occupational exposure to MRI-related magnetic stray fields and sleep quality among
radiographers - A cross-sectional study in the Netherlands
Emre Ozdemir1, Kristel Schaap1, Anke Huss1 & Hans Kromhout1
1Division of Environmental Epidemiology, Institute for Risk Assessment Sciences, Utrecht, the Netherlands,
3508 TD
Keywords: Epidemiology, Static, Review, Commentary, Recommendation, Evaluation
Presented by: Emre Ozdemir
Summary: In this cross-sectional study, we investigated the potential association between
occupational exposure to MRI-related magnetic stray fields and sleep quality among Dutch
radiographers. We observed decreased sleep quality among radiographers who worked with or near
MRI scanners.
Background
Previous studies have consistently shown that working with or near MRI scanners may result in
experiencing transient health effects [1-3]. In addition, several pilot studies found that self-reported
sleeping problems were high among personnel who worked in an MRI room [4,5].
Objectives

78
Our aim was to investigate the potential association between occupational exposure to MRI-related
magnetic stray fields and sleep quality in a cross-sectional study among registered radiographers in
the Netherlands.
Methods
In 2013, online questionnaires were sent to members of the Dutch Society of Radiographers
(NVMBR), inquiring about their lifestyle, health condition, sleep quality, and work practices. Of the
1637 invitees, 490 (30%) responded and filled in questions regarding their exposure, sleep quality
and quantity. Occupational exposure to MRI-related magnetic stray fields was assigned if
participants reported they had entered an MRI room in either 12 months and/or 4 weeks prior to the
survey. Subsequently, we grouped participants as unexposed (not exposed in the past 12 months),
past exposure (not exposed in the past 4 weeks) and recently exposed (exposed in the past 4
weeks). Sleep quality was assessed using the MOS-12 item sleep scale [6]. We used linear and
logistic regression to test associations between stray field exposures and sleep quality, adjusting for
a range of potential confounding factors, such as work-related stress and medication use.
Results
Results were suggestive of an association between stray field exposure in the past 4 weeks and
increased sleep disturbance, and reduced adequacy of sleep. The sleep domains, (overall) sleep
problem index, daytime somnolence, snoring and respiratory problems were not affected by stray
field exposures. Tests for linear trends indicated that the number of days on which radiographers
entered an MRI room in the past 4 weeks, was positively associated with risk of having non-optimal
sleep duration.
Conclusions
In this cross-sectional study, we observed decreased sleep quality among radiographers who
worked with or near MRI scanners. The present study provides new leads for further studies on
sleep quality.
References
[1] De Vocht, F., Van Drooge, H., Engels, H., & Kromhout, H. (2006). Journal of Magnetic
Resonance Imaging, 23(2), 197–204.
[2] LE, van N., Slottje, P., MJ, van Z., F, de V., & Kromhout, H. (2012). Occupational and
Environmental Medicine, 69(10), 759–766.
[3] Schaap, K., Christopher-de Vries, Y., Mason, C. K., de Vocht, F., Portengen, L., & Kromhout, H.
(2014). Occupational and Environmental Medicine, 71(6), 423–429.
[4] Wilén, J., & de Vocht, F. (2011). European Journal of Radiology, 80(2), 510–513.
[5] Zanotti, G., Ligabue, G., & Gobba, F. (2015). Electromagnetic Biology and Medicine, 34(3),
262–264.
[6] Hays, R. D., & Stewart, A. L. (1992). Measuring functioning and well-being: The Medical
Outcomes Study approach (pp. 235-259), Durham, NC: Duke University Press.

S04-2 [16:45]
Survey of the facilities’ policies on allocating pregnant staff for MRI site and an attitude
survey to non-ionizing radiation in Japan
Sachiko Yamaguchi-Sekino1, Shuhei Izawa1, Humio Maeyatsu2, Tsukasa Doi3, Takeo HIkichi4, Hideki Fujita5,
Shinya Imai6, Manabu Akahane7 & Rui-Sheng Wang1
1National Institute of Occupational Safey and Health, Japan, Kawasaki, Japan, 2148585
2Izumi Hospital, Sendai, Japan, 9813212
3Takai Hospital, Tenri, Japan, 6320006

79
4Hirose Hospital, Sendai, Japan, 9893121
5Osaka Saiseikai Nakatsu Hospital, Osaka, Japan, 5300012
6Faculty of Health Sciences, Butsuryo Colleage of Osaka, Osaka, Japan, 5938328
7Department of Public Health, Health Management and Policy, Nara Medical University, Kashihara, Japan,
6348521
Keywords: Occupational, Static, Completed (unpublished)
Presented by: Sachiko Yamaguchi-Sekino
In this study we surveyed policies on allocating MRI scan duties to pregnant employees and
investigated the influence of respondents’ level of awareness of non-ionizing radiation. The results
revealed inconsistent handling within Japan, and that although duties may be restricted in some
situations, 7.8% of facilities increase the opportunity of allocation to MRI duties proactively
compared with the current situation while 51.8% of facilities opt to reduce the allocation to MRI
duties. And it was also observed that the level of awareness of non-ionizing radiation influences the
policy on the allocation of MRI scan duties for pregnant employees.
Background
There have been no clear reports to date of MRI scans exerting a harmful effect on fetal
development and growth, but there are disparate views internationally on pregnant employees
performing MRI scan duties. Although governments and academic societies around the world have
issued recommendations [1], [2], operating criteria differ at the level of the individual hospitals [3].
Japan has the highest number of MRI devices per capita of any OECD country with MRI in
operation at around 6,000 facilities. Since no legislation exists in Japan to protect employees from
non-ionizing radiation, it is estimated that policies on allocating pregnant staff to perform MRI scans
differ between facilities. Therefore, in this study we sent questionnaires to facilities (respondent was
a personnel responsible for MRI scan duties) equipped with MRI devices in Japan and surveyed the
facilities’ policies on allocating pregnant staff to perform MRI scan duties. We also conducted an
attitude survey of the respondents in relation to non-ionizing radiation and surveyed the correlation
with the allocation policy.
Methods
We mailed questionnaires addressed to the personnel responsible for MRI scan duties to Japan’s
5,769 facilities equipped with an MRI system in early November 2017 and set a response deadline
of threeweeks later. The questionnaire was comprised of: 1) Basic information, 2) General questions
about the employees’ pregnancies (e.g. whether the hospital has a policy), 3) The policy on MRI
scan duties while pregnant and alternative duties, and 4) Level of awareness of non-ionizing
radiation in general. IBM SPSS Statistics 19 was used in the statistical analysis. Chi square test and
residual analysis were used to investigate connections between selected policies on working in MRI
scans while pregnant and attitude survey of non-ionizing radiation. Statistical significance was set at
p<0.05.
Results
The results of the 5,769questionnaires dispatched were that 14 were undelivered, 8 of those 14
were redelivered, responses were received from 2,103facilities ultimately, and the response rate
was 36.5% (out of 5,763facilities). In terms of the characteristics of the respondents, 85.6% of
respondents were male, 62.4% were aged in their 40s–50s, and around half (51.0%) held the
decision-making authority to plan the allocation of duties. As for the characteristics of the facilities,
one facility possessed 7.4imaging modalities on average, and the proportion of female technicians
in the sector averaged 27.0%. In terms of the sector’s employee sufficiency level (in terms of the
number of workers at MRI sector), around half responded (51.0%) that it was sufficient, while
around 35.9% responded that there was some insufficiency. The proportion of female MRI
technologists was 26.6% on average.
As for the general questions about the employees’ pregnancies, of those employees that indicated

80
they were pregnant, 67.7% reported that they were in the first trimester. In terms of individual
protection measures taken for pregnant employees in the sector, at 75.5% of facilities there were no
individual measures, at 19.5% there were individual measures, but they were unstipulated, at 0.4%
individual measures were stipulated but not publicized, and at 1.7% individual measures were both
stipulated and common knowledge.
The policies on allocating pregnant staff for MRI site (including scheduled policies) are shown in
Fig.1. Answers that “increase the opportunity of allocation to MRI scan duties compared with the
current situation” accounted for 7.5%, “maintaining the current frequency of allocation to MRI scan
duties” for 32.3%, and “reduce the opportunity of allocation to MRI scan duties compared with
current situation” for 51.8%. Similarly, around half of facilities had investigated altering the content of
MRI scan duties during pregnancy and the details of work options is shown in Fig.2. In terms of the
rationale for alteration of duties, in response to a question on the connection between the concerns
of non-ionizing radiation emitted by MRIsystem and reproduction/development, in over half the
responses the reason given was that the measure was “just a precaution.” Likewise, in response to
a question about physical strain, while 38.4% of responses considered there to be little physical
strain imposed by MRI scan duties during pregnancy (or that the strain can be reduced), 25.8%
perceived the physical strain to be high (and cannot be reduced). Similarly, in nearly 40% there was
no recognition of the risk of falling due to moving rapidly around the MRI device. The responses
mentioning general views of alternative duties noted computed tomography (CT, 41.1%), followed by
MRI scans (38.5%), and reception work (30.7%) in that order.
In an attitude survey of non-ionizing radiation emitted by medical devices, 22.1% responded that
they had attended lectures and independently researched (named as “strong safety attitude”),
65.9% responded that they were concerned but had neither performed independent research nor
attended a lecture (named as “middle safety attitude”), and 7.7% responded that they were
unconcerned (named as “unconcerned”). In relation to the correlation between the level of
awareness of non-ionizing radiation emitted by medical devices and the policy on working in MRI
scan duties while pregnant, a significant effect for awareness was observed (Chi square test,
p<0.001, Fig.3). The “middle safety attitude” group had opted to significantly reduce the opportunity
of allocation (residual analysis p<0.01), and the group that were unconcerned, rather than opting to
reduce allocation to MRI duties, had opted either to maintain the same frequency of allocation as
previous (residual analysis p<0.01) or opted for proactive allocation (residual analysis p<0.01).
There was no significant difference in residual analysis among “strong safety attitude” population.
Discussion and Conclusion
To date, while there have been no reports of obvious adverse events related to the static magnetic
field emitted by MRI system, no uniform guidelines or safety information have been put forth in
Japan. Therefore, in this study we considered and analyzed the possibility that the level of
awareness of non-ionizing radiation held by those in charge are a factor in judgments and decisions.
The result was that while the selection policy for MRI duties changed depending on the concern and
action taken in relation to non-ionizing radiation (Fig.2), factors at the facilities, such as the
respondent attributes, male/female ratio, and number of imaging modalities also affect selection
policies, and in future we plan to conduct a multivariable analysis. Likewise, even among the
responses of increased or maintained the opportunity of allocation to MRI scan duties, this was
often accompanied by working restrictions such as prohibition on entering the MRI room, and in
future we intend to analyze this breakdown separately. Similarly, it is also known that there are
certain positions within a high magnetic field at which the scan technician’s bodily motion will induce
transient changes in their physical condition, such as dizziness and headaches [4]. Awareness of
the risk of falling and working environment is insufficient, and there is a need to promote recognition
particularly at facilities operating 3T MRI systems.
In this study we surveyed policies on allocating MRI scan duties to pregnant employees and
investigated the influence of respondents’ level of awareness of non-ionizing radiation. The results
revealed inconsistent handling within Japan, and that although duties may be restricted in some
situations, 7.8% of facilities increase the opportunity of allocation to MRI duties proactively
81
compared with the current situation while 51.8% of facilities opt to reduce the allocation to MRI
duties. And it was also observed that the level of awareness of non-ionizing radiation influences the
policy on the allocation of MRI scan duties for pregnant employees.
References
[1] D.H.Temperton, "Pregnancy and Work in Diagnostic Imaging Departments 2nd Edition", British
Institute of Radiology (2009)
[2] Expert Panel on MR Safety et al., "ACR guidance document on MR safe practices: 2013" J Magn
Reson Imaging. 2013 Mar;37(3):501-30.
[3] De Wilde JP, Rivers AW, Price DL."A review of the current use of magnetic resonance imaging in
pregnancy and safety implications for the fetus." Prog Biophys Mol Biol. 2005 Feb-
Apr;87(2-3):335-53.
[4] ICNIRP, "ICNIRP Guidelines for limiting exposure to electric fields induced by movement of the
human body in astatic magnetic field and by time-varying magnetic fields below 1 Hz.", Health
Pysics 2014 106 (3): 418-425.
Figures

Figure 1. Facility policies on working in MRI scan duties during pregnant (including scheduled
policies)

82
Figure 2. The details of work options working in MRI scans while pregnancy in each selected
policy at Table 1.

Figure 3. The relationship between the selected policies and level of awareness of non-
ionizing radiation.

83
S04-3 [17:00]
Risk assessment for workers with pacemakers working in the near-field of a high-frequency
EMF source
Carsten Alteköster1 & Marc Wittlich1
1Unit 4.4: Radiation, Institute for Occupational Safety and Health (IFA), Sankt Augustin, Germany, 53757
Keywords: Occupational, RF/Microwaves, Completed (unpublished)
Presented by: Carsten Alteköster
An important task of the Institute of Occupational Safety and Health (IFA) is to carry out risk
assessments at workplaces in order to protect workers with active implantable medical devices
exposed to electromagnetic fields. For this purpose the induced voltages at the input stage of these
devices must be determined. A problem in the high-frequency range arises from the fact that the
formulae used to calculate these voltages are derived under far-field conditions. However, this is
often not the case. Therefore, we investigated the induced voltage level during a near-field exposure
in a high-frequency electromagnetic field.The results suggest that the measuring of the power
density could be sufficient in the scope of a risk assessment.
Introduction
The EMF Directive 2013/35/EU was implemented into German national legislation by the
Occupational Safety and Health Ordinance on Electromagnetic Fields (EMF Ordinance). The
purpose of the EMF Ordinance is to protect workers exposed to electromagnetic fields. In the
Directive, the term “workers at particular risk” refers also to workers with implantable cardiac
pacemakers (PM) or defibrillators (ICD). For this group an appropriate risk assessment and
protective measures are required. Since no limit values for time-varying electromagnetic fields are
given to fulfill this requirement other sources of information are usually used [1, 2, 3, 4]. Hereby it is
made possible to assess the voltage at the input stage of a PM or ICD induced by external
electromagnetic fields. This voltage is then compared to the required values given in [5, 6] to avoid
interference. However, the provided formulae for the high-frequency range are commonly based on
far-field conditions and, hence, it is supposed, that the worker stands in a distance more than one
wavelength λ to the EMF source. If you consider for example a high frequency welding unit with a
working frequency of 27.12 MHz (λ ≈ 11 m), it is obvious that the worker stays within the near-field.
Another problem in the near-field is that the well-defined relationship between the magnetic and
electric field, as it can be found in the far-field, no longer exists. This leads to the fact that both
quantities must be measured and the effort increases finally.
Therefore, the aim of our work was to study whether the far-field formulae lead to advisable results
under near-field conditions and on the other hand, whether we could find another appropriate way to
accomplish a near-field risk assessment. In this context, we focused on the power density in the
near-field which could be measured with a diode or thermal power sensor.
Materials and Methods
In order to investigate the induced voltages at the input stage of an active implantable medical
device we conducted electromagnetic field simulations using the finite-difference time-domain
(FDTD) simulation platform Sim4Life V 3.2 [7].
For the simulation model we selected the Virtual Population phantom DUKE V3.0 (male, 34y, 1.77m
and 72.4 kg) [8]. A simple CAD model of a unipolar pacemaker was inserted in the chest of the
phantom. A lead was fit into the heart and connected to the pacemaker model. An edge sensor was
placed at the point of connection (input stage) to record the voltage at this point. The lead consisted
of a 410 mm long insulated wire with a total diameter of 5 mm, while the inner conductor wire had a
diameter of 1.5 mm. To simplify matters, we used either a linear dipole or a ring dipole as the source
of the electromagnetic field and positioned it in front of the left chest area where the pacemaker is
located in the phantom.
The simulation frequencies were chosen from the industrial, scientific and medical (ISM) radio

84
bands in consideration of common applications. Therefore, we did calculations with 27.12 MHz (e.g.
dielectric or induction heating), 433 MHz (e.g. diathermy) and 868 MHz (e.g. RFID applications).
The dielectric tissue properties at these frequencies were assigned to the DUKE phantom according
to the material database found in [9].
For every combination of dipole type and frequency we evaluated the induced voltage at the input
stage for different distances of the phantom to the dipole source. The distances were chosen
relative to the wavelength λ as 0.1 λ, 0.2 λ,…, 1 λ, 3 λ to cover near-field conditions as well as the
transition to far-field conditions.
Additionally, we ran the same simulations, but without the phantom to represent the situation like it is
usually during a common EMF measurement. Based on this we calculated the mean power density
as well as the electric and magnetic field strength in the area covered by the pacemaker and the
lead in the corresponding simulations with the phantom present. This was named “effective area”.
To reduce the computational load for the simulations with larger distances between source and
phantom we made use of the Huygens source method, which is provided by Sim4Life.
Finally, for every frequency four datasets were derived from the described simulations (induced
voltage vs. distance; mean power density vs. distance; mean electric/magnetic field strength vs.
distance).
To assess how well the provided formulae found in [2], which predominantly agree with those in [3]
and [4], described the simulated situations, the induced voltages at the input stage were calculated
from the simulated field strengths and compared to the simulated induced voltages.
Additionally, we investigated whether the mean power density could be a reasonable parameter to
estimate the induced voltages. Therefore, we performed nonlinear least squares analysis on the
datasets to construct mathematical functions, which combines the mean power density with the
induced voltage. For this we used the Curve Fitting Toolbox of MATLAB [10].
Results
The comparison of the calculated induced voltages and the simulated induced voltages led to
alternating results. It was found that for some combinations of dipole type and frequency the
simulated values are consistently lower than the calculated values. This is especially valid for all
simulations at 27.12 MHz. For the other frequencies the results are not that consistent and the sign
of the difference between simulated voltages and calculated voltages is negative or changing.
Two functions resulted from the performed curve fitting analysis. Function 1 describes the
relationship between the mean power density S outside the phantom body, the wavelength λ and
the distance d between dipole and phantom. Function 2 gives the link between the induced voltage
at the input stage (peak-to-peak value) UPP and the mean power density S outside the phantom
body.
Both functions contain unknown parameters, which were determined by nonlinear least squares
analysis. Thus it was revealed that the parameters are depending on the kind of dipole source and
frequency and therefore both functions look different regarding these parameters for each case.
Making the assumption, that in the “effective area” the action values of the mean power density
according to the table A3.4 of the German EMF Ordnance are met, we calculate UPP. It is
remarkable that in every case the resulting UPP is significantly lower than 10 V, which is the required
value given in [5, 6] to avoid pacemakers or implantable cardioverter defibrillators to suffer
interference in the considered frequency range here.
Discussion and Conclusions
Workers with an active implantable medical device who work in the near-field area of an
electromagnetic field source need an adequate risk assessment for their workplace.
This study demonstrates that this is possible only by measuring the mean power density instead of
measuring the electric and magnetic field strength. Furthermore, no calculations are needed, but it is

85
sufficient to show that the relevant action values given by the German EMF Ordinance are met. This
finding is especially helpful if one considers that the comparison of the simulated induced voltages
with the calculated induced voltages indicate no clear trend.
The performance of further simulations, which take into account other field sources, frequencies or
implant types, will be necessary to confirm our findings. Furthermore, it has to be investigated which
experimental methods will be suitable for verification of the simulation results.
Disclaimer
The mention of commercial products, their sources, or their use in connection with material reported
herein is not to be construed as either an actual or implied endorsement of such products by the
Institute for Occupational Safety and Health of the German Social Accident Insurance (IFA).
References
[1] German Social Accident Insurance (DGUV). DGUV Information 203-043 (BGI 5111).
Beeinflussung von Implantaten durch elektromagnetische Felder: Eine Handlungshilfe für die
betriebliche Praxis (in German). March 2012.
[2] German Federal Ministry of Labour and Social Affairs. Forschungsbericht 451:
Elektromagnetische Felder am Arbeitsplatz Sicherheit von Beschäftigten mit aktiven und passiven
Körperhilfsmitteln bei Exposition gegenüber elektromagnetischen Feldern (in German). January
2015
[3] EN 50527-2-1:2016. Procedure for the assessment of the exposure to electromagnetic fields of
workers bearing active implantable medical devices - Part 2-1: Specific assessment for workers with
cardiac pacemakers.
[4] prEN 50527-2-2:2017. Procedure for the assessment of the exposure to electromagnetic fields of
workers bearing active implantable medical devices - Part 2-2: Specific assessment for workers with
cardioverter defibrillators (ICDs).
[5] EN 45502-2-1:2003. Active implantable medical devices - Part 2-1: Particular requirements for
active implantable medical devices intended to treat bradyarrhythmia (cardiac pacemakers).
[6] EN 45502-2-2:2008. Active implantable medical devices - Part 2-2: Particular requirements for
active implantable medical devices intended to treat tachyarrhythmia (includes implantable
defibrillators).
[7] ZMT Sim4Life V3.2, ZMT Zürich MedTech AG, 8004 Zurich, Switzerland
[8] Gosselin M C, Neufeld E, Moser H, Huber E, Farcito S, Gerber L, Jedensjö M, Hilber I, Di
Gennaro F, Lloyd B, Cherubini E, Szczerba D, Kainz W, Kuster N, Development of a new generation
of high-resolution anatomical models for medical device evaluation: the Virtual Population 3.0,
Physics in Medicine and Biology, 59(18):5287-5303, 2014.
[9] Hasgall PA, Di Gennaro F, Baumgartner C, Neufeld E, Gosselin M C, Payne D, Klingenböck A,
Kuster N, “IT’IS Database for thermal and electromagnetic parameters of biological tissues,” Version
3.0, September 01st, 2015, www.itis.ethz.ch/database
[10] MATLAB R2017a. The Mathworks, Inc., Natick, Massachusetts (USA)

S04-4 [17:15]
Novel mechanistic model for implant safety assessment at WPT and MRI frequencies
Ilaria Liorni1, Esra Neufeld1, Sven Kuehn1 & Niels Kuster1, 2
1Foundation for Research on Information Technologies in Society (IT'IS Foundation), Zuerich, Switzerland,
8004
2Swiss Federal Institute of Technology (ETHZ), Zuerich, Switzerland, 8092
Keywords: Mechanistic/Theoretical, All Frequencies, Completed (unpublished)
Presented by: Ilaria Liorni

86
Current safety guidelines for exposure to electromagnetic fields (EMF) exclude medical implant
wearers, and there is further no established method for assessing risks posed by EMF in this group.
We propose a new mechanistic model for determining local exposure in the presence of implants
based on incident field values valid over a wide frequency range. To validate the approach,
predictions of the mechanistic model were compared to results of numerical simulations. The effects
of parameters such as tissue properties, tip shape, and insulation thickness were studied. The
findings support the formulation of safety standards for implant wearers exposed to EMF over a
wide range of frequencies, including those used in wireless power transfer.
Introduction
Current safety guidelines [1- 3] for exposure to electromagnetic fields (EMF) exclude persons with
medical implants. At the same time, there is no established method to determine the associated risk
for this population group. Standards previously established for evaluating implants under magnetic
resonance imaging (MRI) frequencies, e.g., ISO/TS 10974 [4] and ASTM F2182-11a [5], are based
on models describing the coupling of implants with typical MRI frequencies, i.e., 64 and 128MHz
[6-8].
The scientific literature about implant-related power deposition and resulting tissue heating or
neuromodulation at low frequencies is, however, sparse. For example, strong magnetic (H-) field
exposures are expected from wireless power transfer (WPT) technologies operating in the 3kHz –
10MHz range. Furthermore, the safety of passive implants, e.g., rods, clips, and stents, at MRI
frequencies has not yet been systematically assessed. Indeed, each of these passive implants must
be tested individually by means of a standard method such as ASTM F2182-11a [5].
The mechanistic view applicable in MRI safety assessment for elongated implants is based on pick-
up by the lead of energy deposited at critical implant locations according to measurable transfer
functions. At lower frequencies, the observed dependence on tissue conductivity and tip insulation is
inconsistent with this mechanistic view.
In this study, we have identified a suitable mechanistic theory valid for electrically small (l < λ/2,
where λ is the wavelength in the specific tissue) and partially insulated implants at WPT frequencies.
In addition, the applicability of such a theory to higher frequencies of up to 128MHz has been
investigated as well. The model is based on the hypothesis that the current J through an implant
short-circuit is driven by the potential difference for exposure to electric (E-) fields and by the
electromotive force for exposure to H-fields and is constrained by tissue resistance at the lead tip,
which is modeled as serial resistances dependent on tip shape and local tissue conductivity σ. The
levels of induced E-field and the absorbed power, expressed as specific absorption rate (SAR), at
the lead tip can be derived from J and information about tip shape. Thus, local exposure quantities
can be computed based on a single simulation of the lead tip in a homogeneous environment in
combination with both the information on the integrated tangential E-field along the implant path and
tissue properties at the tips and exposed locations. This approach permits precomputed in vivo
exposure conditions to be reused independently of precise details about the implant.
Numerical simulations in homogeneous environments were performed to validate the mechanistic
model and to study the impact of various factors. Further, to confirm the validity of the model in
realistic inhomogeneous environments, additional simulations involving detailed anatomical models
were performed. From these findings, conclusions and approaches that are valuable for
standardization purposes have been derived.
Material and Methods
Validation of the mechanism was performed for exposure to E- and H-fields within i) homogeneous
setups, ii) semi-homogeneous setups with local inhomogeneity at the lead tip, and iii) complex
human anatomical models.
All simulations were carried out with the simulation platform Sim4Life Version 3.0 (ZMT, Zurich,
Switzerland).
Exposures to E-fields generated by straight wires – of lengths ranging from 10 – 80cm with 0.5mm

87
insulation and 10mm long bare metallic tips – embedded in homogeneous muscle tissue were
studied. Uniform exposure to an E-field tangential to the wire over the frequency range 3kHz –
128MHz was simulated. The influence of different parameters on deposited power, induced E-field,
and current density J in tissue surrounding the lead tip was studied: i) the effect of environment was
assessed by replacing muscle tissue with fat or grey matter (which have different dielectric and
thermal properties) globally or in the vicinity of one or both tips; ii) the influence of the insulation
thickness (0 – 1mm) was checked at 10, 30, 64, and 128MHz for all lead lengths; iii) an analysis of
sensitivity with respect to the tip shape was performed as well at the same frequencies.
For exposure to H-fields, two circular wires embedded in a cylindrical structure filled with muscle
tissue were exposed to a coil-generated H-field. Wires were inserted along trajectories with constant
tangential E-fields, with one wire covering ¼ and the other ¾ of the circle.
For setups representing exposure to both E- and H-fields, the peak-spatial SAR (pSAR) averaged
over 0.1g, 1g, and 10g tissue mass, as well the maximum induced E-field, were extracted at the
lead tip. Further, average quantities (i.e., the average E-field vector induced in a 2mm × 2mm ×
2mm cube, according to ICNIRP 2010 [2]), the induced E-field vector averaged over a 5mm straight
line according to IEEE C95.1 2005 [3], and the induced current density J averaged over the area of
a 1cm2 disc according to ICNIRP 1998 [1] were analyzed. Finally, thermal simulations were
performed for each case over the whole frequency range to estimate local temperature increases at
the tip.
Predictions of local exposures in human anatomical models, based on the mechanistic model, were
made by combining the tip-environment interaction characterizations obtained from single
homogeneous environment simulations with knowledge about in vivo incident field conditions. These
predictions were then used to validate the mechanistic model and the separation approach by
comparing the predictions with results obtained by simulating combined setups featuring the
detailed anatomical model Duke of the Virtual Population (ViP) [9] with integrated implants under
conditions of exposure to E- and coil-generated H- fields at 1MHz.
Results
For frequencies ≤10MHz, agreement within the estimated uncertainty was found for all the quantities
assessed (i.e., induced E-field, J, and SAR) with differences between numerical simulations and
mechanistic model predictions consistently being <5% (Fig. 1). After normalization of the assessed
quantities with respect to the tissue conductivity σ, the tangential E-field, and the lead length
according to the dependence predicted by the mechanistic model, negligible deviations were found.
The frequency-related variability is <2% and tissue dielectric properties and lead-length-related
variabilities <5%, in support of the proposed mechanistic model, whereby the voltage-driven current
through an implant short circuit is limited by tissue resistance where the current is squeezed through
exposed tips. The normalized temperature increase, which is almost constant (variation <1%)
across the frequency range in muscle tissue, shows the expected dependence on thermal
properties when muscle is replaced with other tissues.
For frequencies >10MHz, the mechanistic model predicts pSAR and parameter dependence with
<10% variability. With increasing frequency, the impact of lead length and frequency on normalized
temperature increase and average SAR quantities is more relevant. However, when normalized with
respect to electrical length, a universally scalable behavior relationship across frequencies could be
extracted (Fig. 2).
The impact of insulation is more complex, as capacitive coupling through the insulation becomes
increasingly relevant at higher frequencies. The result is a situation that approximates a bare
implant with a larger surface area through which current enters, thus decreasing resistance (Fig. 3).
The tip shape affects highly localized, induced-field-related quantities and peak-averaged SAR by
>20%, while the variability in temperature increase is typically <3% for short implants at <10MHz,
and up to 13% at higher frequencies.
For exposure to H-fields, similar results were obtained, with 2% deviation in normalized pSAR and

88
induced E-field.
The validation in the human anatomical model with a generic implant exposed at 1MHz via
comparison of model predictions with simulated complete setups shows a difference in pSAR of 5%
for exposure to H-fields and 15% for exposure to E-fields.
Discussion And Conclusions
Exposure of persons with medical implants to EMF is a topic of increasing interest, since current
guidelines do not provide suitable limits and no method has so far been developed to estimate the
associated risk for this population group.
In this study, we developed a quantitative, generalized EM safety model for electrically small
implants applicable to frequencies used with both WPT and MRI. For insulated or partially insulated
leads shorter than λ/2 in tissue, the amount of power deposited at the implant tip depends on the
voltage difference or electromotive force between the two tips, is a simple function of the local tissue
resistivity at the tips, and is only weakly dependent on tip shape. A universal relationship describing
deposited power as a function of the electrical length obtained for partially insulated implants
demonstrates the applicability of the mechanistic model to a large frequency range.
The results can provide valuable support for the formulation of a general approach applicable to a
large frequency range for use in safety guidelines and standards. The tip-related resistance and field
distribution was characterized by means of a single simulation of homogeneous exposure within
homogeneous tissue, and these parameters are combined with precomputed in vivo incident field
conditions and information about the tissues surrounding critical implant locations according to the
mechanistic model. In this way, local exposure metrics can be predicted and worst-case conditions
can be identified. As the next step, extended validation and a comprehensive uncertainty
assessment will be performed.
References
[1] ICNIRP 1998, Guidelines for limiting exposure to time-varying electric, magnetic and
electromagnetic fields (up to 300 GHz), Health Phys., 74(4): 494–522, 1998.
[2] ICNIRP 2010, Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz
to 100 kHz), Health Phys., 99(6): 818–836, 2010.
[3] IEEE C95.1 2005, IEEE standard for safety levels with respect to human exposure to radio
frequency electromagnetic fields, 3 kHz to 300 GHz, Institute of Electrical and Electronics
Engineers, Inc. 3 Park Avenue, New York, New York 10016-5997, USA, 2006.
[4] ISO-TS 10974:2012(E), Assessment of the safety of magnetic resonance imaging for patients
with an active implantable medical device, 2012.
[5] ASTM, F2182-11a. Standard test method for measurement of radio frequency induced heating
near passive implants during magnetic resonance imaging, pp. 1–8, 2011.
[6] J.A. Nyenhuis, S.M. Park, R. Kamondetdacha, A. Amjad, F.G. Shellock, and A.R. Rezai, MRI and
implanted medical devices: basic interactions with an emphasis on heating, IEEE Trans. Device
Mater. Rel., 5(3): 467–480, 2005.
[7] S.M. Park, R. Kamondetdacha, A. Amjad, and J.A. Nyenhuis, MRI safety: RF-induced heating
near straight wires, IEEE Trans. Magn., 41(10): 4197–4199, 2005.
[8] C. J. Yeung, R.C. Susil, and E. Atalar, RF safety of wires in interventional MRI: Using a safety
index, Magn. Reson. Med., 47: 187–193, 2002.
[9] M.C. Gosselin, E. Neufeld, H. Moser, E. Huber, S. Farcito, L. Gerber, M. Jedensjö, I. Hilber, F. Di
Gennaro, B. Lloyd, E. Cherubini, D. Szczerba, W. Kainz, and N. Kuster, Development of a new
generation of high-resolution anatomical models for medical device evaluation: The Virtual
Population 3.0, Phys. Med. Biol., 59(18), 5287–5303, 2014.
[10] R.W.P King, K.M. Lee, S.R. Mishra, and G.S. Smith, Insulated linear antenna: Theory and
experiment, J. Appl. Phys., 45(4), 1688–1697, 1974.

89
Figures

Figure 1. Volume SAR averaged over 0.1g across the frequency range 150kHz – 10MHz
estimated through numerical simulations for exposure to E-fields, normalized to square of the
incident E-field (solid line), and predicted by the mechanistic model (dashed line).

Figure 2. Temperature increase (dT) normalized according to the mechanistic model to lead
length, muscle tissue conductivity at each specific frequency, and tangential E-field as a
function of the electrical length of partially insulated wires, estimated as ratio of the lead length
to the wavelength in medium calculated with King’s formula [10], which accounts for the
presence of an insulated antenna in the medium. Universal relationships that are independent
of frequency are obtained.

90
 Figure 3. pSAR 0.1g – normalized to lead length, muscle tissue conductivity, and tangential
E-field – as a function of frequency for insulation thicknesses of 0, 0.1, 0.5, and 1mm; the
impact of frequency on insulation effectivity through capacitive coupling is apparent.

S04-5 [17:30]
STUDENT PAPER
Integration of big-data analytics in safety assessment of patients with medical implants
during MRI exposure
Aiping Yao1, 2, Earl Zastrow1, Pedro Crespo-Valero3 & Niels Kuster1, 2
1IT'IS Foundation, Zurich, Switzerland, CH-8004
2Department of Information Technology and Electrical Engineering, Zurich, Switzerland, CH-8092
3Schmid & Partner Engineering AG, Zurich, Switzerland, CH-8004
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Aiping Yao
The RF-induced heating of the patient with implant under MR exposure is a complex function of
multi-factors, e.g., implant characteristic, patient anatomy, imaging position, RF coil, etc. A
comprehensive safety assessment cannot be derived from limited clinical scenarios and in silico
trials are usually required to assist the evaluation process. To address further needs for in silico
trials, we have established a safety assessment workflow comprises a data library and toolset to
perform a comprehensive evaluation in a timely and traceable manner. We demonstrate the
proposed workflow through an evaluation of RF-induced heating of a spinal cord stimulator. More
than 39 million unique clinical scenarios were emulated in silico.
Introduction
The RF-induced effects on implantable medical devices and the subsequent health risk poses for
the patient during MRI depends on the implant characteristics (RF and otherwise) and the RF
exposure to the patient. At minimum, a comprehensive safety evaluation must consider significant
cases of relevant RF exposures and a concrete implant RF-characterization method. Therefore,

91
clinical trials of limited scenarios are likely insufficient to ensure safety. Due to these practical
constraints (average time and cost to complete a significant clinical trial), assessment of patients
risk from RF-implant interactions during MRI is established through a synergy of experimental
testing (in phantoms and/or animals) and in silico analysis [1]. It is essential that a standardized
framework that can provide, not only relevant, but also comprehensive, consistent, extendable, and
traceable results, be established --- such that a reliable knowledge base can be built upon the
corroboration of existing and emerging results. With today’s available digital resources (computing
clusters, storage, and software), in silico trials can now be realized in a systematic and
comprehensive manner.

We have constructed a library of RF-induced fields inside the human body, resulting from exposure
to RF coils used in close-bore MRI scanners. The library comprises the exposure of a population of
virtual patients (different patient anatomy) of the Virtual Population (ViP) [2] to a set of generic RF
cylindrical body-coils (different coil dimensions) at different clinical imaging positions. It can be
readily extended to include additional patients as new computable human phantoms become
available. As patient safety is derived from these components, it is essential that the needed digital
components are, not only relevant, but also consistent, extendable, and traceable. We envision that
for each context of use, a pre-selected dataset from the library can be assembled and pre-verified
via a cross-validation process conducted by different users; the assembled dataset can then be
used towards a standardized safety assessment process with a clear digital evidence.

In this work, we establish a standardized workflow, based on Tier 3 evaluation of [1] (see also [3, 4]
for further description), for the safety evaluation of patients with elongated active medical implants
(e.g. pacemakers and neuro-stimulators). The proposed workflow is applied to quantify the RF-
induced heating of a spinal cord stimulator at 128 MHz (3T MRI).
Method
The proposed workflow (see Figure 1) is comprised of the following components:
1. RF-exposure library: This component is the pre-computed RF-induced fields during RF of MRI
exposures inside multiple patients. It is derived from the human phantoms of the ViP and
computational models of RF birdcage coils. It can be standardized for the safety evaluation of
any implants with a clear and traceable digital evidence.
2. Implant-specific objects:Clinical routings of the implant under investigation: This digital object
defines the clinical routings of the implant within each human phantom of the ViP. It can be
standardized for the safety evaluation of any implant that follows the same clinical routings with
a clear and traceable digital identification.Tier 3 model of RF-induced heating of the implant
under investigation: This digital object defines the transfer function of RF-induced characteristics
(heating or terminal voltage, see Clause 10 and 19 of [1] for further description). It is associated
with the implant under investigation and can be assigned a clear and traceable digital
identification.
3. Data analysis module: This component is implemented and released as a software module with
a clear and traceable digital evidence.
4. Output data/results: The results from the evaluation, using the components 1 – 3, with a clear
and traceable digital evidence.

The components of the workflow are designed such that each component can be revised and
evolved independently.

Two virtual patients --- Duke and Fats of the VIP --- were selected for the RF-induced heating
evaluation of a generic spinal cord stimulator. Two clinical routing groups of the spinal cord
stimulator were then defined for the selected patients. The routing groups were defined to represent
the left- and right-routings of the spinal cord stimulator lead (SCSL and SCSR), with each group

92
containing 100 individual routings.

A commercial 90-cm spinal cord stimulator lead was used as the implant under investigation. The
Tier 3 model of RF-induced heating of the implant was derived experimentally using the piece-wise
excitation method [4].
Results
The in vivo RF-induced heating of the spinal cord stimulator was estimated for the selected patients.
Figure 2 shows the distribution of the estimated in vivo power deposition at the electrode of the
spinal cord stimulator in Duke and Fats at 17 imaging positions from head to foot at every 10 cm
increment. The dashed line indicates the envelope of the maximum power deposition under the 65
imaging positions (from head to foot at every 2.5 cm). Each histogram contains the RF-induced
heating of the spinal cord stimulator from 360,000 clinical scenarios (360 B1 excitations x 10 RF-
coils x 100 routings). A total of 39 million clinical scenarios were considered in the RF-induced
heating evaluation. The maximum heating was found in Duke at imaging position 6 during exposure
to the 60-cm-long RF-coil of the 60-cm-bore MRI. This clinical scenario was selected for further
evaluation. Figure 3 illustrates the clinical scenario associated with the maximum estimatedin vivo
RF-induced heating. The B1 fields information was demonstrated for the clockwise and anti-
clockwise quadrature excitations.
Conclusions
With the growing computing power and advancement in the modeling techniques to capture the
interactions of medical implants with external inputs within the human body, it is conceivable that
clinical trials can be replaced or reduced by the use of in silico trials to eliminate or reduce the risk
imposed on patients during clinical assessment. Through the establishment of a common data
library and analysis tools, sufficiently large clinical scenarios (permutation of patients, implant clinical
routings, imaging positions, and MRI coils) can be considered in the implant RF-safety evaluation
procedure with a clear and traceable digital chain of custody. We demonstrated that evaluation of
large clinical scenarios (39 million) is highly feasible.
References
[1] ISO/TS. 10974:2012, Requirements for the safety of magnetic resonance imaging for patients
with an active implantable medical device. ISO/TS 10974 2012.
[2] M. C. Gosselin, E. Neufeld, H. Moser, E. Huber, S. Farcito, L. Gerber, M. Jedensjo, I. Hilber, F. D.
Gennaro, B. Lloyd, E. Cherubini, D. Szczerba, W. Kainz and N. Kuster. Development of a new
generation of high-resolution anatomical models for medical device evaluation: the Virtual
Population 3.0. Phys. Med. Biol. 59(18): 5287–5303, 2014.
[3] S. M. Park, R. Kamondetdacha, J. A. Nyenhuis. Calculation of MRI-Induced Heating of an
Implanted Medical Lead Wire with an Electric Field Transfer Function. J. Mag. Res. Imag.
26:1278–1285, 2007.
[4] E. Zastrow, M. H. Capstick, and N. Kuster.Experimental System for RF-Heating Characterization
of Medical Implants during MRI. ISMRM, 2016.
1278–1285
Figures

93
Figure 1. The proposed standardized workflow based on Tier 3 evaluation of RF-induced
effects of medical implants [1].

Figure 2. Histograms of estimated in vivo power deposition of a 90-cm spinal cord stimulator of
39 million clinical scenarios. The distributions at every 10-cm imaging position from head to
foot are shown. The envelope, indicated by dashed line, is obtained from the maximum
estimated power deposition extracted from the distributions at every 2.5-cm imaging position
from head to foot.

94
 Figure 3. (Left) Clinical scenario with maximum estimatedin vivo RF-induced heating, found in
Duke exposed to a 60-cm bore 3T MRI at imaging position 6. (Top-right) B1 distributions at iso-
centre plane under anti-clockwise quadrature excitation; (Bottom-right) B1 distributions at iso-
centre plane under clockwise quadrature excitation.

S04-6 [17:45]
Novel MRI RF safety assessment of orthopedic implants via bone surface tangential E-fields
Manuel Murbach1, Earl Zastrow1, Esra Neufeld1, Bryn Lloyd1, Wolfgang Kainz3 & Niels Kuster1, 2
1IT'IS Foundation, Zurich, Switzerland, 8004
2ETH Zurich, Zurich, Switzerland, 8092
3US Food and Drug Administration (FDA), Silver Spring, Maryland, USA, 20993
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Manuel Murbach
Orthopedic implant manufacturers produce a large portfolio of on-bone devices. Current implant
radiofrequency (RF) safety standards (e.g., ASTM F2182-11a, TS/ISO 10974), however, do not take
advantage of very well-defined data on relevant RF exposures of these devices. In this study, we
used the bone-surface-averaged tangential E-field as a close approximation of the incident field
impacting an on-bone orthopedic device. Results show that the relevant surface-averaged tangential
E-fields are less than half of the corresponding peak volume-averaged E-fields, which allows
conservative but not overly conservative assessment of RF implant safety in magnetic resonance
imaging.
Introduction
Orthopedic implant manufacturers produce a large portfolio of on-bone devices – such as plates and

95
meshes – of tremendously diverse geometries, sizes, and device combinations. Specific application
of a device on a given bone structure generates a very well-defined relevant incident electric (E-)
field tangential (Etan) to the respective bone-surface. However, the Tier2 method of the technical
specification ISO/TS 109741 does not take advantage of Etan data, but accepts the volumetric 10g
averaged maximum E-field at the implant location, which, as a result, leads to an overestimation of
the maximum E-field at the location of the implant. When only the surface-averaged Etan is
considered, overestimation is eliminated, which allows for robust and conservative safety
assessments of this implant category. A comparison between the volumetric and tangential E-fields
is shown in Figure 1. The ASTM F-21822 test method is testing passive implants in homogenous
tissues only.
In this study, we evaluated a cranial bone plate and mesh, with the respective incident E-fields
evaluated at the skull bone. Our “modified Tier2” approach included three steps: (1) assessing the
maximum induced in silico E-field on the surface of the patient’s bone; (2) using the in silico Etan to
calculate the resulting additional power deposition and temperature increase caused by the implant
in vitro; and (3) validating the conservativeness of this approach with full-wave simulations for
selected scenarios. The concept is currently being extended to other body regions and implant
categories.
Methods
The bone-surface E-fields (Step 1) were evaluated for the anatomical models DUKE and FATS of
the Virtual Population3 in the head imaging position for a generic 1.5T circularly polarized birdcage
body-coil, normalized to the IEC limit4 of 3.2W/kg head-averaged specific absorption rate (SAR).
Evaluations are currently being extended to the whole body, and with more human models and
various radiofrequency (RF) birdcage geometries. The tangential fields are determined with
Sim4Life (ZMT, Zurich, Switzerland), and averaged over a 460mm2 area (Etan460mm2), which
corresponds to the two-dimensional cross-section of a 10 g-cube of muscle tissue.
The implant-enhanced in vitro fields (Step 2) were computed by exposing the implants in a simplified
planar tissue model to the E-field estimated in Step1. We investigated a generic cranial plate
(120mm long, 5mm wide, 1mm thick) and a cranial mesh (82mm diagonal long, 0.1mm thick), both
shown in Figure 2. The lengths of the two implant types were also varied to assess relative impact.
The human skull is interfaced with fat or muscle on the outer surface, therefore the implant is placed
onto a bone surface and embedded in the muscle tissue, which results in a larger amount of power
deposited and higher temperature rise than at the bone-fat interface. The electrothermal simulations
estimate the deposited power (within a 20dB-falloff region) and the peak temperature increase at the
implant tip, with the discretization adjusted to ascertain sufficient convergence.
Results
The maximum E-fields induced in DUKE and FATS in the head imaging position are (values in
parentheses denote the 95th percentile): 348 (147)V/m E10g in the whole head and 226 (183)V/m
E10g in the skull bone; the skull bone surface-averaged tangential fields are 177 (114) V/m
Etan460mm2. Thus, our proposed approach returns 95th percentile incident fields almost 3dB lower
for the implant than obtained with the classic Tier2 approach. As very short implants (< 20mm) are
less of a concern for excessive heating, we applied the 95th percentile Etan460mm2 of 114V/m for
the incident field of the in vitro exposure of Step 2. The estimated steady-state peak temperature
increases are 14.8°C for the cranial plate and 5.0°C for the mesh (Figure 2). These values
conservatively cover all possible implant placements on the skull. Variation of the implant lengths
show the expected behavior when approaching the resonant length in tissue (Figure 3A). The Tier4
validation full-wave simulations are illustrated in Figure 3B. The full-wave simulations for the two
selected exposure scenarios were based on placing the implant in a high-E-field location, and show
temperature increases well below the estimated conservative values of Step 2.

96
Conclusions
Conservative but not overly restrictive magnetic resonance imaging (MRI) RF implant safety
assessment is essential to make MRI scanning available to patients with implants. This presented
novel modified Tier-2 approach has two main advantages: (1) it reduces the overestimation
compared to the original Tier-2 approach while still being conservative and (2) allows opening the
path for device class certifications. This is demonstrated for estimation of thermal load due to
orthopedic on-bone devices. The show-case is currently being extended to the rest of the body and
investigations of the feasibility of surrogate modelling to cover entire on-bone implant groups
throughout the body are ongoing. To further reduce the overestimation of our approach, we
investigate the impact of limiting the surface-averaging area to locations where the implant is
typically located.
Disclaimer
The mention of commercial products, their sources, or their use in connection with material reported
herein is not to be construed as either an actual or suggested endorsement of such products by the
Department of Health and Human Services.
References
1. ISO/TS. 10974:2012, Requirements for the safety of magnetic resonance imaging for patients
with an active implantable medical device. ISO/TS 10974 2012.
2. ASTM. F2182-11a. Standard Test Method for Measurement of Radio Frequency Induced Heating
Near Passive Implants During Magnetic Resonance Imaging. 2011:1–8.
3. Gosselin M-C, Neufeld E, Moser H, et al. Development of a new generation of high-resolution
anatomical models for medical device evaluation: the Virtual Population 3.0. Phys Med Biol
2014;59:5287–5303.
4. IEC. Medical electrical equipment - Part 2-33: Particular requirements for the basic safety and
essential performance of magnetic resonance equipment for medical diagnosis, Edition 3.2. IEC
60601-2-332010+AMD12013+AMD22015 CSV 2015.
Figures

97
Figure 1. Illustration of the difference between the volumetrically averaged E-field (E10g), and
the bone-surface averaged tangential E-field (Etan460mm2) in the FATS anatomical model.
The high E-field regions of the volumetric E-field are often not close to bone structures (*). The
maximum Etan460mm2 of 177 V/m is located at the bridge of the nose (#).

Figure 2. Generic orthopedic cranial plate (A, nominal length 120mm, thickness 1mm) and
mesh (B, diagonal length 82mm, thickness 0.1mm), in a straight configuration. (C) Simplified
planar in vitro exposure configuration. Top: Slice view of the E-field (uniform background E0-
field of 114V/m at implant location). Center: Implant E-field with subtracted background field.
Bottom: Temperature increase distribution. The peak temperature increase reaches 14.8°C for
the 120mm long cranial plate, and 5.0°C for the mesh of 82mm diagonal length. These values
should conservatively cover all possible placements and orientations on the skull.

98
 Figure 3. (A) Peak temperature increase of the generic plate and mesh as a function of the
E-field-aligned implant length. The incident E-field is 114V/m at the location of the implant (on
the bone-muscle interface). Curves are interpolated, and the actual simulation data is shown
as dots. Thin lines are values from 3T (128MHz) exposure for comparison. The resonant
length corresponds to the approximate half-wavelength in muscle. (B) Full-wave simulation
(Tier 4) of FATS with the cranial plate and mesh morphed onto the back of the head. The high
E-field/SAR region can be seen at the lower back of the skull (skull-surface SAR in upper two
images). The additional temperature increase (lower two images) shows peak values (5°C for
the cranial plate and 1.9°C for the mesh) that remain well below the estimated conservative
values of 14.8°C for the cranial plate and 5.0°C for the mesh.

Session: S05
Epidemiology
Tuesday June 26, 2018 • 11:00 - 12:30
Europa C
Chairs: Geza Benke & Leeka Kheifets

S05-1 [11:00]
Validation of self-reported mobile phone use in the smartphone era: Findings from the
SCAMP cohort
Mireku MO1, William Mueller1, Charlotte Fleming1, Irene Chang1, Iroise Dumontheil2, Michael Thomas2,
Marloes Eeftens3, Paul Elliott1, Martin Röösli3 & Mireille Toledano1
1MRC-PHE Centre for Environment and Health, Department of Epidemiology and Biostatistics, Imperial
College London, London, United Kingdom, W2 1PG
2Department of Psychological Sciences, Birkbeck, University of London, London, United Kingdom, WC1E
7HX
3Department of Epidemiology and Public Health, Swiss Tropical and Public Health Institute, Basel,

99
Switzerland
Keywords: Epidemiology, RF/Microwaves, Completed (published)
Presented by: Mireille Toledano
Although adolescents’ self-reported mobile phone use contains inaccuracies with differential
agreement, our findings demonstrate that self-reported usage adequately distinguishes between
high and low use in the Study of Cognition, Adolescents and Mobile Phones (SCAMP) cohort.
Introduction
Mobile phone use is almost ubiquitous amongst adolescents. However, the reliability of self-reported
mobile phone use is an important methodological challenge of epidemiologic research. As such, we
investigated the agreement between self-reported mobile phone use and objective traffic data from
network operators in a subset of adolescents aged 11–12 years participating in the Study of
Cognition, Adolescents and Mobile Phones (SCAMP) cohort. We also investigated the sensitivity
and specificity of self-reported mobile phone use.
Methods
The SCAMP cohort is a prospective secondary school-based cohort study of adolescents across
London-UK, which aims to investigate whether children’s use of mobile phones and other wireless
devices influences their cognitive, behavioural and educational development. Pupils in Year 7 from
39 participating schools (n=6627) completed a computer-based assessment and provided self-
reported data on mobile phone use, including phone call frequency, cumulative call time duration,
and text messages sent per day. In a subset of the SCAMP cohort (n=350), we obtained parental
consent to access mobile phone traffic data from network operators and matched these records with
the self-reported data on mobile phone use. The extent of agreement between self-reported mobile
phone use and objective traffic data use was assessed by Cohen’s weighted Kappa (κ) statistics.
Sensitivity and specificity of self-reported low (< 1 call/day, ≤ 5min of call/day or ≤ 5 text messages
sent/day) and high (≥ 11 calls/day, > 30min of call/day or ≥ 11 text messages sent /day) use were
estimated. The extent of agreement, sensitivity and specificity were analysed separately for
weekdays and weekends.
Results
The agreement between self-reported mobile phone use and mobile operator traffic data was
highest for the call time duration per day on weekdays (38.9%) and weekends (29.4%) compared to
frequency of calls and number of text messages sent. Adolescents overestimated their mobile
phone use during weekends compared to weekdays. Weighted kappa between self-reported and
mobile operator traffic data for call frequency during weekdays was κ = 0.12, 95% CI 0.06-0.18. Call
frequency was the most sensitive for low mobile phone users on weekdays and weekends (77.1,
95% CI: 69.3-83.7 and 72.0, 95% CI: 65.0-78.4, respectively). Specificity was moderate to high for
high users with the highest for call frequency during weekdays (98.4, 95% CI: 96.4-99.5).
Conclusions
Although adolescents’ self-reported mobile phone use contains inaccuracies with differential
agreement, our findings demonstrate that self-reported usage adequately distinguishes between
high and low use. The greater use of mobile smartphones over Wi-Fi networks, compared to mobile
phone networks, indicates that operator traffic data are not the gold standard for exposure
assessment in adolescents. This has important implications for epidemiologic research on the health
effects of mobile phone use in adolescents.

S05-2 [11:15]
Large area mapping of general public radiofrequency power density exposure in Sweden
during the years 2012-2017
Jimmy Estenberg1 & Torsten Augustsson1

100
1SwedishRadiation Safety Authority, Stockholm, Sweden
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Jimmy Estenberg
We have developed a system] for mobile, car based, isotropic spectral measurements of power
densities within the frequency range of 30 MHz to 3 GHz. The system can quickly cover large areas
and the high sampling rate results in extensive data that allows computation of robust medians and
means. 225’000 measurements were performed over the last five years. The median power
densities were within 0.00087 - 2.6 mW/m2 for different types of areas. The highest power density
that we found was 690 mW/m2. We have found a clear relation between population density and
exposure level. Specific cumulative distribution functions of power densities will be presented for
different sources and types of areas.
Introduction
The general public is exposed to a wide range of radiofrequency fields from a variety of manmade
sources. The most significant sources for the public are within 30MHz to 3GHz. Even though the
public’s exposure is known to be well below established guidelines [1], finding hot spots and
monitoring long-term exposure trends are important issues from an environmental assessment point
of view. To do this, a large material for statistical analysis is needed. The Swedish Radiation Safety
Authority has therefore developed a system [2] for mobile, car based, isotropic spectral
measurements of power densities within the frequency range of 30 MHz to 3 GHz. The monitoring
system was first presented at the BEMS meeting in 2012 and we are now ready to present the
results from five years of measurements, including 225’000 spectral power density measurements.
Methods
Measurements are performed while driving and the system is capable to do about 60 detailed full-
spectrum isotropic measurements per minute. This means that a three-hour drive will allow up to
10’800 measurements, covering significant city areas. The monitoring system is based on a
spectrum analyzer and an isotropic three-axis antenna. The resolution bandwidth is set to 1MHz.
The output data includes full-spectrum, polarization, time, date and GPS-coordinates. Single data
has limitations in accuracy due to continuous data acquisition during movement. The output data
has therefore mainly been used to compute average exposure for larger areas. The spectral data
makes it possible to distinguish different sources, e.g. uplink, downlink, specific service providers,
etc.
Results
The median power densities were within 0.00087 - 2.6 mW/m2 for different types of areas. The
highest single total power density that we found was 690 mW/m2. The main contribution was from
base stations for mobile communication (GSM, UMTS and LTE). We have found a clear relation
between population density and exposure level. Specific cumulative distribution functions of power
densities will be presented for different sources and types of areas.
Discussion
Challenges we faced analyzing the huge amount of measurement data involved interpreting and
finding a proper and appropriate format for the presentation of data. This includes of course the
delicate issue of how to consider measurement uncertainties and natural variations in field strength.
The proposed method has some limitations but also advantages compared to other methods that
also performs large area analysis [3]. Our system assesses environmental exposure levels, can
quickly cover large areas, and the high sampling rate results in extensive statistical material that
reduces measurement uncertainties and allows computation of robust medians and means.
However, we do not measure personal exposure from the mobile phone and other handheld
products and our data is limited to roadside exposure.
A strong advantage is that we have a good chance of finding hot spots due to the high sampling rate

101
and the fact that we detect the power density 2.5 meters above the ground. On the other hand, we
might systematically overestimate the public’s exposure as we measure closer to the transmitters,
and also measure in areas that are relatively open and directly exposed to nearby base station
antennas.
Conclusion
All measurements are well below the public’s established guidelines [1]. However, we found a wide
variety in exposure levels that is not only related to the population density but also a result of the
positioning of single transmitting antennas. We believe that there are diversities in how e.g. base
station antennas are positioned in different countries and that this has direct impact on the public’s
exposure to radio waves.
References
[1] International Commission on Non-Ionising Radiation Protection. Guidelines for limiting exposure
to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). Health Physics 74(4):
494-522(1998).
[2] Estenberg, Augustsson (2014). Extensive frequency selective measurements of radiofrequency
fields in outdoor environments performed with a novel mobile monitoring system.
Bioelectromagnetics, 35(3):227-230.
[3] Sagarab, Struchenab, Fintac, Eeftensab, Röösli (2016). Use of portable exposimeters to monitor
radiofrequency electromagnetic field exposure in the everyday environment. Environmental
Research 150:289-298.

S05-3 [11:30]
STUDENT PAPER
Computer/internet use and console gaming: The effect on cognitive function in Australian
primary school students
Christopher Brzozek1, Kurt Benke2, 3, Berihun Zeleke1, Michael Abramson1 & Geza Benke1
1Centre for Population Health Research on Electromagnetic Energy (PRESEE), Monash University,
Melbourne, Australia, 3004
2School of Engineering, University of Melbourne, Melbourne, Australia, 3010
3Department of Economic Development, Jobs, Transport and Resources (DEDJTR), AgriBio Centre,
Melbourne, Australia, 3083
Keywords: Behavioural, Static, Work in Progress
Presented by: Christopher Brzozek
Computer/internet use and console gaming has greatly increased over the last two decades with
children and adolescents becoming frequent users of these technologies at younger ages. The aims
of this study were to investigate any potential associations between computer/internet use and
console gaming with cognitive function in a cohort of primary school children. There were 412
participants from a representative sample of 36 state, private and Catholic schools. Overall
computer/internet use and total screen time were found to have no associations, but console
gaming was found to have significant associations with many of the cognitive functions by either
shortening the response times or reducing the accuracy of responses.
Introduction
Computer/internet use and console gaming has greatly increased over the last two decades, with
children and adolescents becoming frequent users of these technologies at younger ages.
Therefore there has been growing interest on the impact that these technologies may have on the
cognitive skills of frequent users. Previous studies have shown frequent console gaming to be
associated with improved cognitive performance for: change detection (1), visual attention and
inhibition (2, 3), working memory (4) and spatial memory (5). However, the participants from these
102
studies were all young adults. It is plausible that console gaming and computer/internet use may
have different effects on cognitive function depending on the age at which these technologies were
first used. Therefore, the aims of this study were to investigate any potential associations between
computer/internet use and console gaming with cognitive function in a cohort of Australian primary
school children. This study also aimed to investigate if gender modified these associations.
Methods
Study Design
The ExPOSURE study is a prospective cohort study conducted of 4th grade Australian primary
school children in Melbourne and Wollongong. In 2011, 619 students were recruited, and
longitudinal data were collected for 412 students from a representative sample of 36 state, private
and catholic schools across 2012 and 2013. Participating schools selected one or more 4th grade
classes to be invited to the study. The parents or guardians of the children in the selected classes
were then sent plain language information packs detailing the study and consent forms. Participants
completed two examinations, the first at baseline and the second approximately one year later at
follow-up. Each examination required the parents/guardians to complete a questionnaire, which
gathered information on sociodemographic data. While participating students completed a shorter
questionnaire regarding use of certain electronics; including time spent on the computer/internet and
time spent on gaming consoles (such as a PlayStation, Xbox, Gameboy etc.) per week. They also
completed computerized cognitive function tests and the Stroop word colour test. Principals,
teachers, parents/ guardians and students gave written informed consent.
Exposure Assessment
Students’ computer/internet use and time spent console gaming was assessed via a questionnaire.
Student participants were asked “on average how long do you spend on the computer/internet on a
school day?” as well as “on average, how long do you spend on the computer/ internet on weekend
days?” These questions were repeated for time spent console gaming. Total screen time was also
calculated by adding the time spent on both the computer/internet and console gaming.
Outcome Assessment
The validated computerized psychometric test battery CogHealth (CogState, Melbourne, Australia,
2005) and the Stroop colour-word test were used to assess several cognitive function domains.
CogHealth evaluated the response time (ms) and accuracy (%) of simple reaction time with the
detection test, choice reaction time with the identification task, working memory with the one-back
task, visual recognition and memory with the one card learning task and response inhibition with the
go/no-go task. Spatial and executive ability were also assessed by the total number of errors
recorded while completing the Groton maze learning task.
Statistical Analysis
Computer/Internet use was assessed in tertiles of self-reported time spent per week. Quintile
analysis was originally planned but as a large number of students reported exactly 3.5 hours of use
per week, tertiles were the only alternative to have sufficient numbers in each group. Time spent
console gaming was split into quintiles of self-reported use per week. Total screen time was
calculated as the sum of self-reported computer/internet use and self-reported time spent console
gaming per week and was also split into quintiles for the analysis. The mean response time for
cognitive parameters was log transformed, while accuracy was arcsine transformed as
recommended by CogState Research for analysis. The Stroop colour-word tests response times
and error rates were analysed by comparing the results from form B with form A, and the results of
form D with form C. Multiple linear regression models were fitted to each of these cognitive
parameter outcomes for all 3 exposure metrics. Robust standard errors were used to allow for
clustering of students within schools. Age, sex, country of birth, physical activity (times active per
week), socioeconomic status (in quintiles based on postcodes and the Socioeconomic Index for
Areas) and time between examinations were fitted as covariates. An interaction between sex and
computerized console gaming was also included in the model when assessing time spent console

103
gaming per week and cognitive function outcomes. Data analysis was conducted with STATA
version 13.
Results
At baseline 619 students agreed to participate in the ExPOSURE study (52% participation rate), with
412 students participating in both baseline and follow up (66.5% follow up rate). At baseline the
participants’ mean age was 9.9 (SD 0.5) years with 329 females (53%). At follow up the mean age
had increased to 11 (SD 0.5) years with 227 females (55%). Regression models were fitted to each
of the CogHealth and Stroop tasks for Computer/Internet use. Visual recognition memory and
attention (one card learning task) were found to be significantly faster in the students in the 2nd
tertile of computer/ internet use compared with the first. Students in the 2nd tertile were also found
to take significantly longer to complete form D of the Stroop colour word task compared to form C
indicating greater interference. For console gaming and cognitive function, there was a significant
reduction in accuracy seen for visual recognition memory and attention (one card learning task) and
choice reaction time (identification task) in the 4th quintile compared to those who did no console
gaming (first quintile). There were also a significantly higher number of errors made in the Groton
maze learning task (spatial and executive ability) in the 5th quintile. However there was strong
evidence of an interaction between gender and console gaming, and once an interaction term
between console gaming and gender was included in the model more associations were seen. The
reaction times were found to be faster in the 3rd and 5th quintiles of the detection task (Simple
reaction time) and the one-back memory task (working memory). A shorter reaction time was also
seen in the 4th quintiles for these tasks but not at statistically significant levels. The reactions times
were also faster in the 5th quintile of the one card learning task (visual recognition memory and
attention) and in the 3rd quintile of the go/ no go task (response inhibition). However accuracy was
found to decrease in the 5th quintile of the detection task and in the 4th quintile of the one card
learning task. There was strong evidence of a decrease in accuracy in the identification task (choice
reaction time) as the 3rd, 4th and 5th quintiles all showed statistically significant evidence of a
reduction in accuracy compared to those who did no console gaming. For the total screen time
exposure metric, statistically significant results were only seen in the 5th quintile of the Stroop time
ratio for form C and D. This indicated that it took significantly longer to complete form D of the
Stroop colour word task compared to form C indicating greater interference. No associations were
seen between any of the other cognitive functions in these models.
Discussion
There was a weak cross sectional association between computer/internet use and cognitive
function. As only 2 statistically significant results were found from 26 comparisons, these may have
been chance findings. This was also seen in the total screen time models, where only 1 statistically
significant association was found. However, strong and consistent associations were seen between
cognitive function scores and console gaming particularly amongst boys. These associations
showed boys in the highest quintiles of console gaming to be significantly faster, but less accurate
across a range of cognitive function tasks. The faster response times seen for these tasks are
consistent with other studies that have investigated console gaming and cognitive function, but in
older participants (1-4). However, the reduction in accuracy is a concern. It is plausible that these
cognitive function tasks were less interesting than console gaming and therefore the participants
may ‘get bored’ and lose concentration. The strengths of this study included the relatively large
sample size compared to other studies in this field (1-6). Similarly, the sample was representative of
all three school sectors. The use of a well validated and age-appropriate cognitive function test
battery (both CogHealth and Stroop) for outcome assessment is another notable strength of the
study. The main limitations of this study were the cross sectional analysis which prevents any
conclusions about cause and effect being made, as well as the self-reported data used for the
exposure metrics Future studies should adopt a cohort design to determine the temporal sequence
of these associations. It is also imperative to collect more complete exposure data, particularly on
the types of games being played as it is hypothesized that certain genres of game, such as first
person shooter which appeal more greatly to boys, could have a greater effect on cognitive function.
104
Conclusions
Computer/internet use and total screen time were found to have no consistent association with
cognitive function in Australian primary school children. However console gaming had significant
associations with many cognitive functions by either shortening response times or reducing the
accuracy of tasks. These associations were strongly seen in boys in the highest quintiles of console
gaming. However due to the limitations of this study, results need to be interpreted cautiously.
References
1. Clark K, Fleck MS, Mitroff SR. Enhanced change detection performance reveals improved
strategy use in avid action video game players. Acta Psychol (Amst). 2011;136(1):67-72.
2. Cain MS, Prinzmetal W, Shimamura AP, Landau AN. Improved control of exogenous attention in
action video game players. Front Psychol. 2014;5:69.
3. Bavelier D, Achtman RL, Mani M, Focker J. Neural bases of selective attention in action video
game players. Vision Res. 2012;61:132-43.
4. Colzato LS, van den Wildenberg WP, Zmigrod S, Hommel B. Action video gaming and cognitive
control: playing first person shooter games is associated with improvement in working memory but
not action inhibition. Psychol Res. 2013;77(2):234-9.
5. Blacker KJ, Curby KM. Enhanced visual short-term memory in action video game players.
Attention, perception & psychophysics. 2013;75(6):1128-36.
6. Huang V, Young M, Fiocco AJ. The Association Between Video Game Play and Cognitive
Function: Does Gaming Platform Matter? Cyberpsychology, behavior and social networking.
2017;20(11):689-94.

S05-4 [11:45]
Assessment of workers EM exposure during 50 Hz welding process
Mélina Bouldi1
1Electromagnetism, Vibration and Optics Laboratory, French National Institute for Occupational Health and
Safety, Vandoeuvre lès Nancy, France, 54519
Keywords: Occupational, ELF/LF, Work in Progress
Presented by: Melina Bouldi
The low frequency magnetic field radiated from a spot-welding gun, commonly used in French
industries, was experimentally characterized using an automated 3D mapping method. A numerical
model of this source was also developed, and tuned for getting good concordance between
simulated and experimental magnetic fields. For a first risk’s assessment, the volume in which the
2013/35/UE limit is exceeded was determined. Coupling between the numerical source and body
model was then validated, using an analytical approach. New simulations were finally performed to
calculate the electric field induced in the F-VHP human model. This two steps study allows a
complete evaluation of the risk’s area, leading to preventive actions.
Introduction
In France, 80 % of the machines emitting electromagnetic (EM) radiation are welding systems [1],
with a fundamental frequency of 50 Hz for classical metallic welding. Because of the use of a high
current, it seems important to assess the safety of workers exposed to these fields. The welder
should be placed far enough for having an acceptable EMF exposure, but as close as possible to
the welding device so as to obtain a low work cycle time and make the process competitive. In situ
observations of exposure conditions were carried out in several French metallurgical industries [1]; it
has been noted that welders usually place themselves very close to the welding point, in front of the
machine, so as to visually check the soldering. It is in that precise configuration that the exposure
level was evaluated in this study.

105
The European Directive 2013/35/UE [2] defines two relevant limits for frequencies lower than 100
kHz: (1) the “derived values”, DV, i.e. the magnetic – B – field density measured around the EM
source; and (2) the “basic values”, BV, i.e. the body electric – E – field density. When the B field
exceeds the DV, it is necessary to assess the human body intern exposure. The EU Directive
recommends doing this assessment by using EM simulation with human body models [3].
Methods and first results
Because of the use of high current and relatively low voltage, only the magnetic component of the
radiated field is relevant. The corresponding DV for the head – sensory effects – and limbs are
respectively 1 mT and 18 mT [2].
A hand welding gun with integrated transformer was used (Figure 1). The copper electrode has a
dimension of 30 cm by 20 cm, and a mean current of I = 18.87 kA was chosen for welding process
without any metal load. Using a Rogowski probe, it was checked that the current is preponderant in
the electrode, leading to an overriding radiation. A commercial field-meter (Narda – ELT400)
enabled a simultaneous and real-time visualization of the three B field components. Ninety
measurements were carried out using a 3D automatic positioning robot, at different locations
centred on the welding point, on three (x, y) horizontal planes at the heights of z = -10 cm, 0 cm and
10 cm respecting to the welding point (WP). A mapping volume of 20 x 40 x 30 cm3 was finally
covered allowing reconstructing the B field in a 20 x 80 x 30 cm3 volume thanks to system
symmetry. The exposure index values R was calculated according to the Weighted Peak Method [3],
R corresponding to the frequency weighted ratio of the measured field to the DV.
EM simulations were performed using CST Studio Suite, 2017. A simplified model of the copper
electrode was developed [4], leading to a mean difference between Bexp and Bsim of 23% for a 10
cm3 volume next to the WP, and 7% all around. Preliminary results show that, in our specific case of
study, with a worker standing in front of the WP, the DV-head is exceeded in a semi-ellipsoid centred
on the WP, of radius rX = 15 cm, rY = 30 cm, and rZ = 15 cm, as it is shown in Figure 2. Contrarily to
the experiment, BRMSsim has a toroid shape in the immediate vicinity of the WP, with a weak
magnitude in the centre. For getting a better simulation near the WP, model optimization is on-going
by (1) using stranded wire approximation, in which the current density is homogeneous in
conductor’s cross section; and (2) implementing a realistic source respecting the experimental
current profile. First simulation results show that changing the main part of the electrode by a
stranded wire gives a better B field shape around the WP than before.
Furthermore, the CST Low-frequency Time-domain Magneto-Quasi-Static (LT-MQS) solver was
validated in term of coupling between a LF B field source and induced E field in human body. B
fields generated by a system of Helmholtz coil (1 m radius, sinusoidal input current of 10 A
magnitude and 50 Hz frequency) were both calculated analytically using Matlab, and numerically by
CST, with a converging rate of 10-12 accuracy. Preliminary results show a maximum absolute
difference of 10 %. The next step consists on centring a virtual cylinder of body tissue of 30 cm
diameter and σ = 0.2 S/m in the system, and calculating the induced E field, likewise analytically and
numerically.
Perspectives and conclusion
3D LT-MQS simulations are on-going so as to determine the electric field within the F-VHP human
model “Nelly”, depending (1) on the body position respect to the welding electrode, and (2) on the
welding current magnitude. As the DV is depending on which body part is exposed, specific
evaluation are done whether we focus on the head, torso or limb. The total system is already
meshed into nearly 105,000 tetrahedrons using a surface based method. The properties of each
tissue – i.e. density, electrical conductivity and permittivity – have been adjusted in order to be close
to the real human values at 50 Hz, using 4-Cole Cole method.
Finally, statistics will be achieved using the 99 percentile discrimination method defined in [3] so as
to carry on comparisons with the BV fixed at 1.1 V/m at 50 Hz. This will allow a refinement of the

106
location at risks, by defining the 3D effective restriction zone in which the BV at 50 Hz is not
respected. Meanwhile this work, comparison with the EN 62822-3 exposure assessment procedure
[5] will be done. In terms, it will allow a better assessment of the safe area and actions like grounded
zones and/or remote control might be implemented.
References
[1] French National Institute for Occupational Safety and Health – Factsheets ED4200-4217 on
Electromagnetic Safety - 2011
[2] Directive 2013/35/EU - electromagnetic fields - of 26 June 2013 on the minimum health and
safety requirements regarding the exposure of workers to the risks arising from physical agents
(electromagnetic fields)
[3] Non-binding guide to good practice for implementing Directive 2013/35/EU Electromagnetic
Fields.
[4] Bouldi M., Experimental and simulation study of welder’s electromagnetic exposure, in Proc. of
BioEM 2017, Hangzhou, China, June 2017, p158-162.
[5] EN 62822-3:2018 - Electric welding equipment. Assessment of restrictions related to human
exposure to electromagnetic fields (0 Hz to 300 Hz). Resistance welding equipment, Feb. 2018.
Acknowledgements
Thanks to Lucien Hammen and Alexandre Klingler for taking part in the study.
Figures

Figure 1. 50 Hz welding gun with a B field probe fixed on a 3D displacement system.

107
 Figure 2. 3D surface – in grey – in which R = 1 in front of the electrode (simulation results at
t=0.04s, LT-MQS solver, 85,000 tetrahedrons, 10-12 accuracy)

S05-5 [12:00]
STUDENT PAPER
Proximity to overhead power lines and childhood leukemia: An international pooled analysis
Aryana Amoon1, Catherine M. Crespi2, Anders Ahlbom3, Megha Bhatnagar1, Issy Bray4, Katherine Bunch5,
Jacqueline Clavel6, 7, Maria Feychting3, Denis Hemon6, Christoffer Johansen8, 9, Christian Kreis10, Carlotta
Malagoli11, Fabienne Marquant6, Camilla Pedersen8, Ole Raaschou-Nielsen8, 12, Martin Röösli13, 14, Ben
Spycher10, Madhuri Sudan1, 15, 16, John Swanson17, Andrea Tittarelli18, Deirdre Tuck19, 20, Tore Tynes21,
Ximena Vergara1, 22, Marco Vinceti11, 23, Victor Wünsch-Filho24 & Leeka Kheifets1
1Epidemiology, Los Angeles, CA, USA, 90095
2Biostatistics, Los Angeles, CA, USA, 90095
3Epidemiology, Karolinska Institutet, Stockholm, Sweden
4Health and Social Sciences, Bristol, United Kingdom
5Nuffield Department of Population Health, Oxford, United Kingdom
6Epidemiology and Biostatistics Sorbonne Paris Cite Center, Paris, France
7French National Registry of Childhood Hematological Malignancies, Villejuif, France
8Danish Cancer Society Research Center, Copenhagen, Denmark
9Oncology Clinic, Finsen Center, Copenhagen, Denmark
10Institute of Social and Preventive Medicine, Bern, Switzerland
11CREAGEN Research Center of Environmental, Genetic, and Nutritional Epidemiology, Modena, Italy
12Environmental Science, Roskilde, Denmark
13Epidemiology and Public Health, Swiss Tropical and Public Health Institute, Basel, Switzerland
14Basel, Switzerland

108
15Department of Public Health, Aarhus, Denmark
16Pomona, CA, USA, 91766
17National Grid, London, United Kingdom
18Cancer Registry Unit, National Cancer Institute, Milano, Italy
19Hobart, Australia
20Royal Hobart Hospital, Hobart, Australia
21Occupational Health Surveillance, National Institute of Occupational Health, Oslo, Norway
22Energy and Environment Sector, Electric Power Research Institute, Palo Alto, CA, USA, 94304
23Epidemiology, Boston, MA, USA
24Epidemiology, São Paulo, Brazil
Keywords: Epidemiology, ELF/LF, Completed (unpublished)
Presented by: Aryana Amoon
The association between childhood leukemia risk and distance to power lines has been examined in
numerous studies with inconsistent results. In this paper, we provide a comprehensive assessment
of this association and evaluate whether the association is due to magnetic fields or other factors
associated with distance from overhead power lines. We present a pooled analysis combining
individual-level data (29,049 cases and 68,231 controls) from 11 record-based studies. A series of
subgroup, confounder, and sensitivity analyses were performed to evaluate whether bias,
confounding, or other methodologic challenges inherent in these studies have substantial influence
on the results.
Introduction
Thirty-five epidemiologic studies have examined the association between exposure to extremely low
frequency magnetic fields (MF) and childhood leukemia. Analyses that have pooled data from
multiple studies (Ahlbom et al., 2000; Greenland, Sheppard, Kaune, Poole, & Kelsh, 2000; Kheifets
et al., 2010; Schuz et al., 2007) report a small but consistent increased risk of childhood leukemia
associated with exposures above 0.3 or 0.4 microtesla (μT). In one of these analyses, Kheifets et al.
(2010) pooled six studies for an analysis of the association between distance from power lines and
childhood leukemia. They found an odds ratio (OR) of 1.59 (95% CI: 1.02-2.50) for the closest
distance category, which was comparable to the result for MF. High MF can occur close (e.g. <100
meters (m)) to high voltage power lines (Vergara et al., 2015). However, distance is known to be a
poor predictor of MF exposure (Feychting & Ahlbom, 1994), and therefore the question arises as to
whether the association of increased childhood leukemia risk with distance is due to MF or to other
factors associated with distance from overhead power lines that are unrelated to long-term average
MF. In this study, we pool data from multiple studies to provide a comprehensive assessment of the
association between childhood leukemia risk and distance to power lines. We also evaluate whether
the association is due to MF or other factors, and further consider whether bias, confounding, or
other methodologic challenges inherent in these studies have substantial influence on the results.
Methods
The present study is a pooled analysis combining raw individual-level data from multiple studies
identified through PubMed, EMF Portal, and an informal survey of epidemiologist involved in MF
research. To be included in our analysis, a study must have used record-based exposure
assessment, i.e., not requiring active participation of study subjects, with exposure (i.e., distance to
power lines) determined at the individual level; thus, studies with ecologic or area-based exposure
assessment were excluded. We also excluded wire code studies and studies with hospital controls.
We identified 21 studies published between 1993 and 2016, of which 13 met our inclusion criteria
(Figure 1). Eight studies were excluded (Figure 2). Figure 3 shows the distribution of participants
and study years over time.
All variables were recoded to make them as compatible as possible. Distance to power lines was
coded into four categories as the primary analysis (<50 m, 50-<150 m, 150-<300 m, and ≥300 m as

109
the reference); these cut points were selected based on available data and previous literature. The
primary analyses estimated risk of any type of childhood leukemia associated with distance of
residence from power lines and was restricted to participants who had study-defined accurate
geocoding. A mixture of birth and diagnosis homes were used, based on available data, with the
home used in prior publications given preference. We estimated risk for distance from closest
overhead power line of any voltage and from closest power line with voltage of 200 kV or greater.
Analyses were adjusted for age at diagnosis, sex, and SES.
We used two statistical approaches: one-stage meta-analysis and two-stage meta-analysis (Burke,
Ensor, & Riley, 2017). In the one-stage approach, a traditional pooled analysis, data from all studies
were entered simultaneously into a single mixed-effects logistic regression model with random
intercepts for study. In the two-stage approach, effect estimates (log ORs) were obtained for each
study separately and then combined using a random-effects meta-analysis model. A sensitivity
analysis using the two-stage approach included Japan and Finland for which only summary data
were available. For the primary analyses, estimates from these two methods were compared. For all
further analyses, we used the one-stage approach.
Results
We will present subgroup, confounder, and sensitivity analyses. We fitted models for various
subgroups: comparing subtypes of leukemia (ALL and AML), excluding children with Down
syndrome, and comparing subjects younger than five years to those five years or older at diagnosis.
To evaluate whether the strength of the association changed over time, we stratified by decade of
diagnosis in a manner similar to that of Bunch et al (2014), except that due to small numbers, we
grouped the decades as 1960-1980, 1980-2000, and 2000 and later. The latter analysis was
conducted both with and without the UK study, because it was the hypothesis generating study.
We examined the effects of confounder adjustments on risk estimates. Confounders examined
included residential mobility (moving between the time of birth and diagnosis) for five studies, type of
dwelling (single-family home or other) for four studies, traffic exposure (high, medium, or low) for
four studies, urban versus rural setting for seven studies, ecological measures of SES for six
studies, individual measures of SES for five studies, and MF for eight studies. The latter analysis
was performed both with and without Brazil, the only country with measured rather than calculated
fields. Completeness of collected confounder information varied across studies; many studies with
confounder information had substantial subject-level missing data. We further analyzed the
association between childhood leukemia risk adjusting for each confounder individually, controlling
for age, sex, and SES. As confounder information was available only for subsets of studies, we
present ORs from both minimally adjusted models (adjusted for age, sex, and SES) and models with
confounders fit to the same subset of data.
Sensitivity analyses included comparing the association based on birth homes to that in diagnosis
homes, as well as the choice of the reference category (e.g. ≥300 m vs. ≥600 m). To assess how
geocoding accuracy may result in exposure misclassification, we conducted an analysis of all
observations, regardless of geocoding quality, compared to one including only observations with
good geocoding. Finally, we repeated the primary analysis using alternative controls. These
analyses used data from studies that assessed other cancers in addition to leukemia (California,
Italy 2, Sweden, Switzerland, Tasmania, and the UK). We used controls matched to cases of other
cancers (central nervous system tumors, lymphoma, and other cancers), and conducted an analysis
combining all alternative controls.
Further results will be presented at the meeting.
Discussion
This is a first comprehensive assessment of the association between distance to power lines and
childhood leukemia to evaluate whether magnetic fields or other factors related to distance are
responsible for the association. This international pooling effort allows for an insight not possible in
the individual studies previously published. We discuss how bias, confounding, and other

110
methodologic challenges inherent in these studies influence the results.
References
Ahlbom, A., Day, N., Feychting, M., Roman, E., Skinner, J., Dockerty, J., . . . Verkasalo, P. K. (2000).
A pooled analysis of magnetic fields and childhood leukaemia. Br J Cancer, 83(5), 692-698.
doi:10.1054/bjoc.2000.1376
Burke, D. L., Ensor, J., & Riley, R. D. (2017). Meta-analysis using individual participant data: one-
stage and two-stage approaches, and why they may differ. Stat Med, 36(5), 855-875.
doi:10.1002/sim.7141
Feychting, M., & Ahlbom, A. (1994). THE AUTHORS REPLY. American Journal of Epidemiology,
140(1), 75-75. doi:10.1093/oxfordjournals.aje.a117169
Greenland, S., Sheppard, A. R., Kaune, W. T., Poole, C., & Kelsh, M. A. (2000). A pooled analysis of
magnetic fields, wire codes, and childhood leukemia. Childhood Leukemia-EMF Study Group.
Epidemiology, 11(6), 624-634.
Kheifets, L., Ahlbom, A., Crespi, C. M., Draper, G., Hagihara, J., Lowenthal, R. M., . . . Wunsch
Filho, V. (2010). Pooled analysis of recent studies on magnetic fields and childhood leukaemia. Br J
Cancer, 103(7), 1128-1135. doi:10.1038/sj.bjc.6605838
Schuz, J., Svendsen, A. L., Linet, M. S., McBride, M. L., Roman, E., Feychting, M., . . . Ahlbom, A.
(2007). Nighttime exposure to electromagnetic fields and childhood leukemia: an extended pooled
analysis. Am J Epidemiol, 166(3), 263-269. doi:10.1093/aje/kwm080
Vergara, X. P., Kavet, R., Crespi, C. M., Hooper, C., Silva, J. M., & Kheifets, L. (2015). Estimating
magnetic fields of homes near transmission lines in the California Power Line Study. Environ Res,
140, 514-523. doi:10.1016/j.envres.2015.04.020
Figures

111
Figure 1. Characteristics of Studies Meeting Criteria for Pooled Analysis of Childhood
Leukemia and Distance to Power Lines

112
Figure 2. List of Excluded Studies and Reasons for Exclusion

113
 Figure 3. Distribution of Cases and Controls and Studies by Years of Diagnosis

S05-6 [12:15]
Residential mobility and childhood leukemia in California Power Lines Study
Aryana Amoon1, Sona Oksuzyan2, Catherine M. Crespi3, Onyebuchi Arah1, Myles Cockburn4, Ximena
Vergara5 & Leeka Kheifets1
1Epidemiology, Los Angeles, California, USA, 90095
2Los Angeles County Department of Public Health, Los Angeles, California, USA, 90005
3Biostatistics, Los Angeles, California, USA, 90095

114
4Preventative
Medicine, Los Angeles, California, USA, 90089
5Energyand Environment Sector, Electric Power Research Institute, Palo Alto, California, USA, 94304
Keywords: Epidemiology, ELF/LF, Completed (unpublished)
Presented by: Aryana Amoon
Studies of environmental exposures and childhood leukemia do not usually account for residential
mobility. In addition to being a potential risk factor, mobility can induce selection bias, confounding,
or measurement error in such studies. We attempt to disentangle this using stratified logistic
regression, case-only analysis, and propensity-score adjustments to assess predictors of residential
mobility between birth and diagnosis, and account for potential confounding due to it. Mobility varied
by several sociodemographic characteristics, but not distance to the nearest power line nor
calculated magnetic fields. Mobility appears to be an unlikely explanation for observed associations
between power lines exposure and childhood leukemia.
Introduction
The majority of studies that have evaluated the role that environmental exposures play in the
development of childhood leukemia have not considered the mobility of subjects. Residential
mobility, or moving between time of birth and diagnosis, can involve short distances, such as moving
within the same neighborhood, or longer distance moves; the likelihood of experiencing similar
environmental exposures before and after a move may depend on distance. Subjects can also move
out of the study area and be lost to follow-up. As only one residential address is available in most
studies, few studies can directly assess residential mobility.
Mobility has been considered a source of potential bias in childhood leukemia studies as it can
affect study participation and selection, result in exposure misclassification, or confound the results
(Kheifets, Swanson, Yuan, Kusters, & Vergara, 2017). Figure 1 provides a simplified directed acyclic
graph illustrating how mobility could affect studies of childhood leukemia and electro-magnetic fields
(EMF) in particular.
It has been hypothesized that mobility can explain an association between EMF and childhood
leukemia (Sahl, 1994). A previous study (Jones, Shih, Thurston, Ware, & Cole, 1993) found that
people who moved had a higher proportion of “high” wire codes (an imperfect exposure surrogate)
than those who were residentially stable. Another study evaluated residential mobility of adults and
proximity to power lines in the UK (Swanson, 2013), but found that proximity did not appear to
clearly affect the likelihood of moving. Direct data on mobility of children is lacking.
We conducted a large epidemiologic case-control study in California to examine the associations of
childhood leukemia with calculated magnetic fields and with distance from the birth address to the
nearest high-voltage overhead transmission line. In common with other case-control studies of
childhood cancers, cases, but not controls, had to reside in California at time of diagnosis. The aims
of this analysis are to: (i) describe factors that affect or predict mobility among childhood leukemia
cases; (ii) use such factors as proxies to adjust for mobility; and (iii) evaluate potential confounding
due to residential mobility in the study of the potential effect of EMF exposure from nearby power
lines on childhood leukemia.
Methods
The California Power Lines Study (CAPS) included childhood leukemia cases younger than 16
years diagnosed in California between 1988 and 2008 who were also born in California. Controls
were randomly selected from the California Birth Registry (CBR) and matched to cases (1:1).
Controls were excluded if they were diagnosed with any type of cancer in California before the
matched case’s date of diagnosis. Although cases had to be both born in and diagnosed in
California, because controls were selected from birth records, they were born in California, but were
not required to be residing in the state at time of diagnosis of the corresponding case. Thus, we had
birth addresses for both cases and controls, but address at diagnosis for cases only.
We determined geocoded latitudes and longitudes for cases’ residential addresses using the
University of Southern California (USC) GIS Laboratory’s open-source geocoder, which uses parcel
115
level data for Los Angeles County and street level data for the whole of California (Goldberg &
Cockburn, 2010). Only addresses with parcel or street segment matching, which corresponds to
more precise geocoding, were included in this analysis.
We created three categories of residential mobility for cases: 1) not moved, 2) moved within a
neighborhood, defined as distance between birth and diagnosis addresses 50-2000 meters (m), and
3) moved outside of a neighborhood, defined as distance between birth and diagnosis addresses of
2000 m or further. For the primary analysis, the latter two were collapsed and cases were classified
as either residentially stable (did not move) or residentially mobile (moved). To allow for minor
geocoding differences over the years, if the distance between birth and diagnosis addresses was 0
to 50 m, we assumed the subject lived in the same property and did not move. This assumption was
verified by examining Google satellite images for a larger set of residences (with distances <100 m
between birth and diagnosis addresses). 50 m was chosen to increase specificity and make
estimates more conservative.
Proximity to power lines was defined as distance from the child’s address to any power line or to the
nearest power line of 200 kV and above (Kheifets et al., 2015) and was classified into 8 categories:
<50 m, 50-<100 m, 100-<200 m, 200-<300 m, 300-<400 m, 400-<500 m, 500-<600 m, and no lines
within 600 m. Due to small numbers in one of the categories, a sensitivity analyses was run in which
the closest two categories were combined (<100 m). Birth homes located close to lines were site-
visited to verify distance, collect additional information needed for magnetic fields calculations, and
ascertain dwelling type (single-family homes vs other). Site visits, only available for a subset of
subjects (n=178), were conducted blind to case-control status to reduce bias. Calculated fields
estimating fields at time of birth were classified into three categories: ≥0.4 microTesla (μT), 0.1-<0.4
μT, and <0.1 μT (Vergara et al., 2015).
We considered the following variables as covariates: child’s age at diagnosis, number of siblings
living, census-based SES, race/ethnicity, mother's age, mother’s years of education, father's years
of education, mother's place of birth, the payment source for delivery, type of dwelling, proximity to
high voltage power lines, and calculated fields. All covariates were modeled as categorical
variables. For more details, see a previous study (Oksuzyan et al., 2015b).
We will present results on power lines exposure on childhood leukemia stratified by mobility status
of cases, case-only analysis assessing variables predicting mobility, comparing birth and diagnosis
homes in movers, and using mobility predictors and propensity score for mobility as surrogates for
mobility in the case-control analysis of EMF exposure on leukemia. The propensity scores were
created using multinomial logistic regression with the variables associated with mobility as
predictors. We estimated propensity scores for each subject as the predicted probability from the
model based on their covariate values. We then fit logistic regression models for the outcome of
childhood leukemia that included the exposure variable (proximity or calculated field) with and
without adjusting for the propensity score, to assess whether the adjustment changed the childhood
leukemia risk estimate.
Results
4,879 childhood leukemia cases and 4,835 matched controls met the eligibility criteria and had
study-defined good geocoding. Most cases (2,982, 61.1%) moved between birth and diagnosis.
Among those who moved, 618 stayed within 2 km of their birth home, while 1,992 moved outside of
their birth neighborhood. There were no differences in characteristics between cases and controls.
However, among cases who moved, children tended to be older, live in housing other than single-
family homes, and have younger mothers. Fewer siblings and lower SES were also more common
among children who moved.
Odds ratios for leukemia among non-movers living <50 meters from a 200+ kilovolt line (OR: 1.62;
95% CI: 0.72-3.65) and for calculated fields ≥0.4 microTesla (OR: 1.71; 95% CI: 0.65-4.52) were
slightly higher than previously reported overall results.
Further results will be presented at the meeting.

116
Discussion
The residential mobility of childhood leukemia cases varied by several sociodemographic
characteristics, but not by the distance to the nearest power line or calculated magnetic fields.
Mobility appears to be an unlikely explanation for the associations observed between power lines
exposure and childhood leukemia.
Strengths of this study include the use of population registries to obtain data, thus avoiding
participation bias and exposure assessment blind to case-control status to reduce information bias.
To increase accuracy of exposure and outcome assessment, we exclude from analyses all cases
and controls with imprecise geocode matching for birth or diagnosis address. Another strength is the
large sample size, which increases the power to detect associations, should they exist. Despite the
large sample size, in some analyses, especially those involving dwelling type, the analytic sample is
reduced because the variable is only available for site-visited residences. Since site visits were
conducted blind to case-control status, the potential for biases is probably small, and the impact is
mainly on the precision of the estimates. Lastly, the propensity scores allowed us to adjust for the
propensity to move for both cases and controls and thus partially overcome lack of mobility
information for controls.
References
Crespi, C. M., Vergara, X. P., Hooper, C., Oksuzyan, S., Wu, S., Cockburn, M., & Kheifets, L. (2016).
Childhood leukaemia and distance from power lines in California: a population-based case-control
study. Br J Cancer, 115(1), 122-128. doi:10.1038/bjc.2016.142
Jones, T. L., Shih, C. H., Thurston, D. H., Ware, B. J., & Cole, P. (1993). Selection bias from
differential residential mobility as an explanation for associations of wire codes with childhood
cancer. J Clin Epidemiol, 46(6), 545-548.
Kheifets, L., Crespi, C. M., Hooper, C., Cockburn, M., Amoon, A. T., & Vergara, X. P. (2017).
Residential magnetic fields exposure and childhood leukemia: a population-based case-control
study in California. Cancer Causes Control. doi:10.1007/s10552-017-0951-6
Kheifets, L., Crespi, C. M., Hooper, C., Oksuzyan, S., Cockburn, M., Ly, T., & Mezei, G. (2015).
Epidemiologic study of residential proximity to transmission lines and childhood cancer in California:
description of design, epidemiologic methods and study population. J Expo Sci Environ Epidemiol,
25(1), 45-52. doi:10.1038/jes.2013.48
Kheifets, L., Swanson, J., Yuan, Y., Kusters, C., & Vergara, X. (2017). Comparative analyses of
studies of childhood leukemia and magnetic fields, radon and gamma radiation. J Radiol Prot, 37(2),
459-491. doi:10.1088/1361-6498/aa5fc7
Oksuzyan, S., Crespi, C. M., Cockburn, M., Mezei, G., & Kheifets, L. (2012). Birth weight and other
perinatal characteristics and childhood leukemia in California. Cancer Epidemiol, 36(6), e359-365.
doi:10.1016/j.canep.2012.08.002
Oksuzyan, S., Crespi, C. M., Cockburn, M., Mezei, G., Vergara, X., & Kheifets, L. (2015a).
Race/ethnicity and the risk of childhood leukaemia: a case-control study in California. J Epidemiol
Community Health, 69(8), 795-802. doi:10.1136/jech-2014-204975
Oksuzyan, S., Crespi, C. M., Cockburn, M., Mezei, G., Vergara, X., & Kheifets, L. (2015b). Socio-
economic status and childhood leukemia in California. Journal of Cancer Prevention & Current
Research, 3(4). doi:10.15406/jcpcr.2015.03.00087
Sahl, J. D. (1994). Viral contacts confound studies of childhood leukemia and high-voltage
transmission lines. Cancer Causes Control, 5(3), 279-283.
Swanson, J. (2013). Residential mobility of populations near UK power lines and implications for
childhood leukaemia. J Radiol Prot, 33(3), N9-14. doi:10.1088/0952-4746/33/3/N9
Vergara, X. P., Kavet, R., Crespi, C. M., Hooper, C., Silva, J. M., & Kheifets, L. (2015). Estimating
magnetic fields of homes near transmission lines in the California Power Line Study. Environ Res,
140, 514-523. doi:10.1016/j.envres.2015.04.020

117
Yost, K., Perkins, C., Cohen, R., Morris, C., & Wright, W. (2001). Socioeconomic status and breast
cancer incidence in California for different race/ethnic groups. Cancer Causes Control, 12(8),
703-711.
Figures

Figure 1. Simplified directed acyclic graph (DAG) depicting possible connections of residential
mobility in the study of EMF exposures on childhood leukemia.

Session: S06
In vitro II
Tuesday June 26, 2018 • 11:00 - 12:30
Europa D
Chairs: Guangdi Chen & Carmela Marino

S06-1 [11:00]
Weak static magnetic fields induce changes in fibrosarcoma cell growth, hydrogen peroxide,
membrane potential and mitochondrial calcium
Hakki Gurhan1, Rodolfo Bruzon1, Sahithi Kandala1, Yanyu Xiong1 & Frank Barnes1
1Electrical, Computer and Energy Engineering, University of Colorado at Boulder, Boulder, Colorado, USA,
80309
Keywords: In vitro, Static, Work in Progress
Presented by: Hakki Gurhan
Fibrosarcoma cells are exposed to weak static magnetic fields between 0.5 mictoT and 600 microT.
Changes in cell growth, hydrogen peroxide, membrane potential and mitochondrial calcium are
observed.
Reactive oxygen species (ROS) are useful as signaling molecules in mammalians but also they

118
cause cellular damage if produced in an uncontrolled manner. Multiple factors such as electric and
magnetic field type, intensity, frequency, exposure time and the cell confluence and temperature
could cause changes in ROS levels. We have used a pair of square Helmholtz Coils to study the
effect of the static magnetic field (SMF) on ROS production. The windings of the Helmholtz coils are
made of 1.63 mm wire to reduce heating problems produced by the Joule effect and to keep the
temperature of the cell culture around 37 °C. Reduction of background magnetic field is achieved by
using mu-metal boxes.
Human fibrosarcoma cancer cell line HT1080 is used to observe the cell growth rate and changes in
mitochondrial calcium, superoxide, hydrogen peroxide and membrane potential under the exposure
of external static magnetic fields. Cells are exposed to SMF between 0.5 microT and 600 microT for
4 days. Increase in cell growth rate is observed until 400 microT. Cell growth rate started to
decrease at 500 microT and it continue to decrease at 600 microT. Also increase in mitochondrial
calcium, superoxide and membrane potential are observed. However, hydrogen peroxide
concentration decreased while exposed SMF strength increases. Hydrogen peroxide belongs to the
ROS and its concentration is a potential determining factor in the final biological outcome. The low
intensity ROS presence can act as second messenger and activate signaling cascades in response
to gene expression, cell proliferation. On the other hand moderate levels of ROS can activate
apoptosis and high level of ROS can lead to necrosis. Controlling mitochondrial ROS production in
cancer cells can be exploited for therapeutic benefits.

S06-2 [11:15]
Effects of flux density, field direction, frequency and presence of blue light on in vitro
responses to magnetic fields: implications for mechanisms
Jukka Juutilainen1, Jukka Luukkonen1, Mikko Herrala1, Anne Höytö1, Kavindra Kesari1 & Jonne Naarala1
1Department of Environmental and Biological Sciences, University of Eastern Finland, Kuopio, Finland, 70211
Keywords: In vitro, ELF/LF, Completed (unpublished)
Presented by: Jukka Juutilainen
Understanding biophysical mechanism for biological effects of weak alternating magnetic fields (MF)
would greatly enhance causal inferences concerning the suggested health effects. In this
presentation, we describe experiments with varying MF parameters and experimental conditions.
We then use the findings as a basis for discussing possible mechanistic explanations, including the
radical pair mechanism, magnetite hypothesis and resonance interactions. There are problems with
all explanations. While many findings support the radical pair mechanism, explaining the effects of
weak alternating MFs in the presence of the static geomagnetic field remains a challenge.
Introduction
Epidemiological studies have shown associations between health effects and weak environmental
extremely low frequency (ELF) magnetic fields (MF), but causality of these associations remains
unclear. Understanding any biophysical mechanism for biological effects of weak alternating MFs
would greatly enhance causal inferences concerning the suggested health effects. In this paper, we
describe experiments that were done using varying MF parameters and experimental conditions,
and discuss possible mechanistic explanations in light of the findings.
Methods
Human SH-SY5Y neuroblastoma cells, rat C6 glioma cells or rat primary astrocytes were cultured
as described previously [1-8]. The cells were exposed for 24 h to MFs in Helmholz-type coils or
sham-exposed and (after MF exposure or during the last 3 h of MF exposure) for 3 h to menadione

119
(MQ) or methyl methane sulfonate (MMS) or incubated for 3 h. The chemicals are known DNA-
damaging agents (with different modes of action) that were used as cofactors in the experiments.
The measurements, performed as described previously [1-8], included micronucleus frequency,
relative cell number, mitochondrial and cytosolic superoxide levels, DNA damage level and DNA
repair using the Comet assay, and induced genomic instability assessed as delayed genotoxicity.
Results
Summary of some previous findings
We found evidence that exposure to a 50 Hz, 100 µT MF affected responses to MQ-induced DNA
damage [1] and induced genomic instability [2,3] in cultured cell lines, but there was no immediate
DNA damage from MF exposure alone. However, we reasoned that there must be such cellular
changes that occur already during the 24 h of MF exposure, and should be observable before (or
without) exposure to MQ. Altered cytosolic and mitochondrial superoxide levels [2,4] and changes in
cell cycle distribution and p21 protein expression [6] appeared to be such changes. Of these,
superoxide levels were used in studies on the role of magnetic flux density, MF direction and
exposure to blue light. In exposure-response studies, effects were found down to 10 µT (the lowest
flux density used), and all data were not consistent with a conventional rising exposure-response
relationship [4]. In the experiments with varying MF direction, a 50 Hz, 30 µT MF affected cytosolic
superoxide level when it was horizontal (perpendicular to the 33 µT ambient static MF in the
incubator), while a similar vertical MF had no effect [7]. The experiments involving blue light were
performed to test the radical pair mechanism. The hypothesis was that MFs affect light-induced
radical reactions in cryptochrome proteins, and MF effects should therefore occur only in the
presence of blue light. The results did not confirm this simple hypothesis; on the contrary, MF effects
were observed without blue light, and blue light suppressed MF effects [8].
New results with 7.5 kHz magnetic fields
We have conducted studies to assess possible health risks of intermediate frequency (IF) MFs.
These studies included in vitro experiments with rat primary astrocytes exposed to horizontal or
vertical 7.5 kHz MFs. The exposure system as well as the MF directions and flux densitites were the
same as those used in the corresponding 50 Hz experiments. The most obvious and consistent
effect in the IF MF experiments was increased relative cell number in the exposed samples. The
relative cell number is measured (to assay possible cytotoxicity) in the flow cytometric micronucleus
assay, which is performed after a sufficient time post exposure (144 h in case of astrocytes) to allow
micronuclei to form during cell division. It thus seems that the 7.5 kHz MF either suppresses cell
death or stimulates proliferation of rat primary astrocytes. This effect did not depend on the direction
(horizontal or vertical) of the IF MF (Figure 1). It was consistently observed in experiments with or
without the genotoxic chemicals, and the effect size did not increase when the magnetic flux density
was increased from 30 µT to 300 µT. Some effects on genotoxicity indicators were observed, and
also these findings mainly indicated (slight) “protective” effects of the IF MF.
Discussion
Our discussion focuses on potential mechanistic explanations for the findings described above. In
this discussion, we assume that all effects observed are based on the same fundamental
mechanism, and that the different biological responses observed at 7.5 kHz and 50 Hz are related to
the different biological models used (primary astrocytes vs. secondary cell lines). Only further
research can show whether these assumptions are correct.
Effects based on induced electric fields and currents are not likely to explain the effects observed,
as effects resulted only from a horizontal alternating MF (at 50 Hz) or from both MF orientations (at
7.5 kHz); a vertical alternating MF induces stronger internal fields in the cell culture medium than a
horizontal MF. Therefore, we focus the rest of the discussion on proposed mechanisms explaining
direct MF effects.
The 50 Hz MF was effective only when it was perpendicular to the ambient static MF, indicating that
the static MF is somehow involved. This could be interpreted to fit with the proposed mechanisms

120
involving resonance effects due to combinations of static MF strength and alternating MF frequency,
such as the ion cyclotron resonance, ion parametric resonance and parametric resonance models.
However, these resonance models have mainly focused on explaining effects of parallel alternating
and static fields. MF effects based on the radical pair mechanism and on magnetic particles
(magnetite) are both believed to be involved in animal navigation and orientation in the geomagnetic
field, so it is conceivable that these mechanisms would respond to the ambient static field. However,
these theories do not currently offer predictions concerning combined effects of static and oscillating
MFs at the frequencies used in these experiments.
Effects of light are not included in the magnetite hypothesis or the resonance models, so the
observed interaction between MF and blue light is not easily explained by these theories. It is also
obvious that the results do not fit with the prediction that (only) light-induced radical reactions are
responsive to MF exposure. If magnetosensitive radical pair reactions are involved, they are likely to
be light-independent. However, as interactions between blue light and MF were observed, light-
sensitive molecules such as cryptochromes may nevertheless be involved (possibly as downstream
components) in MF detection.
The lack of a conventional rising exposure-response relationship (both at 50 Hz and 7.5 kHz)
indicates that the MF effects may result from the “information” content of the field rather than from its
“energy” content. If MFs are received as signals informing cells about the environment (e.g.,
reception of the geomagnetic field) or if they disturb reception of such information, increasing field
strength does not necessarily lead to a more pronounced biological response (once the threshold for
signal reception or interference has been exceeded). From this point of view, the findings fit with the
hypothesis that the effects of oscillating MFs are mediated by the same detectors that are known to
be involved in sensing the geomagnetic field (the radical pair mechanism and/or magnetite). It
should also be noted that the radical pair mechanism should result in a highly non-linear exposure-
response relationship.
Effects observed at 7.5 kHz are not likely to be explained by the resonance mechanisms, as the
resonance frequencies predicted by these models are much lower. It is currently unclear whether a
magnetite-based mechanism could explain effects at this frequency. The radical pair mechanism
should work at 7.5 kHz: because of the short lifetime of spin coherence in radical pairs, alternating
MFs up to the MHz range should be ‘perceived’ by the radical pair similarly as static fields. On the
other hand, this is also a major problem in using the radical pair mechanism to explain biological
effects of weak alternating MFs: how could a 10 µT alternating MF (weakest field used in these
experiments) markedly affect a radical pair reaction in the presence of the stronger (33 µT) static
field? This problem becomes even more obvious if we try to use the radical pair mechanism as an
explanation for the epidemiological findings suggesting effects above 0.4 µT.
In conclusion, many of the empirical findings fit with involvement of radical pair mechanism.
However, it will be necessary to identify credible mechanisms that would allow an alternating MF to
cause significant effects in the presence of a static field. This question has been discussed
elsewhere [9].
References
1. Luukkonen J, Liimatainen A, Höytö A, Juutilainen J, Naarala J: Pre-exposure to 50 Hz magnetic
fields modifies menadione-induced genotoxic effects in Human SH-SY5Y neuroblastoma cells.
PLoS ONE 6(3): e18021 (2011).
2. Luukkonen J, Liimatainen A, Juutilainen J, Naarala J: Induction of genomic instability, oxidative
processes, and mitochondrial activity by 50 Hz magnetic fields in human SH-SY5Y neuroblastoma
cells. Mutat Res 760: 33–41 (2014).
3. Kesari KK, Luukkonen J, Juutilainen J, Naarala J: Genomic instability induced by 50 Hz magnetic
fields is a dynamically evolving process not blocked by antioxidant treatment. Mutat Res 794: 46-51
(2015).

121
4. Kesari KK, Juutilainen J, Luukkonen J, Naarala J: Induction of micronuclei and superoxide
production in neuroblastoma and glioma cell lines exposed to weak 50 Hz magnetic fields. J Royal
Soc Interface 13: 20150995 (2016).
5. Herrala M, Mustafa E, Naarala J, Juutilainen J: Assessment of genotoxicity and genomic
instability in rat primary astrocytes exposed to 872 MHz radiofrequency radiation and chemicals.
IJRB, in press.
6. Luukkonen J, Höytö A, Sokka M, Liimatainen A, Syväoja J, Juutilainen J, Naarala J: Modification
of p21 level and cell cycle distribution by 50 Hz magnetic fields in human SH-SY5Y neuroblastoma
cells. IJRB 93:240-248 (2017)
7. Naarala J, Kesari KK, McClure I, Chavarriaga C, Juutilainen J, Martino CF: Direction-dependent
effects of combined static and ELF magnetic fields on endothelial cell proliferation and superoxide
radical production. BioMed Res Int 2017: 5675086 (2017).
8. Höytö A, Herrala M, Luukkonen J, Juutilainen J, Naarala J: Cellular detection of 50 Hz magnetic
fields and weak blue light: effects on superoxide levels and genotoxicity. IJRB 93:240-248 (2017).
9. Barnes FS, Greenebaum B: The effects of weak magnetic fields on radical pairs.
Bioelectromagnetics 36:45-54 (2015).
Figures

Figure 1. Relative cell number in rat primary astrocytes exposed to 7.5 kHz, 30 µT vertical or
horizontal magnetic fields (MF) for 24 h and subsequently to methyl methane sulfonate (MMS)
for 3 h. The ambient static MF was 33 µT and almost vertical (inclination 80-85o).

S06-3 [11:30]
STUDENT PAPER
The effects of 50 Hz magnetic fields and low dose cadmium co-exposure on cell viability in
JAR cells
Yumin Jin1, Aziguli Yimaer1, Chuan Sun1, Liling Su1, Zhengping Xu1 & Guangdi Chen1
1Bioelectromagnetics lab , Zhejiang University School of Medicine, Hangzhou, China, 310058
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: Yumin Jin

122
In this study, we used JAR cells as main research cell model, and found that 50 Hz MF exposure
enhanced cell viability in JAR cells in a time dependent way and could attenuate cadmium inhibited
cell viability. Exposure to MF or cadmium, or co-exposure to MF and cadmium changed protein
expression profile, and the differentially expressed proteins were enriched for signal transduction,
protein translation and stress related pathways. Our data suggest that MF exposure enhanced cell
viability and attenuated cadmium inhibited cell viability may be through changes of proteome
profiling.
Background
With the rapid development of electric power industry, public concerns have been raised about the
environmental exposure to extremely low frequency magnetic fields (ELF-MF) and its potential
health effects. Based on the limited epidemiological evidence, the International Agency for Research
on Cancer (IARC) has classified ELF-MF are possible carcinogens to humans (Group 2B) in 2002.
However, the effect and its mechanism of ELF-MF exposure have not been clarified in laboratory
studies. Thus, the mechanisms of ELF-MF exposure induced biological effects remain to be
elucidated.
Methods
In the present study, we used ELF-MF sensitive cells JAR cells as main research cell model, JEG-3
cells as the second research cell model examined the effects of 50 Hz MF exposure at 3.0 mT on
JAR cell viability using CCK-8 assay. Next, we investigated the co-exposure to 3.0 mT, 50 Hz MF
and known carcinogen cadmium on cell viability using CCK-8 assay in JAR cells. These
experimental evaluations were also conducted in JEG-3 cells, and the differential responses of JAR
and JEG-3 cells to MF exposure with or without co-exposure to cadmium were also compared. To
further explore the mechanisms of 50 Hz MF influenced cell viability, we also applied proteomic
approaches and examined the effects of 50 Hz MF exposure with or without co-exposure to
cadmium on proteome profiling in JAR cells.
Results
Part 1: The effects of 50 Hz MF on cell viability in JAR cells
In JAR cells, the results showed that, comparing to sham exposure group, the cell viability was
significantly increased in 3.0 mT of 50 Hz MF exposure group for 6 hours (p<0.05), 24 hours
(p<0.01) and 48 hours (p<0.01). In JEG-3 cells, we found that, comparing to the sham exposure
group, the cell viability was significantly increased in 3.0 mT of 50 Hz MF exposure group for 6
hours (p<0.05) and 24 hours (p<0.01), but the cell viability significantly decreased in 48 hours
(p<0.05) exposure group.
Part 2: The effects of co-exposure to 50 Hz MF and cadmium on cell viability in JAR cells
Firstly, we investigated the effects of co-exposure to low dose cadmium and chromium on cell
viability in JAR cells. Results showed that, comparing to the control group, cell viability was
significantly decreased in co-exposure to 0.25 μM cadmium and 0.50 μM chromium group for 6
hours, 24 hours or 48 hours (p<0.01). Next, we investigated the effects of 50 Hz MF and low dose
cadmium combine exposure on cell viability in JAR cells. Results showed that, comparing to the
0.25 μM cadmium exposure group, cell viability was significantly increseaed in the co-exposure to
50 Hz MF and cadmium group for 6 hour or 24 hour (p<0.01)，but not for 1 hour or 48 hour.
Similarly, we found that, in JEG-3 cells, comparing to the 0.25 μM cadmium exposure group, cell
viability was significantly increseaed in the co-exposure to 50 Hz MF and 0.25 μM cadmium group
for 1 hour, 6 hour or 24 hour (p<0.01), but not for 48 hours.
Part 3: The effects of co-exposure to 50 Hz MF and cadmium on proteome profiling in JAR
cells
To establish the protein expression response to environmental stimuli in JAR cells, we firstly used
p53 as a responsive protein, and evaluated the effects of 3.0 mT of 50 Hz MF exposure and/or 2.5
μM cadmium on p53 expression. As expected, cadmium exposure up-regulated p53 expression,

123
while 50 Hz MF exposure did not affect the p53 expression. Interestingly, comparing to the cadmium
exposure group, p53 expression level was decreased in the co-exposure to 50 Hz MF and cadmium
group.
Next we used Tandom Mass Tag quantitative proteomic technology to investigate the JAR cells
proteome profiling changes after exposure to 3.0 mT of 50 Hz MF and/or 0.25 μM cadmium for 6
hours. The results showed that, comparing to the sham exposure group, MF exposure resulted in 16
differentially expressed proteins, including 13 proteins up regulated and 3 proteins down regulated,
which were related to signal transduction, protein translation and cytoskeleton associated pathways.
Exposure to cadmium resulted in 111 differentially expressed proteins, including 60 proteins up
regulated and 51 proteins down regulated, which were mainly related to cytoskeleton associated,
signal transduction and stress related pathways. Co-exposure to 50 Hz MF and cadmium resulted in
162 differentially expressed proteins, including 78 proteins up regulated and 84 proteins down
regulated, which were related to signal transduction, protein translation and stress related pathways.
Interestingly, comparing to cadmium exposure, co-exposure to MF and cadmium resulted in 163
differentially expressed proteins, including 72 proteins up regulated and 91 proteins down regulated,
which were related to signal transduction, protein translation and cytoskeleton associated pathways.
Conclusion
In conclusions, this study demonstrated that, under current exposure conditions 1) MF exposure
enhanced cell viability in time dependent way; 2) MF exposure could attenuate cadmium inhibited
cell viability in JAR cells; 3) exposure to MF or cadmium, or co-exposure to MF and cadmium
changed protein expression profile, and the differentially expressed proteins were enriched for
signal transduction, protein translation and stress related pathways. Our data suggest that MF
exposure enhanced cell viability and attenuated cadmium inhibited cell viability may be through
changes of proteome profiling, while the role of differentially expressed proteins responding to MF
awaits further investigations.
Key words: 50 Hz magnetic fields, JAR, JEG-3, cell viability, Mass spectrometry
Figures

124
Figure 1. Effects of exposure to 50 Hz MF at 3.0 mT for different durations on cell
viability in JAR (A) and JEG-3 (B) cells. Values are Mean ± SEM from 5 independent
experiments (Students’t test).

125
Figure 2. Effects of co-exposure with 3.0 mT of 50 Hz MF and 2.5 μM Cadmium for up to
48 hours on cell viability of JAR (A) and JEG-3 (B) cells. Values are Mean ± SEM from 5
independent experiments (Students’ t test).

126
Figure 3. Differentially expressed proteins in JAR cells treated with 3.0 mT, 50 Hz
magnetic field and/or with 2.5 μM Cadmium. S: Sham exposure group; E: MF exposure
group; S+Cd: Cadmium exposure group; E+Cd: Cadmium and MF co-exposure group.

127
S06-4 [11:45]
Microwave electric field at 2.45 GHz modulates the β-adrenergic response of human
embryonic stem cell-derived cardiomyocytes
Catrin Williams1, Catherine Hather2, Adrian Porch1, Heungjae Choi1, Rhodri Williams3, David Lloyd1, 4 &
Christopher George5
1Centre for High Frequency Engineering, Cardiff University, Cardiff, United Kingdom, CF24 3AA
2ReNeuron, Bridgend, United Kingdom, CF35 5HY
3Cardiff University, Cardiff, United Kingdom, CF10 3NB
4Cardiff University, Cardiff, United Kingdom, CF10 3AT
5Institute of Life Sciences, Swansea University, Swansea, United Kingdom, SA2 8PP
Keywords: In vitro, RF/Microwaves, Completed (unpublished)
Presented by: Catrin Williams
Human embryonic stem cell-derived cardiomyocytes were exposed to pulsed microwave electric
field (5.47 kV/m, 10 ms pulses at 50% duty cycle for 60 s, Tmax=37°C) in a TM010 mode resonant
cavity. This had a long-lasting effect on intracellular calcium signalling, caspase-3 and -7 activity and
apoptotic cell death, as compared to a conventional heating control. Membrane poration did not play
a role in the non-thermal mechanism of action. Future work will investigate the effects of non-
thermal microwave electric field on voltage-gated calcium channels.
Background
Microwave frequency electromagnetic fields are ubiquitous in our modern, urban environment. This
work investigates the thermal and non-thermal biological effects of microwaves used in mobile/ Wi-
Fi communication as well as the potential therapeutic use in tumour ablation and the treatment of
cardiac arrhythmias.
Methods
Human embryonic stem cell-derived cardiomyocytes (CM) were exposed to pulsed microwave
electric field (MEF) at 2.45 GHz using a TM010 mode resonant cavity. This corresponded to a
maximum field strength of 5.47 kV/m. The maximum temperature obtained by MEF heating was
restricted to 37°C, using pulsed MEF (10 ms, 50% duty cycle) for a period of approximately 60 s. As
a control, separate CM populations were heated conventionally to match the degree of microwave
heating. After a 4 hour recovery period, calcium signals in fluo-4-loaded microwave-exposed or
control CM were visualised using confocal microscopy. The patterning of calcium signals under
basal conditions and following the addition of a β-adrenergic receptor agonist, isoproterenol (ISO; 30
nM), were profiled using SALVO software. The extent of CM apoptosis was determined using
caspase (Caspase Glo-3/7 assay, Promega) and TUNEL (DeadEnd TUNEL assay, Promega)
assays. MEF-induced membrane permeabilization was investigated by looking at the influx of
fluorescently-labelled dextrans (10 and 70 kDa) into CM after exposure.
Results
MEF did not affect basal calcium cycling in CM exposed to MEF or control heating (n= 41 separate
populations). However, in ISO-treated CM, MEF increased calcium spike amplitude, area and the
rate of calcium release. MEF also markedly decreased signal variability between calcium spikes in
ISO-treated CM. Significant differences between MEF-treated and conventional heating controls
were observed for caspase-3 and -7 activity and apoptotic cell death. No infiltration of fluorescent
dextrans was observed after exposure to MEF.
Conclusions
The microwave treatment regime investigated in this study affects the calcium handling fingerprint of
ISO-treated CM. This also led to significant differences in apoptosis, as compared to the

128
conventional heating control. These potentially non-thermal effects could not be explained by
membrane poration. Future work will investigate the role of voltage-gated calcium channels in the
non-thermal mechanism of action of MEF.

S06-5 [12:00]
Study on the non-thermal effects of exposing cells to 0.07-0.3 THz using a widely tunable
source
Noriko Yaekashiwa1, Sato Otsuki1, Shin’ichiro Hayashi1, 2 & Kodo Kawase1, 3
1RIKEN, Sendai, Japan, 980-0845
2National Institute of Information and Communications Technology, Koganei, Japan, 184-8795
3Nagoya University, Nagoya, Japan, 464-8603
Keywords: In vitro, THz, Work in Progress
Presented by: Noriko Yaekashiwa
This study investigated the effects of Terahertz (THz) wave irradiation on the cell activity of normal
human fibroblast (NTI-4) cells. Our research examined non-thermal effects on cells exposed to
radiation at low power (below 10 μW) with tunable frequencies from 0.07 THz to 0.3 THz. Cells were
exposed to THz wave for either 70 or 94 h. We found no difference in cell proliferation and cellular
activity between THz-exposed cells and sham cells.
Introduction
Some studies have been carried out on the biological effects of terahertz wave (THz) irradiation
[1–6]. However very few studies have focused on low-power, long-term irradiation of cells with a
widely tunable source [7]. The European Commission’s Scientific Committee on Emerging and
Newly Identified Health Risks (SCENIHR) reported recent research trends and found that few
studies have examined the biological effects of the THz fields. And the SCENIHR recommended
new studies into the effects of exposure of skin (low-level, long-term exposure) and eyes (high-level,
short-term exposure) to radiation [8].
In addition, Fröhlich proposed resonance vibrations in the cellular phospholipid bilayer membrane at
frequencies ranging from 0.1 THz to 1 THz [9]. This hypothesis was based on the expectation that
irradiation at weak intensities might cause some kind of non-thermal effect.
The aim of our study was to examine non-thermal effects of cells exposed to 0.07-0.3 THz at low
power using widely tunable source.
Materials and Methods
THz wave source
As THz source variable over wide frequencies, we used the Uni-Traveling Carrier Photodiode (UTC-
PD, NTT Electronics Co., Japan). Using photomixing of two wavelength laser lights on the
photodiode, the UTC-PD can generate THz wave in the range of 0.1 THz to 1 THz [10]. To avoid the
thermal effects of irradiation, we used a THz wave power source of 10 μW or less. We irradiated the
cells for 94h or 70 h with frequencies sweeping (231 sweeping steps) from 0.07 THz to 0.3 THz,
with a 1 GHz step-up every 24.17 min or 18 min, respectively.
Cells used for exposure tests
We used normal human fibroblast, NTI-4 (RIKEN BioResource Center, Japan) in this study. DMEM-
low glucose (Sigma-Aldrich, Japan) with ten percent fetal bovine serum (Thermo Fisher Scientific
K.K., Japan) was used as the medium for the NTI-4 cells. Cells were maintained in an incubator set
at 37°C with 5% CO2 in a saturated vapor state.
Exposure apparatus for Cell proliferation measurement
As the exposure apparatus for cell proliferation measurements, the head of the wave source UTC-
PD was inserted into an aluminum pipe and set under the cells to be exposed. To avoid exposure of

129
other cells, the bottoms of wells containing unexposed cells were covered in aluminum foil to block
the cells from THz wave. This exposure apparatus was set in the incubator.
The alternating current (AC) impedance method (BM2401; HIOKI E.E.Co. Japan) was used to
measure cell proliferation. For this method, we used a dedicated measurement well (bottom area of
0.785 cm2), to the bottom of which was fixed an ultra-thin indium tin oxide (ITO) membrane
electrode on quartz glass which shows transmittance of approximately 10% for THz wave. Cells that
adhere to and grow on the ITO electrode impede current flow, changing the impedance value and
providing the curve of cells growth for 94 h.
Exposure apparatus for Cell activity measurement
The exposure apparatus for cell activity measurements was set up inside the incubator. A 96-well
plate (Corning, 3596, USA) was used to measure cell activity. The THz wave transmittance of the
wells was 90%. To compare the effects on exposed and sham cells, these two sets of cells were
kept in individual, identical incubators.
We used a novel colorimetric component, CellTiter96® AQueous One Solution Cell Proliferation
Assay (Promega Co., USA) to investigate the activity of the cells exposed to the THz wave. We
seeded 4000 cells/well in a 96-well plate and cultured the cells for 2 h. After exposure for 70 h,
assay reagent was added directly to all cells and incubated for 3 h. We used iMark microplate
reader (BIO-RAD, Canada) and measured the absorbance at 490 nm.
Results and Discussion
The results of the cell proliferation for 94 h with frequencies ranging from 0.07 THz to 0.30 THz did
not differ from that of unexposed cells. Seventy hours of irradiation did not significantly change the
activity rates of NTI-4 cell. Positive control cells were cultured for 70 h in an incubator chamber
heated to 42°C. Compared with sham cells, the activity of positive control NTI-4 cells declined by
50% (P < 0.05).
We irradiated cultured cells long-term at a low power, in order to investigate non-thermal effects for
THz fields. We found no difference in the cell proliferation and cell activity between irradiated cells of
94h or 70 h and those of sham cells. Our study found no non-thermal effect as a result of exposure
of cells to 0.07 THz to 0.3 THz of radiation.
Acknowledgement
We thank to Ms. S. Kato for experimental assistance. This work was supported by a Grant-in-Aid
from the Ministry of Internal Affairs and Communications (MIC) in Japan.
References
[1] Wilmink GJ, Grundt JE. Invited Review Article: Current state of research on biological effects of
terahertz radiation. J Infrared Milli Terahz Waves 2011;32:1074–122.
[2] Korenstein-Ilan A, Barbul A, Hasin P et al. Terahertz radiation increases genomic instability in
human lymphocytes. Radiat Res 2008;170:224–34.
[3] Hintzsche H, Jastrow C, Kleine-Ostmann T et al. Terahertz radiation induces spindle
disturbances in human-hamster hybrid cells. Radiat Res 2011;175:569-74.
[4] Hintzsche H, Jastrow C, Kleine-Ostmann T, et al. Terahertz Electromagnetic Fields (0.106 THz)
do not induce manifest genomic damage in vitro. PLoS ONE 2012;7:e46397.
[5] Bourne N, Clothier RH, D’Arienzo M et al. The effects of terahertz radiation on human
keratinocyte primary cultures and neural cell cultures. ATLA 2008;36:667-84.
[6] De Amicis A, Sanctis SD, Cristofaro SD et al. Biological effects of in vitro THz radiation exposure
in human foetal fibroblasts. Mutat Res Genet Toxicol Environ Mutagen 2015;793:150-60.
[7] Yaekashiwa, N., Otsuki, S., Hayashi, S., & Kawase, K. (2017). Investigation on the non-thermal
effects of exposing cells to 70-300 GHz using a widely tunable source. Journal of Radiation
Research. P.1-6. DOI:10.1093/jrr/rrx075.

130
[8] Scientific Committee on Emerging and Newly Identified Health Risks (SCENIHR). Final opinion
on potential health effects of exposure to electromagnetic fields (EMF). 2015 Jan.
https://ec.europa.eu/health/scientific_committees/emerging/docs/scenihr_o_041.pdf2015 Jan. (3
October 2017, date last accessed).
[9] Fröhlich H. Long-range coherence and energy storage in biological systems. Int J Quant Chem 2
1968;641-9.
[10] Ito H, Nagatsuma T, Ishibashi T. Uni-traveling-carrier photodiodes for high-speed detection and
broadband sensing. Proc SPIE 2007;6479:64790X.

S06-6 [12:15]
Effects of combined exposure of 0.3 THz with ionomycin on cytokine production in human
keratinocyte cells
Shin Koyama1, Eijiro Narita1, Yoko Shimizu1, Kensuke Sasaki2, Maya Mizuno2, Soichi Watanabe2, Masao
Taki3, Naoki Shinohara1 & Junji Miyakoshi1
1Laboratory of Applied Radio Engineering for Humanosphere, Research Institute for Sustainable
Humanosphere, Kyoto University, Uji, Japan, 611-0011
2Electromagnetic Compatibility Laboratory, National Institute of Information and Communications Technology,
Koganei, Japan, 184-8795
3Department of Electrical & Electronic Engineering, Tokyo Metropolitan University, Hachioji, Japan, 192-0397
Keywords: In vitro, THz, Work in Progress
Presented by: Shin Koyama
The cellular effects of 24-hour combined exposure to 0.3 terahertz (THz) with ionomycin in human
keratinocyte (HaCaT) cells were investigated. There was no significant increase on three kinds of
cytokine production in the cells exposed to 0.3 THz with ionomycin compared with the ionomycin-
treated cells alone. This result indicates that the combined exposure to 0.3 THz with ionomycin
would be considered to have no or very little effect on the three kinds of cytokine production in
HaCaT cells.
Introduction
In recent years, THz waves, which are the frequency region from 0.1-10 THz, have been used for
many kinds of technologies. The THz region had been underused due to the limited availability of
sources and detectors, however the recent development of various THz sources enabled to detect
some kinds of diseases or to scan for weapons, explosives and biohazards. These rapid
introduction of THz devices leads some public concern about possible adverse effect on human
health. Previously it has been reported that the exposure to THz increased genomic instability in
human lymphocytes [1]. On the other hand, Scarfì et al. indicated that the exposure to THz has not
induced direct chromosomal damage and alteration of cell cycle kinetics [2]. Then it is necessary to
evaluate the influence on human body by the low-level and long-term exposure to THz. To
investigate the athermal effects of THz, we manufactured a device which can expose 0.3 THz to
cells. In this study, we examined the production of cytokines, in here IL-1α, -6, and -8 in keratinocyte
(HaCaT) cells by the combined exposure to 0.3 THz with ionomycin.
Materials and Methods
We used a specially designed exposure apparatus which can expose 0.3 THz at the density of
about 0.5 mW/cm2. The system was developed by the research group of Tokyo Metropolitan
University. The cavity in the exposure system was maintained under controlled conditions similar to
those in a conventional incubator, i.e., an atmosphere of 95% air and 5% CO2at a relative humidity
of >95% and a temperature of 37°C. A culture dish of 3.5 mm in diameter was put on the top of
apparatus. Spatially averaged power density was set for about 0.5 mW/cm2at the bottom of culture
dish. Temperature elevation was suppressed less than 0.1ºC for that condition. HaCaT cells were

131
maintained in DMEM medium supplemented with 5% fetal bovine serum. After 24-hour exposure,
the supernatant was collected and the cells were lysed with CelLyticTM (SIGMA). The cells were
treated with or without 1 μg/ml of ionomycin and exposed to 0.3 THz.
The measurement of intercellular and extracellular cytokines was performed using ELISA kit
according to the manufacture’s instruction. Absorbance was measured at 450 nm with a plate
reader and data were analyzed using standard curve.
Results and Discussion
The results of combined exposure of 0.3 THz and ionomycin are shown in Figure 1. The cytokine
productions were increased with ionomycin treatment alone significantly. However, no significant
difference was observed between combined exposure of 0.3 THz with ionomycin and ionomycin
treatment alone.
Conclusions
The data observed in the present study suggest that the combined exposure to 0.3 THz with
ionomycin for 24 hours might have no significant effect of the cytokine production in HaCaT cells. In
conclusions, we do not consider that the exposure to 0.3 THz could enhance any significant effects
on cytokine production in cultured human keratinocyte cells under our experimental conditions.
Acknowledgments
This work has supported in part by a Grant from the Ministry of Internal Affairs and Communications,
Japan and in part by the Research Institute for Sustainable Humanosphere, Kyoto University
(Mission 5-1).
References
[1] A. Korenstein-Ilan, A. Barbul, P. Hasin, A. Eliran, A. Gover, and R. Korenstein, “Terahertz
Radiation Increases Genomic Instability in Human Lymphocytes,” Radiat. Res., 2008, 170: 224-234.
[2] M.R. Scarfì, M. Romanò, R. Di Pietro, O. Zeni, A. Doria, G.P. Gallerano, E. Giovenale, G.
Messina, A. Lai, G. Campurra, D. Coniglio, M. D'Arienzo, “THz Exposure of Whole Blood for the
Study of Biological Effects on Human Lymphocytes,” J. Biol. Physics, 2003, 29: 171–177.
Figures

Figure 1. The IL-1α, IL-6 and IL-8 production in cell lysate or supernatant exposed to 0.3 THz
with or without ionomycin (1 µg/ml) for 24 hours on HaCaT cells. Data are presented as the
mean ± SD from three independent experiments. **p< 0.01.

132
 Session: S07
Electroporation
Tuesday June 26, 2018 • 16:30 - 18:00
Europa D
Chairs: Olga Pakhomova & Richard Nuccitelli

S07-1 [16:30]
STUDENT PAPER
Guidelines for data collection and follow-up reporting for early stage health technology
assessment of electrochemotherapy of skin melanoma and basal-cell carcinoma
Eva Pirc1, Leandro Pecchia2, Matej Reberšek1, Gregor Serša3, Marko Snoj4, Aleš Grošelj5 & Damijan
Miklavčič1
1Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia, 1000
2School of Engineering, University of Warwick, Coventry, United Kingdom, CV4 7AL
3Department of Experimental Oncology, Institute of Oncology Ljubljana,, Ljubljana, Slovenia, 1000
4Faculty of Medicine, University of Ljubljana, Ljubljana, Slovenia, 1000
5Department of Otorhinolaryngology and Cervicofacial Surgery, University Medical Centre Ljubljana,
Ljubljana, Slovenia, 1000
Keywords: Electroporation, Pulsed, Completed (published)
Presented by: Eva Pirc
In this paper, two Markov models are presented that will be used for cost-effectiveness analysis of
electrochemotherapy for treatment of basal cell carcinoma and skin melanoma. Few cost-
effectiveness analyses of electrochemotherapy have already been done, but with a lack of
information about the intervention effect on the quality of life of the treated patients, which may
result in inaccurate or even inadequate conclusions. For this reason, guidelines for data collection
and follow-up reporting are suggested in this paper. A simple table is presented that will simplify the
data collection process. More realistic cost-effectiveness analysis will most likely enable easier
equipment purchase and clinical practice implementation.
I. Introduction
Electroporation is a platform technology which is already well established in medicine, food
processing and gene transfection [1]. Basic mechanisms of electroporation are still unknown, but it
is clear that due to an application of high voltage electric pulses, cell membrane becomes
permeable to molecules, which otherwise cannot cross it. One of the most successful
electroporation-based applications is an antitumor therapy called electrochemotherapy (ECT).
Through the permeabilized cell membrane transport of chemotherapeutic drug bleomycin or
cisplatin is increased and considerably higher cytotoxicity is reached [2]. Currently, ECT is mainly
used in the treatment of cutaneous and subcutaneous tumors, following Standard Operating
Procedures (SOP) [3]. Because clinical use of electroporation in medicine is in a full swing, the
treatment needs a Health Technology Assessment (HTA) in order to justify its effectiveness. In the
frame of HTA, the cost-effectiveness of ECT will be evaluated for the treatment of skin melanoma
and Basal Cell Carcinoma (BCC). These two cancer types were chosen, because of high response
rate when treated with ECT. Current treatment of choice for skin melanoma is surgery and for BCC
in addition to surgery also radiotherapy and topical or injectable delivery of drugs, such as
imiquimod and 5-Fluorouracil (5-FU) and also photodynamic therapy are available [4]. When, in
case of skin melanoma, numerous or extensive lesions are present, the patient is not amenable to
surgery, therefore isolated limb perfusion is used in this case. It has already been shown that eECT
is much less expensive than isolated limb perfusion [5] and it may even be more effective at the
same time [6]. ECT is, for now, available only to stage III patients, whose 5 to 10 years survival rate
is between 30 and 40 % [7].

133
In this paper, an approach to cost-effectiveness analysis (CEA) for emerging application is
presented. Because cancer is a recurrent disease, Markov modeling strategy was chosen. Two
Markov models for skin melanoma and BCC were developed. . Each state in a Markov model
represents a specific state of health, or stage of disease, between which patients migrate with time.
There is always one state that represents death, which is considered an absorbing state as it does
not have a return path. Each state is associated with an average quality of life, measured with QALY
(obtained with EQ-5D questioners), it is a reflection of quality of life of patients in this health status,
and a cost that is reflecting the cost needed to maintain the patient in this state (e.g., avoid
deteriorations) and the probability to move from the current health state to another (in the model
presented with letter P), or to stay in the same state in the following period of time (usually one year,
but in this study three months) [8].
II. Methods and materials
As the structure of the model should always reflect the essential features of the disease and its
interventions, models should be adopted for each cancer type. Developed general model for skin
melanoma (Figure1) has eight states and a general model for BCC has seven states (Figure 2). For
the initial analysis models were simplified/reduced (Figure 1, 2), some states were aggregated in
order to facilitate the calculations and focus only on states that are relevant for the new technology
(i.e. where ECT is currently used).
Patients enter both general models in a state without cancer, however, these two states will be
eliminated at the initial calculations because they are not relevant for the new technology – ECT, we
will only focus on the population with specific cancer type. Following states in case of skin
melanoma represent different cancer staging, such as: primary tumor; distant metastases;
metastases in lymph nodes; distant metastases with predominantly cutaneous symptoms and
distant metastases without predominantly cutaneous symptoms. Because all patients are not
diagnosed with cancer at its initial state, some might enter the model with the disease that combines
several states in the model. Clinical stages of skin melanoma are defined as follows: at stages I and
II the patients only have a primary tumor; stage III a, b means that a patient has a primary tumor and
metastases at lymph nodes; stage III c combines primary tumor, local metastases, and metastases
with predominantly cutaneous symptoms. The last and most severe stage IV includes primary tumor
and distant metastases [9]. ECT is currently used in most cases in stage IIIc and IV patients and in
case of BCC, it is used only on patients with bigger tumors and their recurrence [10, 11], therefore a
reduced - initial models for initial calculations, only have aggregated states that are relevant for
those two cancer types. Aggregated BCC model is really simple it only has three possible outcomes:
complete response, residual disease and progressive disease. Both general models also have the
same two absorbing states, one is death due to cancer and the other is death from other causes, for
initial calculations this two states will be merged to one dead state. The probability of staying in dead
state is always equal to one and the cost is always zero. Because it is reported that significant
changes in quality of life of patients treated with ECT are detected already few months after the
treatment [12], the time step of our study will be three months.
III. Data collection guidelines
In this paper, two general Markov models for a cost-effectiveness analysis (CEA) were developed,
but not yet used in the simulations because for CEA, data on quality of life improvement on a time
step scale, in our case three months are needed, which is a big challenge. Researchers are
currently only reporting that quality of life is improved after the treatment, but the numerical data that
are crucial for unbiased CEA are missing [13, 14, 15]. Therefore, for now, only data needed for
evaluation of the cost-effectiveness of ECT treatment of skin melanoma and BCC is defined.
Additionally, to facilitate initial data collection, recommendations for data collection process and
follow-up reporting are suggested. A simple template for reporting (Figure 3) is suggested. For each
patient, every three months, the clinician is supposed to:
define all medical interventions a patient experienced in last period;
evaluate patient state, with the corresponding number in the initial model;

134
 obtain EQ-5D questioner and report numerical value;
record all medical interventions or examinations that patient had in the last period, including type
of anesthesia during ECT, days spend in hospital, biopsy, blood analysis and radiological
assessment;
report about all drugs prescribed to the patient, that are related to his/her condition; also for
management of side effects, analgetics, wound care dressing…
report any other potential costs.
IV. Conclusion
ECT, as an emerging technology, is not jet used on all patients, but as the proof of cost-
effectiveness rises, it is slowly moving up as the treatment of choice. More realistic cost-
effectiveness analysis will give us a better insight into electroporation cost and most likely enable
easier equipment purchase and clinical practice implementation. Additionally, the prediction of cost-
effectiveness can also influence the next round of fund-raising. A final evaluation will be done when
we obtain a sufficient amount of numeric data about the QoL improvement. Nevertheless, we
believe that a two-year data collection will give us a representative pattern. New models, for specific
cancer type, should be developed for each disease separately and data collection should be done
simultaneously. With this paper, we would like to emphasize the importance of EQ-5D questioners
and numerical reporting of QoL improvement.
V. Acknowledgment
This study was supported by COST Action BM1309 (COST EMF-MED) within the STSM (ECOST-
STSM-BM1309-110416-073660) and the Slovenian Research Agency (ARRS) (MRIC UL IP-0510,
P2-0249), conducted within the scope of the European Associated Laboratory on the
Electroporation in Biology and Medicine (LEA- EBAM).
VI. Literature
[1] S. Haberl, D. Miklavcic, G. Sersa, W. Frey, and B. Rubinsky, ‘Cell membrane electroporation-Part
2: the applications’, IEEE Electrical Insulation Magazine, vol. 29, no. 1, pp. 29–37, Jan. 2013.
[2] D. Miklavčič, B. Mali, B. Kos, R. Heller, and G. Serša, ‘Electrochemotherapy: from the drawing
board into medical practice’, BioMedical Engineering OnLine, vol. 13, p. 29, 2014.
[3] L. M. Mir et al., ‘Standard operating procedures of the electrochemotherapy: Instructions for the
use of bleomycin or cisplatin administered either systemically or locally and electric pulses delivered
by the CliniporatorTM by means of invasive or non-invasive electrodes’, EJC Supplements, vol. 4,
no. 11, pp. 14–25, Nov. 2006.
[4] A. I. Rubin, E. H. Chen, and D. Ratner, ‘Basal-Cell Carcinoma’, New England Journal of
Medicine, vol. 353, no. 21, pp. 2262–2269, Nov. 2005.
[5] G. L. Colombo, S. D. Matteo, and L. M. Mir, ‘Cost-effectiveness analysis of electrochemotherapy
with the Cliniporatortrade mark vs other methods for the control and treatment of cutaneous and
subcutaneous tumors’, Ther Clin Risk Manag, vol. 4, no. 2, pp. 541–548, Apr. 2008.
[6] D. E. Spratt et al., ‘Efficacy of skin-directed therapy for cutaneous metastases from advanced
cancer: a meta-analysis’, J. Clin. Oncol., vol. 32, no. 28, pp. 3144–3155, Oct. 2014.
[7] C. Garbe et al., ‘Diagnosis and treatment of melanoma. European consensus-based
interdisciplinary guideline – Update 2016’, European Journal of Cancer, vol. 63, pp. 201–217, Aug.
2016.
[8] Weinstein, Milton C., Bernie OBrien, John Hornberger, Joseph Jackson, Magnus Johannesson,
Chris McCabe, and Bryan R. Luce. Principles of Good Practice for Decision Analytic Modeling in
Health-Care Evalua- tion: Report of the ISPOR Task Force on Good Research PracticesMod-eling
Studies. Value in Health 6, no. 1 (January 1, 2003): 917.
[9] Amin, M.B., Edge, S., Greene, F., Byrd, D.R., Brookland, R.K., Washington, M.K., Gershenwald,
J.E., Compton, C.C., Hess, K.R., Sullivan, D.C., Jessup, J.M., Brierley, J.D., Gaspar, L.E., Schilsky,

135
R.L., Balch, C.M., Winchester, D.P., Asare, E.A., Madera, M., Gress, D.M., Meyer, L.R. (Eds.).
“AJCC Cancer Staging Manual”, Springer 2017
[10] Clover A.J.P., J. O’Mahony and D. Soden. ”Electrochemotherpy of Basal Cell Carcinoma”,
Handbook of Electroporation, Ed.: D. Miklavčič, Springer International Publishing Switzerland, vol.1,
2017, DOI: 10.1007/978-3-319-26779-1_97-1
[11] Campana L.G., M. Sepulcri, S. Valpione, L. Corti and C. R. Rossi. ”Electrochemotherpy for
Superficially Metastatic Melanoma.”, in Handbook of Electroporation, Ed.: D. Miklavčič, Springer
International Publishing Switzerland, vol.1, 2017.
[12] Bianchi, Giuseppe, Laura Campanacci, Mattia Ronchetti, and Davide Donati.
“Electrochemotherapy in the Treatment of Bone Metastases: A Phase II Trial.” World Journal of
Surgery 40, no. 12 (2016): 3088–94.
[13] Lido, Paolo, Giovanni Paolino, Andrea Feliziani, Letizia Santurro, Mauro Montuori, Flavio de
Sanctis, Piero Rossi, et al. “Cutaneous Metastasis of Unknown Primary Presenting as Massive and
Invasive Abdominal Lesion: An Elective Approach with Electrochemotherapy.” Anais Brasileiros De
Dermatologia 90, no. 6 (December 2015): 879–82.
[14] Larkin, John O., Christopher G. Collins, Simon Aarons, Mark Tangney, Maria Whelan, Seamus
O’Reily, Oscar Breathnach, Declan M. Soden, and Gerald C. O’Sullivan. “Electrochemotherapy.”
Annals of Surgery 245, no. 3 (March 2007): 469–79.
[15] Serša, Gregor, Tanja Cufer, Snezna Marija Paulin, Maja Čemazar, and Marko Snoj.
“Electrochemotherapy of Chest Wall Breast Cancer Recurrence.” Cancer Treatment Reviews 38, no.
5 (August 2012): 379–86.
Figures

Figure 1. A general Markov model for skin melanoma for the initial analysis model is simplified
in order to facilitate the calculations and focus only on states that are relevant for the new
technology –ECT. Only bold states are kept and two death states are aggregated into one
dead state. Therefore the model for initial calculations only has three states.

136
Figure 2. A general Markov model for BCC for the initial analysis model is simplified in order to
facilitate the calculations and focus only on states that are relevant for the new technology
–ECT. Only bold states are kept and two death states are aggregated into one dead state.
Therefore the model for initial calculations only has five states.

Figure 3. Guidelines for data collection. Data needed for evaluation of the cost-effectiveness of
electrochemotherapy treatment of skin melanoma and BCC is defined. To easily facilitate
those data, template for data collection process and follow-up reporting is suggested.

137
S07-2 [16:45]
Point source electroporation for local, rapid and minimal invasive treatment of brain tumors
Shirley Sharabi2, David Last2, David Guez2, Dianne Daniels2, Sharona Salomon2, Muhammad Hjouj3 & Yael
Mardor1, 2
1Medical science, Tel Aviv university, Tel Aviv, Israel
2Advanced Technology Center, Sheba medical center, Ramat Gan, Israel, 52295
3medical imaging department, Alquds University, Ramat GanAlquds, Palestinian Territory
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Shirley Sharabi
Existing treatments offer poor prognosis for glioblastoma multiforme patients. We have developed a
unique minimally-invasive setup for treating brain tumors employing a single intracranial electrode
placed within the tumor and an external surface electrode. This setup provides intratumoral
irreversible-electroporation with surrounding blood-brain-barrier disruption, enabling efficient delivery
of systemically administered drug to the residual tumor and surrounding infiltrating zone. An efficacy
study conducted with this setup and systemic Cisplatin on glioma bearing rats demonstrated
significantly increased survival rates, decreased tumor growth rates and significant activation of the
immune system compared to control groups.
Introduction:
In spite of aggressive therapy, existing treatments offer poor prognosis for glioblastoma multiforme
patients due to tumor infiltration into the surrounding brain, the ability to evade and suppress the
immune system and poor blood-brain barrier (BBB) penetration of most therapeutic agents.
When applied to the brain, electroporation can induce transient BBB disruption in volumes extending
beyond irreversible-electroporation. We have developed a unique minimally-invasive setup for
treating brain tumors employing a single insulated intracranial electrode with an exposed tip placed
within the tumor and an external surface electrode (Fig 1). We term this setup point-source
electroporation. The setup provides intratumoral irreversible-electroporation with surrounding BBB
disruption, enabling efficient delivery of systemically administered drug to the surrounding infiltrating
zone.
Methods:
A finite element model simulation program was written to determine optimal electrodes design,
setup and treatment parameters for rats studies. The electrodes setup was constructed accordingly.
Electroporation-induced damage and BBB disruption volumes were studied in naive rats brains
using MRI and histology. The efficacy of electroporation combined with/without systemically
administered Methotrexate and Cisplatin was studied in rats bearing intracranial glioma. The effect
of the treatment on the activation of the immune system was studied as well.
Results:
The average volumes of permanent tissue damage calculated from T2-weighted MRI acquired 2
weeks post treastment were significantly smaller than those of BBB disruption calculated from
contrast-enhanced T1-weighted MRI (p<0.0001), acquired immediately post treatment, in naïve rats.
Significant correlation (r2=0.94 ; p<0.0002) was found between BBB disruption and permanent
damage volumes. BBB disruption was visible for at least 24 hours post treatment although the
volume gradually decreased.
The efficacy study resulted in suppressed tumor growth rates in the combined electroporation-
Cisplatin group (1.1±0.1) relative to the growth rates in the control group p<0.047 (5.2±1.0) (Fig 2)
and in the Cisplatin-only group p<0.012 (3.92±1.0) (Welch’s F(2, 12.73)=10.84; p<0.002; ω2=0.28).
There was no significant difference between tumor growth rates in the control and Cisplatin-only

138
groups. Kaplan-Meir analysis revealed that the electroporation-Cisplatin combination prolonged
survival significantly (χ2=7.54; p<0.006).
Immunofluorescence analysis revealed significant infiltration of activated microglia, peripheral
macrophages and CD8+ cells in the remaining tumor as well as activated astrocytes forming glial
scaring surrounding electroporation-induced necrotic regions (Fig 3).
In order to assess the clinical application of our point-source setup a human brain model was
implemented in the finite elements simulation, demonstrating the feasibility for obtaining clinically-
relevant treatment volumes (~6 cm diameter) using a single 3 mm (diameter) intracranial electrode
with a 2 cm exposed tip and an external surface electrode.
Conclusions:
Here we present the application of the point-source approach, combining irreversible electroporation
in the tumor mass with surrounding reversible BBB disruption and systemic chemotherapy for the
treatment of brain tumors. Our results demonstrate the advantages of applying our minimal invasive,
point-source setup for obtaining significant anti-neoplastic effects in a brain tumor model. Finally, the
simulation program demonstrated the feasibility of extending the rat setup to a clinical application.
Figures

Figure 1. Point source setup. A) schematic illustration of the point sourse electrode setup,
consisting of a single intracravial electrode, insulated exept for a 1mm exposed tip and a
surface ground electrode pressed against the rat chest using conductive gel. B) 2D finite
elements geometrical model for the point sourse setup. C) Electical field distribution induced
by the point sourse setup. The electrical field is strongest near the electrode tip and tappers
down with distance in a spherical manner

139
 Figure 2. Examples of tumor growth in treated versus control rats. Top row: Tumor of a control
rat as depicted by contrast-enhanced T1-weighted MRIs at different time points. Bottom row:
Tumor of a rat treated with electroporation (600V, 50μs, 1Hz, 180 pulses) and Cisplatin at
different time points.

Figure 3. Infiltration of cytotoxic T-cells into the tumor for different treatments 1 week post
treatment. Sections were stained with CD8+ and Hoechst 33342 blue. X200. Left: Cisplatin
only. Middle: Electroporation only Right: Combined electroporation+Cisplatin. Lymphocytes
are mainly located on the irreversible electroporation-treated tumor border.

S07-3 [17:00]
Electroporation detection with THz time-domain spectroscopy
Stefania Romeo1, P. Thomas Vernier2 & Olga Zeni1
1Institute for Electromagnetic Sensing of the Environment (IREA), CNR, Napoli, Italy, 80124
2Frank Reidy Research Center for Bioelectrics), Old Dominion University, Norfolk, Virginia, USA, 23508
Keywords: Electroporation, THz, Completed (unpublished)
Presented by: Stefania Romeo
We report here a new procedure for THz time-domain spectroscopic analysis of mammalian cell
140
samples exposed to electroporating pulsed electric fields. Suspensions of human monocytes, Mono-
Mac 6 (MM-6), were exposed to 8, 100 μs, 400 V (100 kV/m between the cuvette electrodes) pulses
at 5 kHz, and then their spectral response between 0.2 and 1.5 THz was recorded with a
commercial THz time-domain spectrometer. Pulse-exposed cells absorb more THz radiation than
sham-exposed cells, suggesting that THz spectroscopy can be used as a powerful, label-free
method for characterization of electroporation.
Introduction
High-voltage, short-duration electric pulses increase the permeability of cell membranes to
otherwise impermeant or low-permeant molecules (electroporation, EP). Results from molecular
dynamics simulations suggest that electroporation begins with the reorganization of water and lipid
head group dipoles in the phospholipid bilayer interface, leading to the formation of defects that
evolve into water-filled, lipid pores [1]. The interior of the electroporated membrane thus contains
water, which should provide a signature for detection of the electropermeabilized state.
THz spectroscopy is a powerful and effective method for the characterization of the properties of
water in biological systems, to detect the hydration state of biomolecules and the dynamics of the
hydration process [2].
In this feasibility study, we use THz time-domain spectroscopy to detect electroporation in human
cells subjected to permeabilizing pulsed electric fields (PEFs).
Methods
Human monocytes, Mono-Mac 6 (MM-6), suspended in 200 μL of RPMI medium at a concentration
of 7.5 x 106 cells/mL were exposed/sham exposed to permeabilizing electric fields in standard, 4
mm EP cuvettes (Sigma Aldrich). A commercial pulse generator (Cliniporator™, IGEA, SpA, Italy)
was used to deliver 8, 100 μs, 400 V (100 kV/m between the cuvette electrodes) pulses at 5 kHz.
After exposure/sham exposure, cell samples were spun onto a mylar sheet (50 μm thick) by
cytocentrifugation, which was then mounted on a plastic holder for THz time-domain spectroscopy
measurements.
The fiber-coupled, THz, time-domain system (FICO, Zomega TeraHertz Corp., NY, USA) used for
the spectroscopic measurements consisting of three main components: control unit, emitter and
receiver heads, and laser source. This system generates and detects broadband THz pulses in the
0.1–3 THz range, both in transmission and reflection geometries. For the measurements, the
sample holder was placed between the emitter and the receiver heads in transmission configuration
(Figure 1). For each sample the time-domain pulse transmitted by air, mylar alone, and cell spot on
mylar (sham or exposed) was recorded as the average of 1000 acquisitions. A total of 10
independent measurements were carried out for each condition.
To quantify the interaction of THz radiation with the cell samples, the measured time-domain data
were converted to frequency domain with a fast Fourier transform. From the amplitude and phase
information obtained, the absorption coefficient, α(ν), and refractive index, n(ν), of each sample were
calculated as a function of frequency ν as from the following equations [3]:

where E and φ are, respectively, the amplitude and phase of the THz spectra, d is the sample
thickness, and T(ν) is the transmission coefficient at the air-sample interface.
Results
The absorption coefficient and refractive index of PEF-exposed and sham-exposed samples were
calculated in the frequency range between 0.2 and 1.5 THz. These parameters are a measure of the

141
absorption and dispersion characteristics of the sample. For both types of cell samples, the
absorption coefficient increases, and the refractive index decreases, with frequency. Both
parameters are greater for PEF-exposed cells than for sham-exposed ones, indicating a higher
absorption of the THz radiation by PEF-exposed samples. These results demonstrate that THz
spectroscopy can be used for the detection of changes in cell properties resulting from EP.
Discussion
The spectral changes observed in electroporated cells are consistent with those reported for
hydrated biological materials, and with the dynamics of the hydration of the various molecular
species present in cell membranes. A higher absorption of the THz radiation is associated to a more
hydrated state of the biological sample [4].
Electroporation-induced cell swelling in isosmotic media, a result of water influx driven by a colloid-
osmotic imbalance across the electropermeabilized plasma membrane, is commonly observed [5,
6]. Our interpretation, albeit provisional, is that the procedure reported here for the analysis of PEF-
exposed samples with THz spectroscopy can provide a qualitative indication of EP. Because of the
complexity represented by a suspension of electropermeabilized cells, new methods and
procedures need to be devised for a quantitative characterization of the changes in cellular
structures involved in the EP process.
References
[1] Vernier PT, Levine ZA. Water bridges in electropermeabilized phospholipid bilayers. Proceedings
of the IEEE. 2013;101(2):11.
[2] Heugen U, Schwaab G, Brundermann E, et al. Solute-induced retardation of water dynamics
probed directly by terahertz spectroscopy. P Natl Acad Sci USA. 2006;103(33):12301-12306.
[3] Naftaly M, Miles RE. Terahertz time-domain spectroscopy for material characterization.
Proceedings of the Ieee. 2007;95(8):1658-1665.
[4] N. Chopra, K. Yang, J. Upton, Q. H. Abbasi, K. Qaraqe, M. Philpott and A. Alomainy Fibroblasts
cell number density based human skin characterization at THz for in-body nanonetworks. Nano
Commun Netw 10:60-67(2016).
[5] Romeo S, Wu YH, Levine ZA, Gundersen MA, Vernier PT. Water influx and cell swelling after
nanosecond electropermeabilization. Biochimica et biophysica acta. 2013;1828(8):1715-1722.
[6] Nesin OM, Pakhomova ON, Xiao S, Pakhomov AG. Manipulation of cell volume and membrane
pore comparison following single cell permeabilization with 60- and 600-ns electric pulses.
Biochimica et biophysica acta. 2011;1808(3):792-801.

Figures

142
 Figure 1. THz time domain spectroscopy set up

S07-4 [17:15]
Imaging using terahertz radiation; modelling of melanoma with computational phantoms
Zoltan Vilagosh1, Alireza Lajevardipour1 & Andrew Wood1
1Department of Health and Medical Sciences, Swinburne University of Technology , Hawthorn, Australia,
3122
Keywords: Clinical (diagnostics), Static, Work in Progress
Presented by: Zoltan Vilagosh
The capacity to gauge skin hydration levels gives the terahertz band (THz) an imaging potential
different to Optical Coherence Tomography and Confocal microscopy. The capability to detect THz
is improving, but imaging needs to meet a clinical need. Finite-difference time-domain skin
computational phantoms aid the analysis of skin interaction with THz radiation. The presented
phantoms have accurate anatomical layering and electromagnetic properties. Simulations have
shown that the increased water content of melanomas and the freezing of the associated skin region
can produce penetration to a depth of 4 mm. No current modality can image skin to 4 mm, thus THz
imaging aids therapeutic decision making, such as the size of excision of melanoma.
Introduction
The development of a reliable computational phantom of the human skin is an invaluable tool in

143
imaging and assessing the interaction of terahertz frequency band (THz) radiation and skin for the
development of medical applications. The use of computational simulations accelerates the initial
scoping and the subsequent “proof of concept” phases of any proposal. From these early
simulations a base of evidence is built to recruit collaborators and to justify the expense of
equipment and clinical trials.
A flexible, accurate, computational phantom needs the capacity to be adjusted to take account of
temperature, hydration changes, anatomical features and variations in skin types. Computational
simulations are also a fast and efficient way of investigating potential confounding factors. Thus,
hydration gradients, the effect of hair follicles and sweat glands, and the variability of skin anatomy
can be tested both in the initial phase and modelled as problems arise during the development of
applications.
Melanoma and the “thickness problem”
One of the most pressing clinical difficulties that can be resolved by THz is the “melanoma thickness
problem”. According to Ferlay et al, (2015), 0.3% of the World’s population will have a melanoma at
some stage in their lives. The Australian Institute of Health and Welfare report (2017), notes 12,744
cases of melanoma in Australia in 2013, with 1,467 reported deaths in 2014. Thomas et al, in 2014
noted “The risk of death from cutaneous melanoma is determined mainly by the thickness of the
tumour”. There is a pressing need to accurately gauge melanoma thickness and subcutaneous
spread before surgery.
Various authors have associated prognosis of melanoma with its thickness. This was first studied by
Alexander Breslow in 1970. In 1975, Breslow gave the “poor prognosis” thickness at >1.5 mm. The
risk-thickness correlation was codified as the “Breslow thickness”. The current estimates give the
best prognosis at tumour thickness at <0.75-0.8 mm(T1) and the worst at >4.0 mm(T4) (Büttner et
al, 1995; Gershenwald et al, 2017). The alternate classification is the Clark Level, where the extent
of the melanoma is described according to the penetration through the levels of the skin (Clark et al,
1989) The surgical margins that are needed for safe excision vary according to the depth of
penetration and lateral spread, however the margin that should be applied for safe excision is
somewhat contested (Sladden et al 2009; Tchernev et al, 2017).
Cronin et al, (2018) reported that the margins recommended by the Irish national guidelines were
reached in only 74% of melanoma excisions.
Role of THz imaging in Melanoma
There are no current technologies that can resolve human skin structure to depths of 2 to 5 mm.
Figure 1 outlines the current alternate technologies and their depth resolution capability.
Figure 1 and caption.
Errors of excision make for increased risk of spread, needless morbidity and the need for repeated
procedures. Currently, the “definitive interpretation of melanoma margins requires a multidisciplinary
approach that includes proper pre-surgical examination of melanoma margins combined with
histologic, immunohistochemical and/or molecular analysis following surgery” (Danialan et al, 2012).
However, histological assessment is only as good as the sample provided; Orzan et al (2015) noted:
“choosing an inadequate biopsy technique can delay or create difficulties in establishing the
diagnosis or, in the worst case, can lead to a false negative result”.
THz electromagnetic radiation based imaging can, in principle, significantly add to the pre-surgical
assessment of melanomas, which would lead to a more accurate choice of margin for excision.
Melanomas contain 81-86% water (Rofstad et al. 1994), compared with 70-75 % in normal tissue.
Hydrated melanin is also a significant absorber of THz radiation, with a frequency absorption pattern
different from that of water.
THz radiation can image the human skin to a depth of only 0.3-0.5 mm at room temperature,
however, with the use of the routinely available carbon dioxide or liquid nitrogen skin freezing
techniques, imaging to a depth of 5 mm can be achieved.

144
THz Penetration of Frozen Skin
Over the range of 0.4 to 1.0 THz, the real permittivity (ε’) of water at 350 C changes from about 5.0
to 4.0, (Rønne and Keiding, 2002), whilst the ε’ of ice is between 3.06 and 3.18 for temperatures
-2330 to -330 C (200 to 2400 K), (Takeya et al (2014). The corresponding values for the imaginary
part of the permittivity (ε”) are 4.5 to 2.5 at 350 C and 0.002 to 0.1 at -2330 to -330 C. The difference
in the absorption coefficients and refractive indexes is tabulated in Figure 2
Figure 2 and caption
The properties of ice translate to an enhancement in the penetration of skin in the order of >1.0 x
107 at 4.0 mm depth. The difference between the absorption coefficient at 0.4 and 1.0 THz gives an
estimation of water content. THz studies of frozen brain tissue by Png et al (2009) at -280 C showed
no clear pattern between diseased and healthy material, but did show a consistent absorption
coefficient in the order of 500-1000 m-1 at 0.4 THz and 1500-2000 m-1 at 1.0 THz. Sim et al (2013)
reported studies of oral melanoma at 200 C and frozen at -200 C which showed the tumour having a
lower absorption coefficient than normal mucosa (closer to ice) at -200 C, suggesting that the higher
hydration rate was detectable by this technique.
Simulations of Frozen Skin
Simulations using a finite-difference time-domain solver (XFdtd) and skin computational phantoms
with accurate anatomical layering and electromagnetic properties have shown that hydration
changes and freezing the skin region can, in principle, produce in-vivo melanoma mapping to a
depth of 4 mm (Figure 3).
Figure 3 and caption
Melanin THz absorption content at 350 C is a significant contributor to the overall skin absorption
(Lipscomb et al, 2013), but the role of melanin in the frozen samples is not clear, and thus difficult to
model with any confidence.
Conclusion
The technique of freezing skin with carbon dioxide or liquid nitrogen has been employed for decades
to treat lesions such as solar keratoses. This tool can be deployed to increase the depth of
penetration of THz radiation. Given the changes in the complex THz permittivity of skin on freezing,
careful, multi-frequency measurements have a prospect of filling an important gap in the
management of melanoma, by providing information regarding excision margins before surgery. The
use of computational phantoms can speed the development of such a technology by providing some
of the preliminary scoping and proof of concept.
References
Australian Institute of Health and Welfare 2017. Cancer in Australia 2017. Cancer series no. 101.
Cat. no. CAN 100. Canberra: AIHW.
Breslow, A, 1970. Thickness, Cross-Sectional Areas and Depth of Invasion in the Prognosis of
Cutaneous Melanoma, Annals of Surgery. 172 (5): 902–8
Breslow, A, 1975. Tumor thickness, level of invasion and node dissection in stage I cutaneous
melanoma. Annals of surgery, 182(5), p.572.
Büttner, P., Garbe, C., Bertz, J., Burg, G., d'Hoedt, B., Drepper, H., Guggenmoos‐Holzmann, I.,
Lechner, W., Lippold, A., Orfanos, C.E. and Peters, A., 1995. Primary cutaneous melanoma.
Optimized cutoff points of tumor thickness and importance of Clark's level for prognostic
classification. Cancer, 75(10), pp.2499-2506.
Clark Jr, W.H., Elder, D.E., Guerry IV, D., Braitman, L.E., Trock, B.J., Schultz, D., Synnestvedt, M.
and Halpern, A.C., 1989. Model predicting survival in stage I melanoma based on tumor
progression. JNCI: Journal of the National Cancer Institute, 81(24), pp.1893-1904.

145
Danialan, R., Gopinath, A., Phelps, A., Murphy, M. and Grant-Kels, J.M., 2012. Accurate
identification of melanoma tumor margins: a review of the literature. Expert Review of Dermatology,
7(4), pp.343-358.
Ferlay, J., Soerjomataram, I., Dikshit, R., Eser, S., Mathers, C., Rebelo, M., Parkin, D.M., Forman,
D. and Bray, F., 2015. Cancer incidence and mortality worldwide: sources, methods and major
patterns in GLOBOCAN 2012. International journal of cancer, 136(5).
Gershenwald, J.E., Scolyer, R.A., Hess, K.R., Sondak, V.K., Long, G.V., Ross, M.I., Lazar, A.J.,
Faries, M.B., Kirkwood, J.M., McArthur, G.A. and Haydu, L.E., 2017. Melanoma staging: Evidence‐
based changes in the American Joint Committee on Cancer eighth edition cancer staging manual.
CA: a cancer journal for clinicians.
Lipscomb, D., Echchgadda, I., Peralta, X.G. and Wilmink, G.J., 2013, February. Determination of the
optical properties of melanin-pigmented human skin equivalents using terahertz time-domain
spectroscopy. In Terahertz and Ultrashort Electromagnetic Pulses for Biomedical Applications (Vol.
8585, p. 85850F). International Society for Optics and Photonics.
Olsovsky, C., Shelton, R., Carrasco-Zevallos, O., Applegate, B.E. and Maitland, K.C., 2013.
Chromatic confocal microscopy for multi-depth imaging of epithelial tissue. Biomedical optics
express, 4(5), pp.732-740.
Orzan, O.A., Șandru, A. and Jecan, C.R., 2015. Controversies in the diagnosis and treatment of
early cutaneous melanoma. Journal of medicine and life, 8(2), p.132.
Png, G. M., Flook, R., Ng, B. H., & Abbott, D. (2009). Terahertz spectroscopy of snap-frozen human
brain tissue: an initial study. Electronics Letters, 45(7), 343-345.
Rofstad, E.K., Steinsland, E., Kaalhus, O., Chang, Y.B., Høvik, B. and Lyng, H., 1994. Magnetic
Resonance Imaging of Human Melanoma Xenografts in Vivo: Proton Spin—lattice and Spin—spin
Relaxation Times Versus Fractional Tumour Water Content and Fraction of Necrotic Tumour Tissue.
International journal of radiation biology, 65(3), pp.387-401.
Rønne, C. and Keiding, S.R., 2002. Low frequency spectroscopy of liquid water using THz-time
domain spectroscopy.Journal of Molecular Liquids,101(1-3), pp.199-218.
Sim, Y.C., Park, J.Y., Ahn, K.M., Park, C. and Son, J.H., 2013. Terahertz imaging of excised oral
cancer at frozen temperature. Biomedical optics express, 4(8), pp.1413-1421.
Sladden, M.J., Balch, C., Barzilai, D.A., Berg, D., Freiman, A., Handiside, T., Hollis, S., Lens, M.B.
and Thompson, J.F., 2009. Surgical excision margins for primary cutaneous melanoma. The
Cochrane Library.
Tchernev, G. and Chokoeva, A.A., 2017. New Safety Margins for Melanoma Surgery: Nice
Possibility for Drinking of “Just That Cup of Coffee”?. Open access Macedonian journal of medical
sciences, 5(3), p.352.
Waldstein, S.M., Faatz, H., Szimacsek, M., Glodan, A.M., Podkowinski, D., Montuoro, A., Simader,
C., Gerendas, B.S. and Schmidt-Erfurth, U., 2015. Comparison of penetration depth in choroidal
imaging using swept source vs spectral domain optical coherence tomography. Eye, 29(3), p.409.
Xia, J., Yao, J. and Wang, L.V., 2014. Photoacoustic tomography: principles and advances.
Electromagnetic waves (Cambridge, Mass.), 147, p.1.
Figures

146
Figure 1. alternate human in-vivo skin imaging modalities and their

Figure 2. the absorption coefficient and refractive index of water at -330 C and 250 C

147
 Figure 3. (A) Skin modelling at -100 and 350 C, (assuming 75% water content and presuming
no other changes in THz absorption parameters), at 0.45 THz, using a modified Gaussian
excitation, (B) The XFdtd model, with part of the dermis rendered transparent, and the skin
containing rete ridges, nevi, hairs and sweat ducts, (C) a cross-section of the meshing
configuration.

S07-5 [17:30]
STUDENT PAPER
Analysis of effective pulsed power excitation for intracellular electropermeabilization
Farzan Zare2, Lakshitha Naranpanawe2 & Negareh Ghasemi1
1Queensland University of Technology (QUT), Brisbane, Australia, 4069
2University of Queensland (UQ), Brisbane, Australia, 4069
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Farzan Zare
Pulsed electric field induced electroporation for biomedical and agricultural purposes is becoming
increasingly ubiquitous. However, the efficacy of various pulse characteristics and stimulation
techniques are widely debated. A Jurkat T Lymphocyte cell is chosen as a model cell for finite
element simulation analysis where the significance of membrane pore density and rotation of the
cell is examined. The impact of these parameters on the spatial distribution of current density and
the transmembrane potential are investigated as a function of frequency (1Hz to 1 GHz) based on a
sinusoidal voltage excitation. A modulated pulse train combining low and high frequency excitation
is proposed allowing for maximum intracellular electromanipulation.
Introduction
Pulsed Electric Field (PEF) technology is widely used in the biomedical [1, 2] and agricultural
industries for transfection or apoptosis of a target cell line. Both applications involve the overt
phenomenon of electroporation. One of the current theories of electroporation is due to Maxwell-
Wagner-Sillars polarization and is speculated to occur when the transmembrane potential (TMP)
reaches approximately 1V. [3] This process initiates pore formation, increasing the permeability of
the cell membrane, through reversible or irreversible electroporation depending on various pulse
148
parameters [1, 4-7].
An emerging topic is targeted intracellular electromanipulation of a cell [8]. An effective technique for
electroporation is with high voltage pulse trains where the frequency, rise time, pulse width and
electric field strengths are carefully considered [4, 9]. Several cycles of rectangular pulses with
pulse-width ranging between nano to microsecond are used for electropermeabilization [2, 10].
Among the proposed characteristics, narrow pulses with faster rise times are found to be more
effective [1, 5].The output voltages of pulsed power systems can be bipolar or unipolar pulses. The
viability of bipolar and unipolar pulses on membrane permeabilization was studied and debated in
different literatures [11, 12]. Kotnik et al., Faurie et al. and Tekle et al. investigated the effects of
rectangular unipolar and bipolar pulses and concluded that the latter are more efficient for
electroporation and achieve the permeabilization of cells at lower applied voltage levels [11].
Comparatively, Pakhomov et al. and Gianulis et al. showed that sinusoidal unipolar pulses are far
more effective for cell membrane electroporation [5, 6]. According to their studies, there was no
attenuation in bioeffects when unipolar pulses were utilised whereas ‘bipolar cancellation’ was
observed when non-symmetrical bipolar pulses were applied. Therefore, more cell membrane
permeabilization was achieved using AC unipolar electric fields.
The transmembrane voltage of a spherical cell is distributed asymmetrically, and its absolute value
is highest at the poles given from the following equation [6, 13]:
TMP = 1.5 E R cosθ (1-e-t/τ) + TMPrest (1)
Where E, R, T and TMPrest are the electric field strength, cellular radius, time constant, and the
voltage across the membrane at rest, respectively. From equation (1) through the consideration of θ
as shown in figure 1 (b) it is trivial that the application of pulses yields an asymmetric distribution of
irreversible and reversible poration [7].
A different approach was introduced by Gianulis et al. where they identified that membrane
discharging is not exclusively responsible for the bipolar cancellation phenomenon [9]. The
relationship between various parameters have been extensively researched, yet it is still difficult to
definitively correlate a set of experimental results to a biological response (i.e. the inverse problem)
[14].
Simulation Model
Cells undergo rotation when placed within an AC electric field [13]. Furthermore, Andrei G.
Pakhomov, et al. has shown that a pulse train of nanosecond electric fields increase the number but
not the size of pores [15]. In this paper, the influence of rotation and number of pores on the TMP
and current density is analysed through finite element simulations.
In the finite element simulations, a theoretical double shell model is considered (Figure. 1 - a). A
Jurkat Lymphocyte cell was chosen for simulation as it has a dominantly spherical morphology that
minimises variances between practical and theoretical results.
A 2D AC-conductivity simulation module of ANSYS Maxwell was used for calculating the electric
field, current density and transmembrane potential that is subjected to a sinusoidal frequency sweep
from 1Hz to 1GHz.
However, due to high impedance value of the plasma membrane at low frequency excitation, the
effectiveness of intracellular electropermeabilization (i.e. nucleus) is less pronounced. Similar
simulation results for current density and TMP values were observed at a frequency range of 1Hz to
1MHz, therefore the analysis of the frequency range of 1 MHz to 1GHz is emphasised.
The first assumption is that no pores exist in the plasma membrane when the first pulse is initiated;
(Figure 1 (a-1)). The second assumption is that the number of pores increases as a function of the
number of applied pulses; (Figure 1 (a-2, 3 and 4) [15].
The model created in (figure 1 (a-1)) is kept consistent throughout the simulations. The only change
in variable is the number of pores and the rotation of the model. Furthermore, current density and
the voltages are consistently measured along the lines shown in Figure. 1 (a) for all simulations.

149
Cells were excited by 40V across a 40um electrode gap, where the electrode length and distance
between the electrode to cellular poles were symmetrical and consistent across all simulations. The
four models of pore density were simulated for each angle of the cell rotation with respect to the
electrodes.
Discussion
Identification of the spatial distribution of current density when exciting cells with PEF technology
can help solve the inverse problem.
The plasma membrane has very low impedance at frequencies above 100 MHz, thus, if the
excitation frequency is adjusted within this range, the electric pulsed excitation can affect
intracellular objects (i.e. nucleus and mitochondria). However, as the frequency drops below 1 MHz,
a higher number of pores is required to achieve similar intracellular results. This can be explained by
the increased local conductivity of the membrane.
As shown in Fig.1.c, when the cells are positioned at the angle of zero, the voltage across the cell
plasma membrane is decreased (16.5V, 2V, 1V, 0.7V) when the number of pores is increased (0, 1,
3, 5), respectively. Furthermore, when the pore placements are changed from zero, the impedance
of the cell model is increased at low frequency as its capacitive coupling becomes significant. This
impacts intracellular excitation due to a diminutive presence of current density within the whole cell
and a significant voltage appears across the cell plasma membrane.
Fig1.d shows the voltage across the nucleus membrane for the same electric field excitation.
Evidently at 1MHz - As the number of pores is increased, the voltage across the cell plasma
membrane is decreased while the voltage across the nucleus membrane is increased. However, at
higher frequency ranges (>100MHz), the electrical impedance of the plasma and nuclear
membranes decrease significantly resulting in higher current densities through the cell.
As shown in Figure 2, it is evident that at high frequency (>100 MHz) the plasma membrane does
not create any electrical shielding or buffering due to low impedances of the membranes.
Comparatively for 1 MHz excitation this differs due to the resistivity of pores affecting the overall
electrical model of the cell.
From figure 3, the electrical field and current density distribution depend on the overall cell
impedance and most importantly how the pores are positioned with respect to the electric field
direction (figure 3).
To achieve effective intracellular manipulation, one must consider the torque present in practical
applications (simulated as snapshots of rotation). To overcome such an obstacle, there is a
necessity for the application of fast pulses in quick succession. Moreover, for efficient application
cellular desensitization needs to be avoided as illustrated in a recent finding by Jensen et al. [4].
Conclusion
A new type of low and high frequency pulse train excitation is hypothesised to generate multiple
pores locally in a small area on the plasma membrane for which electric field distribution can affect
intracellular organelles.
A pure sinusoidal waveform (1Hz to 1GHz) with amplitude of 40V is applied as an excitation signal
in the simulation. Based on the Finite Element simulation results, high frequency has a stronger
effect for intracellular excitation. Thus, the pulsed train modulation should be synthesized to have
higher magnitudes at higher frequency ranges.
The electrical characteristics of the whole cell are significant as the optimum excitation frequency is
determined by these factors corresponding to their equivalent models. According to the simulation
results, we have realized that the excitation at 1MHz is the critical frequency for which the
membrane impedances can be electromanipulated.
References
[1] J. S. Beebe, M. N. Sain, and W. Ren, "Induction of Cell Death Mechanisms and Apoptosis by

150
Nanosecond Pulsed Electric Fields (nsPEFs)," Cells, vol. 2, 2013
[2] H. Hanna, A. Denzi, M. Liberti, F. M. André, and L. M. Mir, "Electropermeabilization of Inner and
Outer Cell Membranes with Microsecond Pulsed Electric Fields: Quantitative Study with Calcium
Ions," Scientific Reports, vol. 7, p. 13079, 2017/10/12
[3] W. Frey, J. White, R. Price, P. Blackmore, R. Joshi, R. Nuccitelli, et al., "Plasma membrane
voltage changes during nanosecond pulsed electric field exposure," Biophysical journal, vol. 90, pp.
3608-3615, 2006
[4] S. D. Jensen, V. A. Khorokhorina, C. Muratori, A. G. Pakhomov, and O. N. Pakhomova, "Delayed
hypersensitivity to nanosecond pulsed electric field in electroporated cells," Scientific Reports, vol. 7,
p. 10992, 2017/09/08
[5] K. H. Schoenbach, S. J. Hargrave, R. P. Joshi, J. F. Kolb, R. Nuccitelli, C. Osgood, et al.,
"Bioelectric Effects of Intense Nanosecond Pulses," IEEE Transactions on Dielectrics and Electrical
Insulation, vol. 14, pp. 1088-1109, 2007
[6] C. Polk and E. Postow, Handbook of Biological Effects of Electromagnetic Fields, -2 Volume Set:
CRC press, 1995
[7] C. Yao, H. Liu, Y. Zhao, Y. Mi, S. Dong, and Y. Lv, "Analysis of Dynamic Processes in Single-Cell
Electroporation and Their Effects on Parameter Selection Based on the Finite-Element Model," IEEE
Transactions on Plasma Science, vol. 45, pp. 889-900, 2017
[8] J. B. Stephen, F. B. Peter, W. Jody, P. J. Ravindra, and H. S. Karl, "Nanosecond pulsed electric
fields modulate cell function through intracellular signal transduction mechanisms," Physiological
Measurement, vol. 25, p. 1077, 2004
[9] E. C. Gianulis, M. Casciola, S. Xiao, O. N. Pakhomova, and A. G. Pakhomov,
"Electropermeabilization by uni- or bipolar nanosecond electric pulses: The impact of extracellular
conductivity," Bioelectrochemistry, vol. 119, pp. 10-19, 2018/02/01
[10] T. Batista Napotnik, M. Reberšek, P. T. Vernier, B. Mali, and D. Miklavčič, "Effects of high
voltage nanosecond electric pulses on eukaryotic cells (in vitro): A systematic review,"
Bioelectrochemistry, vol. 110, pp. 1-12, 2016/08/01
[11] T. Kotnik, L. M. Mir, K. Flisar, M. Puc, and D. Miklavčič, "Cell membrane electropermeabilization
by symmetrical bipolar rectangular pulses: Part I. Increased efficiency of permeabilization,"
Bioelectrochemistry, vol. 54, pp. 83-90, 2001/08/01
[12] C. Faurie, E. Phez, M. Golzio, C. Vossen, J.-C. Lesbordes, C. Delteil, et al., "Effect of electric
field vectoriality on electrically mediated gene delivery in mammalian cells," Biochimica et
Biophysica Acta (BBA) - Biomembranes, vol. 1665, pp. 92-100, 2004/10/11
[13] T. Kotnik, F. Bobanović, and D. Miklavcˇicˇ, "Sensitivity of transmembrane voltage induced by
applied electric fields—a theoretical analysis," Bioelectrochemistry and bioenergetics, vol. 43, pp.
285-291, 1997
[14] S. J. Beebe and K. H. Schoenbach, "Nanosecond pulsed electric fields: a new stimulus to
activate intracellular signaling," BioMed Research International, vol. 2005, pp. 297-300, 2005
[15] A. G. Pakhomov, E. Gianulis, P. T. Vernier, I. Semenov, S. Xiao, and O. N. Pakhomova,
"Multiple nanosecond electric pulses increase the number but not the size of long-lived nanopores in
the cell membrane," Biochimica et Biophysica Acta (BBA) - Biomembranes, vol. 1848, pp. 958-966,
2015/04/01
Figures

151
Figure 1. The cell simulation model (a) the chosen pore densities (no pores, 1, 3 and 5), (b)
the position of the pores (0o, 20o, 55o & 90o) with respect to the electric field excitation, (c) the
voltage across the plasma membrane and (d) voltage across the nuclear membrane.

152
Figure 2. The current density waveforms through each model at 0º (a) no pore, (b) one pore,
(c) three pores and (d) five pores.

153
 Figure 3. (a) five cells with different positions and pore configurations on cell and nucleus, (b)
2D electric field and (c) 2D current density distributions

S07-6 [17:45]
Improved cell permabilization by synergistic interaction between gold nanoparticles and
electromagnetic fields
Matej Kranjc1, Vitalij Novickij2, Tamara Pezić1, Saša Haberl1, Tina Batista Napotnik1 & Damijan Miklavčič1
1Faculty of Electrical Engineering, University of Ljubljana, Ljubljana, Slovenia, 1000
2Institute of High Magnetic Fields, Vilnius Gediminas Technical University, Vilnius,, Lithuania, 03227,
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Matej Kranjc
Pulsed Electromagnetic Field (PEMF) induced cell membrane permeabilization is an emerging
technique of contactless increase of membrane permeability. The inferior efficacy of PEMF
permeabilization compared to conventional electroporation is currently identified as a major
drawback of the methodology. To address the problem, we decided to enhance the PEMF induced
electric field by adding highly conductive gold nanoparticles (Au NPs) to enhance the induced
electric field. We have determined, that the addition of Au NPs enhanced PEMF electroporation and
allowed increase of the contactless permeabilization efficacy over PEMF electroporation alone with
no effect on cell survival.
Introduction
Pulsed electric field treatment is becoming a promising technique for permeabilization of cell
membrane that has gained considerable attention in the last few decades. The technique is based
on a phenomenon, known as electroporation or electropermeabilization, that is characterized by
transient increase of cell membrane permeability to molecules, which otherwise have poor or no
transmembrane transport mechanisms, by exposure of cells to externally applied pulsed electric
fields [1]. All conventional electroporation applications require direct contact between the electrodes
and the treated object, that is either via plate electrodes, which embrace the tissue, or using
invasive needle electrodes, which are inserted into the tissue. Pulsed electromagnetic fields (PEMF)
induced increase of the cell membrane permeability is similar to electroporation with the important
difference of non-invasive establishment of electric field by exposing a treated tissue to a time-
varying magnetic field. PEMF was successfully used for drug as well as for plasmid DNA delivery in
vitro and in vivo [2]–[4]. However, the induced electric field of PEMF was at best few V/cm, which is
at least two orders of magnitude lower than what is usually reported as electric field leading to

154
electroporation. In electroporation research, in order to cause membrane permeabilization, one
needs to overcome the threshold transmembrane voltage, which is induced by the electric field to
which cells are exposed and this usually is in the range of several hundreds V/cm [5].
It has been theoretically predicted and demonstrated that presence of conductive NPs close to the
membrane leads to membrane electroporation at lower fields, presumably due to local enhancement
of electric field [6]–[9]. In this study, we tested on Chinese hamster ovary cells, if adding highly
conductive gold nanoparticles (Au NPs), can enhance the induced electric field and improve cell
permeabilization.
Material and Methods
In order to test this hypothesis, we incubated Chinese hamster ovary (CHO) cells with gold
nanoparticles (Sigma-Aldrich, St. Louis, Mo, USA) of two different diameters (5 nm and 20 nm). To
get 50 µg/ml concentrations of nanoparticles, gold nanoparticles (Au NPs) in citrate buffer, size 5 nm
(cat. no. 741949), 20 nm (cat. no. 741965) were diluted with KPB electroporation buffer [10].
For PEMF generation, the 550 A, 2 kV generator has been used. The magnetic field inductor, which
served as a load of the system, consisted of 11 windings and 6 layers (total of 66 windings) with a
resulting total inductance of 9.8 μH. The duration of one PEMF pulse was 15 μs and maximum
amplitude was 5.5 T. The maximum induced electric field is proportional to the dB/dt of the pulse
and was in the range of 8 V/cm (Fig 1B). Bursts of 50 pulses at low frequency (1 Hz) were used and
a 30-second pause was introduced to allow cooling of the sample. The total number of pulses was
200 (50 x 4) with a total treatment time of approximately 5 min. For comparison, a conventional
electroporation pulses were generated between stainless steel plate electrodes (d = 2 mm) in a
sequence of 8 x 100 μs square wave pulses of 250 V amplitude (Fig 1A).
Before cells were exposed to either pulsed electric filed (PEF) treatment or PEMF treatment, CHO
cell suspension with or without Au NPs was mixed with fluorescent markers – either propidium
iodide (PI) or YO-PRO®-1 (YP). After the PEMF and PEF treatment samples were incubated at
room temperature (22°C) for 3 min in order for fluorescent dye to enter permeabilized cells.
Untreated samples were not exposed to pulse treatment, but were incubated at room temperature
for the same time as treated samples. Cells were then diluted in 100 µl of KPB buffer to stop inflow
of fluorescent dye, following by analysis using flow cytometer (Attune NxT; Life Technologies,
Carlsbad, CA, USA). The percentage of permeabilization was defined by gating.
Results
The effect of PEMF on permeabilization of CHO cells with and without NPs was analyzed. The
representative shifts of fluorescence spectra after PEMF treatment for both PI and YP assays are
shown in Fig. 2.
As it can be seen, with NPs (20 nm, 50 µg/ml) the shift of spectra using the PI assay is more
expressed compared to YP. However, without NPs, the shift due to PEMF was detectable only in the
YP case (Fig. 2C). The post-processed and normalized results of PEMF induced electroporation
without and with NPs are shown in Fig. 3.
PEMF induced permeabilization without NPs was detectable (P<0.05) only using the YP assay
when the median fluorescence was analyzed. Similar to PEF results (data not shown), cell
permeabilization after PEMF treatment was significantly increased when the NPs were added, while
the 5 nm NPs were less efficient if compared to 20 nm NPs.
PEMF treatment has no effect on survival of cells with or without the Au NPs. There is no statistical
difference between PEMF treated (100% of survived cells) and untreated samples (97.4 ± 3.4 %
survived cells).
Discussion
Introduction of nanoparticles during pulsed electric field (PEF) or pulsed electromagnetic field
(PEMF) treatment allows modification of the extracellular medium conductivity and permittivity, while
conglomeration of NPs near cell wall locally enhances electric field and thus the transmembrane

155
voltage potential [7]. The proof of concept for PEF was confirmed both in silico [6], [7] and in vitro [9]
while this study demonstrate enhancement of cell membrane permeability using pulsed
electromagnetic fields.
Mechanisms and pathways of increased molecular transmembrane transport induced by the
electromagnetic pulses are currently not known. There were several possible mechanisms
suggested in the literature, e.g., electro-endocytosis [2], [11], [12], electrophoresis [11], creation of
metastable membrane pores via interaction with membrane-attached magnetic particles [11],
electroporation due to induced electric field [2], [13], changed receptor binding or activation [12] and
induction of magnetomotive force which interacts with ions [3], [14]. Moreover, the speculations on
the PEMF induced permeabilization mechanism were always straightforwardly associated with
electroporation due to linear dependence of the induced electric field on the dB/dt of the pulse. Our
data further supports the hypothesis of electrostatics behind the formation of pores due to observed
increase of permeabilization due to local field enhancement by Au NPs. However, it does not explain
the occurrence of permeabilization in fields that are by several orders lower if compared to
conventional electroporation. High permeabilization of CHO cells was achieved in PEF >400 V/cm
(when enhanced by 20 nm Au NPs), however same levels of permeabilization were detected when
the PEMF induced electric field was merely 8 V/cm. Still, the number of applied electric pulses in
PEF treatment was 8, which is considerably lower than 200 pulses applied in PEMF treatment.
Observed discrepancy suggests to a relation between amplitude and number of pulses that results
in the same level of permeabilization, similar as with conventional electroporation [15].
Acknowledgments
This work was supported by the Slovenian Research Agency (research core funding No. P2-0249).
This work was also partly supported by the Research Council of Lithuania (Grant Nr. LAT-02/2016).
The research was conducted in the scope of the electroporation in Biology and Medicine (EBAM)
European Associated Laboratory (LEA) and within Infrastructure Programme: Network of research
infrastructure centers at University of Ljubljana (MRIC UL IP-0510).
References
[1] T. Kotnik, P. Kramar, G. Pucihar, D. Miklavčič, M. Tarek, D. Miklavcic, and M. Tarek, “Cell
Membrane Electroporation-Part 1: The Phenomenon,” IEEE Electr. Insul. Mag., vol. 28, no. 5, pp.
14–23, Oct. 2012.
[2] S. Kranjc, M. Kranjc, J. Sčančar, J. Jelenc, G. Serša, and D. Miklavčič, “Electrochemotherapy by
pulsed electromagnetic field treatment (PEMF) in mouse melanoma B16F10 in vivo.,” Radiol.
Oncol., vol. 50, no. 1, pp. 39–48, Mar. 2016.
[3] T. J. Kardos and D. P. Rabussay, “Contactless magneto-permeabilization for intracellular plasmid
DNA delivery in-vivo.,” Hum. Vaccin. Immunother., vol. 8, no. 11, pp. 1707–13, Nov. 2012.
[4] V. Novickij, J. Dermol, A. Grainys, M. Kranjc, and D. Miklavčič, “Membrane permeabilization of
mammalian cells using bursts of high magnetic field pulses,” PeerJ, vol. 5, no. April, p. e3267, 2017.
[5] D. Miklavcic, D. Semrov, H. Mekid, and L. M. Mir, “A validated model of in vivo electric field
distribution in tissues for electrochemotherapy and for DNA electrotransfer for gene therapy.,”
Biochim. Biophys. Acta, vol. 1523, no. 1, pp. 73–83, Sep. 2000.
[6] H. Qiu, R. P. Joshi, and A. Pradhan, “Simulation of nanoparticle based enhancement of cellular
electroporation for biomedical applications,” J. Appl. Phys., vol. 116, no. 18, 2014.
[7] J. Lekner, “Electroporation in cancer therapy without insertion of electrodes.,” Phys. Med. Biol.,
vol. 59, no. 20, pp. 6031–6042, Sep. 2014.
[8] Z. Rezaee, A. Yadollahpour, V. Bayati, and F. N. Dehbashi, “Gold nanoparticles and
electroporation impose both separate and synergistic radiosensitizing effects in HT-29 tumor cells:
An in vitro study,” Int. J. Nanomedicine, vol. 12, pp. 1431–1439, 2017.
[9] Y. Zu, S. Huang, W. C. Liao, Y. Lu, and S. Wang, “Gold nanoparticles enhanced electroporation
for mammalian cell transfection,” J. Biomed. Nanotechnol., vol. 10, no. 6, pp. 982–992, 2014.

156
[10] C. Blangero and J. Teissié, “Ionic modulation of electrically induced fusion of mammalian cells.,”
J. Membr. Biol., vol. 86, no. 3, pp. 247–53, 1985.
[11] L. Towhidi, S. M. P. Firoozabadi, H. Mozdarani, and D. Miklavcic, “Lucifer Yellow uptake by CHO
cells exposed to magnetic and electric pulses.,” Radiol. Oncol., vol. 46, no. 2, pp. 119–25, Jun.
2012.
[12] Z. Shankayi, S. M. P. Firoozabadi, M. Mansourian, and A. Mahna, “The effects of pulsed
magnetic field exposure on the permeability of leukemia cancer cells,” Electromagn. Biol. Med., vol.
33, no. 2, pp. 154–158, Jun. 2014.
[13] V. Novickij, A. Grainys, J. Novickij, and S. Markovskaja, “Irreversible magnetoporation of micro-
organisms in high pulsed magnetic fields.,” IET Nanobiotechnol., vol. 8, no. 3, pp. 157–62, Sep.
2014.
[14] V. Novickij, A. Grainys, E. Lastauskienė, R. Kananavičiūtė, D. Pamedytytė, L. Kalėdienė, J.
Novickij, and D. Miklavčič, “Pulsed Electromagnetic Field Assisted in vitro Electroporation: A Pilot
Study,” Sci. Rep., vol. 6, no. 1, p. 33537, Dec. 2016.
[15] G. Pucihar, J. Krmelj, M. Rebersek, T. Napotnik, and D. Miklavcic, “Equivalent Pulse
Parameters for Electroporation,” IEEE Trans. Biomed. Eng., vol. 58, no. 11, pp. 3279–3288, Nov.
2011.
Figures

157
Figure 1. Applied PEF and PEMF pulses, where A - voltage and current waveform of 100 µs
square wave pulse applied in a sequence of 8, B - the waveform of the applied
electromagnetic field pulse.

158
Figure 2. The dependence of PI (A,B) and YP (C,D) fluorescence intensity spectra on PEMF
treatment samples without (A,C) or with (B,D) 20 nm Au NPs. The dependence of PI (E) and
YP (F) fluorescence intensity spectra on PEF treatment without Au Nps. Left of the line (gate)
are non-permeabilized cells and right of the line (gate) are permeabilized cells.

159
Figure 3. Permeabilization of cells after PEMF treatment. Cells without nanoparticles (w/o
NPs) and cell with 5 nm or 20 nm Au NPs were used. The percentage of PI and YP
fluorescent cells (A) as well as median fluorescence (B) are shown. Each sample and
fluorescent dye has their own untreated sample (PI control or YP control). Asterisk (*) marks
statistically significant difference between treated sample and its corresponding untreated
sample (P<0.05).

160
 Session: S08
In vivo
Wednesday June 27, 2018 • 11:30 - 12:30
Europa C
Chairs: Florence Poulletier De Gannes & Jukka Juutilainen

S08-1 [11:30]
Neuroinflammation promotes brain cell responses to 1800 MHz GSM electromagnetic fields
Julie Lameth1, Florian Occelli2, Severine Boillée1, Delia Arnaud-Cormos3, Philippe Leveque3, Jean-Marc
Edeline2 & Michel Mallat1
1INSERM U 1127, Sorbonne Université, CNRS, ICM, Paris, France, F-75013
2UMR 9197 CNRS, Paris Saclay Institute of Neuroscience, Neuro-PSI, Université Paris-Sud, Orsay, France,
F-9197
3UMR-7252, Univ. Limoges, CNRS, XLIM, Limoges, France, F-87000
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Michel Mallat
We have investigated brain cell responses to a two-hour head exposure to GSM-1800 MHz in rats
undergoing neuroinflammation. In the dorso-medial area of the cerebral cortex (Cx), GSM exposure
(mean SAR of 2.94 W/kg) led to transient reduction in Interleukin-1ß gene expression, increase in
the length of cell processes emitted by microglia and post translational modifications of glutamate
receptors. In the auditory Cx, GSM exposure (1.55 W/kg) resulted in both extension of microglial cell
processes and altered responses of neurons to auditory stimuli, including reduced firing rate and
enhanced duration of evoked responses. Comparisons with GSM-exposed healthy rats showed that
neuroinflammation promoted cortical cell responses to GSM signals.
Introduction
Public health concerns raised by the development of wireless communication stimulate
investigations of the biological effects of electromagnetic fields (EMF) such as GSM-900 or1800
MHz. Many studies have looked for possible brain cell responses to GSM EMF in healthy subjects
but few of them have addressed the influence of these EMF under neuropathological states.
It is now well established that virtually all pathologies or lesions affecting the central nervous system
(CNS) involve neuroinflammatory reactions marked by the activation of CNS-resident immune cells
such as microglia. These cell express specific macrophage markers including the Iba1 protein.
Microglia strikingly differ from other CNS cells owing to their functional and morphological plasticity,
which is highlighted by continuous extension and retraction of microglial cell processes. Microglial
cell motility is regulated by neurotransmitters, nucleotide or ions released by neighboring neurons or
glial cells. Activated microglia are thought to play key pathophysiological roles, due to their capacity
to produce a variety of compounds including proinflammatory cytokines or reactive oxygen species
which may impact on the activity or the wiring of neuronal networks (Salter and Stevens, 2017; Wolf
et al., 2017).
In the current studies we used a classical rodent model of acute neuroinflammation set by treating
rats with bacteria-derived lipolysaccharide (LPS) (Dantzer et al., 2001). This allowed us to determine
whether GSM-EMF (1800 MHz) can affect activated microglial cells, excitatory neurotransmission
and electrical activity of neurons in the context of an acute neuroinflammatory reaction.
Single head -only exposures to GSM EMF were applied to normal or LPS-treated rats. Our study
focused on the cerebral cortex (Cx) at time points varying between 3 to 72 hours following the EMF
exposure. The effect of GSM EMF on neuroinflammation was determined by measuring the
expression of proinflammatory genes encoding cytokines or enzymes catalyzing the formation of
reactive oxygen species and by morphological analyses of microglia. Excitatory neurotransmission
was probed by detecting phosphorylation of serine residues in glutamatergic AMPA (alpha amino-

161
3-hydroxy-5-methylisoxazole-4-propionic acid) receptors (AMPAR). Electrical activities of neurons
were analysed in the auditory Cx .
Methods
Detailed experimental procedures are reported in Lameth et al. (2017) and Ocelli et al., (under
revisions).
Adult male Wistar rats pretreated or not with LPS (250 µg/kg) were exposed to a GSM
electromagnetic field (RFPA, 1/8 duty factor) pulsed at 217 Hz (pulse emission every 4.6 ms for 546
µsec).. Head-exposures to EMF were carried out under light anesthesia to prevent movement and
ensure reproducible position of the animal heads below the loop antenna emitting the GSM signal.
The duration of exposure was set at 2 hours. Half of the rats from the same litters were placed in the
device without exposure to RF and served as control (sham-exposure).
Specific absorption rates (SARs) were determined numerically in different brain regions using a
numerical rat model with the Finite Difference Time Domain (FDTD) method.
Biochemical analyses were focused on the dorso-medial area of the Cx. The level of expression of
proinflammatory genes was assessed by quantitative reverse transcription-PCR. The
phosphorylation of serine residues in the GluA1 AMPAR sub units (GluA1) was detected and
quantified in Western blot experiments.
The length of microglial cell processes were determined by computer-assisted analysis of images
captured in fixed cortical sections (dorso-medial and auditory area), following immunofluorescent
staining of microglia using anti-Iba1 antibody.
Extracellular recordings from neurons of the primary auditory Cx (A1) was studied with multi-unit
activity (MUA) collected with arrays of 16 electrodes. Acoustic stimuli consisted in sequences of
pure tones ranging between 140 Hz and 36 kHz. Significant responses were defined as firing rates
above the average levels of the baseline activity plus six times the standard deviation of the
baseline (spontaneous activity). Natural stimuli such as guinea pig whistle were also presented.
Statistical analyses were performed using Student t test and non parametric Mann-Whittney test.
Threshold for significant difference was set at p <0.05
Results
Adults rats were treated with LPS 24h before GSM or Sham exposure. The effect of EMF were first
analyzed in the dorso-medial area of the Cx (DMCx) in which the mean SAR value was estimated to
2.94 W/kg. LPS-activated microglia had a bushy like cell morphology marked by highly ramified
tortuous processes in cortical sections fixed 24h after sham or GSM–exposure. However
comparison of GSM and sham -exposed rats showed that microglial cell processes stained by Iba1
immunoreactivity were more developed in GSM-exposed rats. This morphological cell response was
associated with a significant 50% reduction in the mean level of interleukin 1ß transcripts assessed
in DMCx extracts. Analyses of the proteins recovered from DMCx revealed a decreased
phosphorylation at two residues on the GluA1 AMPAR subunit (serine 831 and 845), suggesting that
GSM-exposure had affected excitatory neurotransmission. The GSM–induced changes in microglial
cell processes, interleukin 1ß gene expression, and GluA1 phosphorylation did not persist 72h after
EMF exposure and were not observed in healthy rats exposed to GSM EMF without LPS
pretreatment (Lameth et al., 2017).
To determine the effect of GSM on neuronal activity in LPS-treated rats exposed to EMF, we
focused on the auditory Cx (ACx) in which the mean SAR value was 1.55 W/kg. Iba1
immunohistochemistry applied to the ACx confirmed the EMF-induced microglial response and
showed that the growth of microglial cell processes already occurred 3h after EMF exposure.
Electrophysiological recording performed between 3 to 6h after GSM or sham exposure revealed a
significant 33 % reduction in firing rate in GSM-exposed rats. Moreover, GSM-exposure resulted in
altered neuronal responses to pure tones or guinea pig vocalizations, as indicated by significant
reductions (20 to 40%) in the number of evoked action potential together with a 12% increase in the
mean duration of the evoked firing (Occelli et al., under revisions).
162
Ongoing work aims at determining whether a same GSM exposure may impact on a chronicle
neuroinflammatory state, in transgenic rats that model a inheritable form of human amyotrophic
lateral sclerosis (Thomsen et al., 2014). Preliminary data indicate that GSM-EMF applied at a
preclinical stage of the disease, can induce reduction in the expression of proinflammatory genes
assessed in the motor area of the DMCx.
Discussion
Our results provide evidence that a single head- exposure of LPS-treated rats to GSM-1800 MHz
can interfere with the LPS-triggered neuroinflammatory reaction as shown by modifications in the
morphology of activated microglial cells and reduced expression of the proinflammatory interleukin
1ß gene. GSM exposure also induced changes in the activity of neuronal network, which are
marked by post translational modifications of AMPA receptors and modifications in spontaneous or
evoked neuronal firing. The effect of the EMF on activated microglia and cortical neurons are
unlikely to be independent, considering that the motility of microglial cell processes is regulated by
neuronal activity, whereas activated microglia can affect neurotransmission (Salter and Stevens,
2017). Importantly however, our analyses of rat exposed to EMF without LPS pretreatment indicate
that the cortical cell responses to GSM required or at least depend on the acute neuroinflammatory
state triggered by LPS. This contention is further supported by our previous studies showing that our
experimental setting used for GSM exposures did not trigger cell stress responses or glial reaction
in healthy juvenile (35 day-old) rats (Watilliaux et al., 2010). Therefore we conclude that an acute
neuroinflammatory state can promote brain cell responses to GSM-EMF applied at cortical SAR
values ranging between 1.55 and 2.94 W/kg. These responses were detectable during the 24h
period that followed the exposure but were reversible as indicated by brain analyses carried out 3
days after the exposure.
Acknowledgments
This research was supported by the French National Research Program for Environmental and
Occupational health of ANSES (grants 2013/2/03 and 2015/2 RF/12), and funding from the program
“Investissements d’avenir” ANR-10-IAIHU-06.
References
Dantzer, R (2001). Cytokine-induced sickness behavior: where do we stand? Brain Behav Immun
15:7-24.
Lameth, J, Gervais, A, Colin,C, Leveque, P, Jay, TM, Edeline, JM, Mallat, M (2017). Acute
Neuroinflammation Promotes Cell Responses to 1800 MHz GSM Electromagnetic Fields in the Rat
Cerebral Cortex. Neurotox Res 32:444-459.
Ocelli, F, Lameth, J, Adenis, V, Huetz C, Leveque P, Jay, TM, Edeline JM, Mallat M. A single
exposure to GSM-1800MHz signals in the course of an acute neuroinflammatory reaction can alter
neuronal responses and microglial morphology in the rat primary auditory cortex. (under review).
Salter, MW, B Stevens (2017). Microglia emerge as central players in brain disease. Nat Med
23:1018-1027.
Thomsen, G.M., Gowing, G., Latter, J., Chen, et al. (2014). Delayed disease onset and extended
survival in the SOD1G93A rat modelof amyotrophic lateral sclerosis after suppression of mutant
SOD1 in the motor cortex. J Neurosci 34, 15587-15600.
Watilliaux, A, Edeline, JM, Leveque, P, Jay, TM, Mallat, M (2011). Effect of exposure to 1,800 MHz
electromagnetic fields on heat shock proteins and glial cells in the brain of developing rats.Neurotox
Res 20:109-119.
Wolf, SA, Boddeke, HW, Kettenmann, H (2017). Microglia in Physiology and Disease. AnnuRev
Physiol 79:619-643.

163
S08-2 [11:45]
Effect of 835 MHz radiofrequency radiation exposure on glucose metabolism in aging mice
model
Hye-Ryoung Lee1, Ju-Hwan Kim1, Hakrim Kim1, Hyung-Gun Kim1 & Jin-Koo Lee1
1Department of Pharmacology, Dankook University, Cheonan, Korea, 31116
Keywords: In vivo, RF/Microwaves, Completed (unpublished)
Presented by: Hye-Ryoung Lee
Chronic exposure to RFR in aging animal models can lead to hypoglycemia by causing
abnormalities in glucose metabolism. The resulting hypoglycemia can aggravate the survival rate in
aging mice model.
In this study we investigate the changes of glucose metabolism regulating long-term exposure to
RFR using an aging animal model. Aging mice exposed to 835 MHz RFR (SAR 4.0 W/kg) for 9
weeks showed a decrease in body weight, blood glucose and survival rate. And also, plasma IL-6
and IL-8 and hypothalamic CRH mRNA levels were increased by RFR in aging mice. These results
show that RFR can act as external stressors and further promote aging. Plasma insulin and leptin
levels and daily food intake were significantly decreased in aging mice exposed to long-term RFR.
Expression of orexigenic neuropeptides and ObRs were further increased by RFR exposure. In
contrast, expression of anorexic neuropeptides were further decreased by RFR exposure compared
to RFR-exposed normal group. These results show that RFR affects the expression of appetite-
related neuropeptides causing abnormalities in glucose metabolism regulation. Lastly, we
investigated insulin signaling in RFR-exposure aging mice. The mRNA expression levels of
AMPKa1, AMPKa2, FoxO1, FoxO3a, SIRT-1 and PGC-1a and the phosphorylation of PI3K and
STAT-3 were significantly increased in the RFR-exposed hypothalamus of aging mice. In addition,
expression of GLUT-4 mRNA level was significantly increased in the RFR-exposed soleus muscle of
aging mice. These results demonstrate that the RFR activates the insulin signaling in the
hypothalamus and promotes GLUT-4 expression in muscle cells. Taken together, chronic exposure
to RFR in aging animal models can lead to hypoglycemia by causing abnormalities in glucose
metabolism. The resulting hypoglycemia can aggravate the survival rate in aging mice model.

S08-3 [12:00]
STUDENT PAPER
Interleukin-6 plays an essential role in pulsed electromagnetic field improving cartilage and
subchondral bone in mice with osteoarthritis
Xiaotian Yang1, 2, Hongchen He1, 2, Siyi Zhu1, 2 & Chengqi He1, 2
1Department of Rehabilitation Medicine, West China Hospital, Sichuan University, Chengdu, China, 610041
2Key Laboratory of Rehabilitation Medicine in Sichuan, Chengdu, China, 610041
Keywords: In vivo, Pulsed, Completed (unpublished)
Presented by: Xiaotian Yang
PEMF therapy preserves cartilage degeneration and subchondral bone microstructure in mice with
knee OA by inhibiting IL-6/Stat3 signaling. Interleukin-6 gene knockout alleviate cartilage
degradation in mice with OA.
Introduction: Pulsed electromagnetic field (PEMF) has been investigated as a noninvasive
alternative method to protect cartilage and subchondral bone from osteoarthritis (OA), and have
anti-inflammatory and pro-regenerative effects. Interleukin-6 (IL-6) is a major cytokine in the
pathogenesis of OA. Inhibition of IL-6 can alleviate cartilage degradation in mice with OA. This study
aimed to investigate whether IL-6 is crucially involved in PEMF preserving cartilage degeneration
and improving subchondral bone microstructure.
Methods: Forty 10-week-old male wild type (WT) C57BL/6 and IL-6 knockout (IL-6-/-) mice were

164
randomly divided into four groups (n=10): IL-6-/-, WT, IL-6-/-+PEMF, WT+PEMF. OA was induced by
the destabilization of the medial meniscus (DMM) of the right knee. We compared effects of pulsed
electromagnetic field therapy at different frequencies (8Hz, 50Hz, 75Hz) in mice with OA in our
previous work, and found 75 Hz PEMF therapy has a positive effect on cartilage and subchondral
bone in OA mice. In this study, the PEMF groups were exposed to 75 Hz, 3.8 mT PEMF for 1 h/day,
7 days/week, 4 weeks. Cartilage destruction was investigated by Safranin O staining. Bone
microstructure was evaluated by micro-computed tomography. Gene expressions were assayed
with real-time quantitative polymerase chain reaction. Protein expressions were observed by the
immunohistochemistry staining.
Results: The WT+PEMF and IL-6-/- group increased bone volume fraction (BV/TV), trabecular
thickness (Tb.Th), trabecular number (Tb.N), and suppressed bone surface/bone volume (BS/BV)
and trabecular separation (Tb.Sp) levels in the micro-computed tomography analysis. Real-time
PCR analysis showed that PEMF decreased cartilage and subchondral bone’s IL-6, Stat3, MMP-13,
Adamts4 and Adamts5 mRNA levels. Similar results were observed in the immunohistochemistry
staining. No significant differences were found between IL-6-/- and IL-6-/-+PEMF group.
Conclusion: PEMF therapy preserves cartilage and subchondral bone in knee OA mice by inhibiting
IL-6/Stat3 signaling.

S08-4 [12:15]
Effects of extremely low frequency electromagnetic field (50 Hz) exposure on the bone
marrow cells of mice
Isabelle Lagroye1, 2, Emmanuelle Poque-Haro2, Rémy Renom2, Yves Le Dréan3, Florence Poulletier De
Gannes2, Annabelle Hurtier2, Yann Percherancier2, Denis Habauzit3 & Bernard Veyret1, 2
1EPHE, PSL, Paris, France, 75014
2IMS laboratory UMR5218, Bordeaux University, Talence, France, 33405
3Inserm U1085, Institut de Recherche sur la Santé, l’Environnement et le Travail – IRSET, Rennes, France,
35000
Keywords: In vivo, ELF/LF, Completed (unpublished)
Presented by: Isabelle Lagroye
BDF1 mice (up to 12 litters per group) were exposed 8 h/d, 5d/w for up to 8w of age at 1000 µT,
beginning at conception. We tested a multi-hit model by irradiating fathers (X-rays) to induce
genomic instability in the pups and further treatment with the genotoxic agent MNU in the young
mice. In the bone marrow cells, genomic instability was evaluated as the presence of DNA damage,
while the presence of gamma-H2AX and micronuclei was also assessed in the peripheral blood
cells. Exposure to MF did not induce nor promote genomic instability in critical tissues. To our
knowledge, this is the first in vivo study focusing on prenatal exposure and bone marrow in the
context of a possible relationship between 50 Hz MF and childhood leukaemia.
Introduction
The International Agency for Research on Cancer (IARC) has classified the extremely low frequency
(ELF) magnetic fields (MF) as "possible human carcinogen" for childhood leukaemia (B-cell acute
lymphoblastic leukaemia: B-ALL). This classification is based on epidemiological data but not
supported by experimental data (especially carcinogenicity and genotoxicity) and the current
understanding of the action mechanisms of low amplitude ELF MF.
The CLeMaN project makes the hypothesis that in-utero and early-life exposure of mice to 50Hz MF,
alone or in combination with a physical and/or chemical treatment could induce or promote genomic
instability (GI) in the bone marrow (BM). This could favour the occurrence of ALL and thus explain
how 50Hz MF may affect B-ALL without inducing immediate genotoxic damage.
Material and Methods

165
The exposure system consists of two identical sets of Merritt coils powered via a voltage controller,
set up in an air-conditioned room. The exposure system generates a vertical MF up to 1000μT, at 50
Hz, uniform to within 5% inside the cage. Free-moving mice are placed in four cages located at the
centre of each, independently-activated set of coils. Magnetic flux density and temperature are
monitored during exposure.
BDF1 mice, hybrids of DBA2 males and C57Bl/6 females [(♂DBA2) x (♀C57Bl/6) F1] are used in
this study. The animals are housed under controlled temperature (21±2°C) and 12 h light/dark
regimen (08:00–20:00). The experimental protocol followed the ethical procedures recommended by
French legislation and was approved (DIR 1371) by the Bordeaux University ethical committee.
To test whether MF exposure can induce genomic instability, BDF1 hybrid mice are either exposed
(MF group) or sham exposed (SHAM group) in a randomized fashion. They are exposed at 1000 µT
(the highest available level), 8hours/day, 5days/week. Mice are provided with food and water during
exposure. Exposure begins the day of copulation (E0.5) and last up to 8 weeks postnatal, when at
maturity. Pups are exposed with their mothers until weaning (3-4 weeks post-natal).
To generate BDF1 mice with induced genomic instability (IGI), DBA2 male mice were given a whole-
body acute irradiation of 1Gy of X-rays and mated 6 weeks later to non-irradiated C57Bl/6 females.
The effect of MF exposure is tested on these mice with (SHAM MNU IGI and MF MNU IGI groups)
or without (SHAM IGI and MF IGI groups) a further methyl nitrosourea treatment (MNU, 50 mg/kg).
MNU treated-mice were sacrificed 72 hrs later.
Here, we report on DNA damage in BM and peripheral blood cells using the alkaline comet assay
(mainly DNA single strand breaks and alkali-labile lesions) and/or gamma-H2AX staining (DNA
double strand breaks) and the micronucleus assay (chromosome fragments and/or whole
chromosomes). These endpoints were measured after exposure in up to 12 litters per group (1
mouse per litter). All samples were coded.
The alkaline comet assay was adapted from Olive (1999). A total of 100 comets from duplicate
slides were analysed per animal using the Comet Assay IV software (Perceptive Instrument, UK).
Olive tail moment (OTM) and the percentage of DNA intensity in the tail were used as DNA damage
parameters.
The number of micronuclei and the phosphorylation of H2AX were assessed using flow cytometry
(BD FACSCanto IITM flow cytometer, France) equipped with BD FACSDiva software v.8.0.1. To
detect micronuclei, cells were stained using the Microflow® kit (Litron Laboratoires New York, USA),
according to the supplier protocol. 10000 gated events were acquired from each sample.
Statistical analysis used the non-parametric Kruskall-Wallis test for small samples followed, if
significant, by the Mann-Whitney test (Anastats®).
Results
MF exposure alone didn’t induce genomic instability in BM cells and peripheral blood cells as shown
by the lack of DNA damage, gH2AX expression and micronuclei induction in those cells.
In BM cells, MNU induced significant DNA damage. X-rays irradiation of the fathers induced
genomic instability in the BM cells of their pups (with or without further MNU treatment) as shown by
the significant increase in DNA damage in comparison to pups born from non-irradiated fathers (Fig.
1).
In the peripheral blood cells, the genotoxic effect of MNU was detected through the induction of
DNA damage only. We saw a trend for an increase in micronuclei but it didn’t reach the statistical
significance. Finally, X-rays irradiation of the fathers protected their pups from the genotoxic effects
of MNU in both sham- and MF-exposed mice. This effect was detected using the comet assay.
Conclusion
Using a combination of physical and chemical agents, the present project is consistent with the
current understanding of childhood leukaemia as a multi-hit disease. To our knowledge, this is the

166
first in vivo study focusing on prenatal exposure and bone marrow in the context of a possible
relationship between 50 Hz MF and childhood leukaemia. While the magnetic flux density tested
here was much higher than can be found in the environment, we found no evidence that early-life,
including prenatal, exposure to ELF-MF could induce or promote genomic instability in the mouse
bone marrow and peripheral blood cells.
Acknowledgement
Sponsor: French National Alliance for Life Sciences and Health (Aviesan) N°ENV201309
Figures

Figure 1. DNA damage expressed as the Olive Tail moment (mean ± SEM) in the bone
marrow cells on BDF1 pups sham-exposed (SHAM) or exposed to 50 Hz MF 1 mT (Exposed)
± MNU ± induced genomic instability through father irradiation (IGI). n=9-12 mice (1
mice/litter).

167
 Session: S09
Mechanisms
Wednesday June 27, 2018 • 11:30 - 12:30
Europa D
Chairs: Michal Cifra & Francis Hart

S09-1 [11:30]
How can 10-100 μT, 10-100 Hz magnetic field influence human cardiovascular regulation?
Joel J Niederhauser1
1Bioelectromagnetics, InnoClever GmbH, Liestal, Switzerland, 4410
Keywords: Human, ELF/LF, Completed (unpublished)
Presented by: Joel J Niederhauser
Since 1896, studies exposing human subjects to 10-100 μT, 10-100 Hz magnetic fields (MF) have
recurrently shown acute influence on cardiovascular regulation. MF exposures above 100-200 μT
have not shown an influence. Unlike magnetophosphenes, known to appear at 5-10 mT,
mechanisms for influence of MF on human cardiovascular regulation have received little attention.
Based on magnetophosphene knowledge and baroreceptor sensor geometry of 5 mm diameter and
10 mm length, baroreceptor MF sensitivity threshold is expected to be in the order of 10-20 μT. We
present a simple converging neural network model demonstrating amplitude window behaviour as
possible explanation for observed MF influence on human cardiovascular regulation.
Introduction
In 1896 Arsène d'Arsonval reported symptoms of altered cardiovascular regulation in subjects
plunging the head into 42 Hz magnetic field (MF) in the order of 100 μT in a large solenoid [1]. He
also reported magnetophosphenes induced by magnetic field of a smaller coil.
Magnetophospenes have been extensively studied and a frequency dependent threshold in the
order of 5-10 mT MF at the retina has been established [2].
The influence of MF on cardiovascular regulation has received less attention. Studies below 1 μT
and above 100 μT reported no effect [2]. As shown in Figure 1 there are repeated reports in the
10-100 μT exposure range, especially when challenging the sympathetic part of cardiovascular
regulation, that showed influence of 16-60 Hz MF on slowing heart rate and altered heart rate
variability in human subjects [1–21].
We present a putative mechanism for influence of 10-100 μT, 10-100 Hz magnetic field on human
cardiovascular regulation.
Materials and Methods
Anatomy and physiology
Human cardiovascular regulation stabilizes blood pressure and oxygen supply to the human brain.
Baroreceptors in the carotid arteries (neck) are sensors for blood pressure and flow. The pressure
dependent firing rate of the mechanoreceptors shows no firing below a threshold (e.g. 80 mmHg)
and pressure dependent firing rate above the threshold.
The baroreceptor signal is processed in the nucleus of the solitary tract in the brainstem. Depending
on systolic (e.g. 120 mmHg) and diastolic (e.g. 80 mmHg) pressure, there are two distinct feedback
mechanisms: (1) parasympathetic autonomous control slows heart rate if systolic pressure is too
high, (2) sympathetic autonomous control increases blood vessel constriction and slightly increases
heart rate if diastolic pressure is too low. The sympathetic effect on low frequency heart rate
fluctuation is an order of magnitude below the parasympathetic effect.
This work has focus on the sympathetic autonomous control mechanism operating at the noise level
of the baroreceptors. The sympathetic system resembles a time discrete signal processing design

168
clocked at the heart beat. The output generates sympathetic bursts. For each heartbeat there either
is a sympathetic burst or there is no sympathetic burst. The probability of having a sympathetic burst
depends on the diastolic pressure level. There is a 1.4 second latency between the sampling of the
diastolic pressure and the corresponding generated output sympathetic burst.
Increased muscular sympathetic nerve activity (MSNA) results in increased blood vessel constriction
by direct nerve path and indirectly by the renin-angiotensin hormone system.
Baroreceptor signal encoding
The pressure dependent baroreceptor signal encodes information in two ways. Systolic pressure is
encoded using frequency modulation and diastolic pressure is encoded using pulse width
modulation. The silence period is extended for low diastolic pressure below the baroreceptor firing
threshold.
External MF at or above the noise level may interfere with the silence period resulting in a encoded
diastolic pressure signal that appears to be higher than the actual diastolic pressure. This may
influence the sympathetic control and result in widening of blood vessels, slightly slowing heart rate
and reduced oxygen supply to the human brain.
Sensor noise level
The noise level of a sensor scales with square root sqrt(n) for n times spatial or temporal averaging.
The human visual system can not only detect light but also magnetic field (magnetophosphenes)
and pressure (pressure phosphenes). Assuming similar technology, the MF threshold of the
baroreceptors may be calculated from the well established visual system magnetophosphene
threshold using temporal (bandwidth) and spatial (area) averaging properties of the two sensors.
The peripheral retina has a rod density of 40'000/mm2. With 1000 converging rods the sensor area
is 1/40 mm2 = 0.025 mm2. The bandwidth of the visual system is 40 Hz and the MF noise-level is
5-10 mT (magnetophosphene threshold).
Carotid arteries with a diameter of 5 mm have 15.7 mm circumference. Assuming 10 mm length, the
baroreceptor sensor area can be estimated to be 157 mm2. The bandwidth of the sensor is in the
order of 1 Hz.
Compared with retina sensors the baroreceptor sensors have relative spatial averaging narea =
157/0.025 = 6283 and relative temporal averaging nbandwidth = 40/1 = 40.
Simple converging sensory neuron network model
We propose a simple converging sensory neural network model with following properties. Non-
linear, discrete time simulation at 1 ms resolution. 1000 input firing signals with modulated firing
probability density function with three stage 10:1 converging coincident detection. If two or more
inputs receive a firing signal within the 1 ms timeslot then a firing event is triggered at the output.
One 50 Hz cycle of 20 ms duration was simulated for an amplitude range of cosine firing probability
density functions.
The simple model was implemented in MATLAB.
Results
Sensor sensitivity threshold
The noise level of a sensor scales with square root sqrt(n) for n times spatial or temporal averaging.
The MF threshold is expected to scale with 1/sqrt(n) = 1/sqrt(narea ꞏ nbandwidth) = 1/sqrt(6283ꞏ40) =
1/500.
Based on the magnetophosphene threshold of 5-10 mT the MF threshold for baroreceptor sensor
can be calculated to be at 5-10 mT / 500 = 10-20 μT.
This threshold is in the same order of magnitude as previously estimated theoretical limit for human
sensitivity to weak fields [2, 22].

169
Simple converging sensory neural network model
The output of the MATLAB simulation is shown in Figure 2 and Figure 3. For modulation below the
sensitivity threshold there are only one or two firing events at the output of the converging network.
For modulation factor 3 the firing rises up to 8 or 9 events. For larger modulation the number of firing
events decreases to one or two.
Discussion
The MF sensitivity threshold of the baroreceptor sensors was calculated from the known phosphene
threshold using simple geometry and bandwidth considerations. The result of 10-20 μT is in the
same order of magnitude as previous theoretical human sensitivity threshold estimation [2, 22].
The MF field may interfere with the silent period of the baroreceptor diastolic pressure measurement
influencing the correct regulation of the blood vessel constriction. If the sympathetic regulation is
influenced or inhibited by MF interference a change in dynamic control behavior may be observed
when properly challenged (e.g. changing ambient temperature or drawing blood).
The simple converging neural network may provide an explanation why influence of MF on
cardiovascular regulation was reported at 10-100μT human studies [1–21] but not observed in
studies with higher MF [2].
Acute cardiovascular effects may reduce cerebral oxygen supply and should be considered in safety
regulations.
Conclusions
Based on phosphene threshold we estimate a magnetic field sensitivity threshold for baroreceptor
and cardiovascular regulation in the 10-20 μT range.
A simple converging neural network model demonstrates amplitude window behaviour as possible
explanation for lack of observed human cardiovascular effect in higher magnetic field exposure
studies.
References
[1] Arsène d’Arsonval. Dispositifs pour la mesure des courants alternatifs de toutes fréquences. In
Comptes rendus des séances de la Société de biologie et de ses filiales, volume 48, pages
450–451. Société de biologie (Paris, France), 1896.
[2] World Health Organization et al. Extremely low frequency fields. 2007.
[3] AA Marino and RO Becker. Biological effects of extremely low frequency electric and magnetic
fields: a review. Physiol Chem Phys, 9(2):131–147, 1977.
[4] CM Maresh, MR Cook, HD Cohen, C Graham, and WS Gunn. Exercise testing in the evaluation
of human responses to powerline frequency fields. Aviation, space, and environmental medicine,
59(12):1139–1145, 1988.
[5] MR Cook, C Graham, HD Cohen, and MM Gerkovich. A replication study of human exposure to
60-hz fields: Effects on neurobehavioral measures. Bioelectromagnetics, 13(4):261–285, 1992.
[6] L Korpinen, J Partanen, and A Uusitalo. Influence of 50 hz electric and magnetic fields on the
human heart. Bioelectromagnetics, 14(4):329–340, 1993.
[7] L Korpinen and J Partanen. The influence of 50 hz electric and magnetic fields on the
extrasystoles of human heart. Reviews on environmental health, 10(2):105–112, 1994.
[8] C Graham, MR Cook, HD Cohen, and MM Gerkovich. Dose response study of human exposure
to 60 hz electric and magnetic fields. Bioelectromagnetics, 15(5):447–463, 1994.
[9] C Graham, MR Cook, DW Riffle, MM Gerkovich, and HD Cohen. Nocturnal melatonin levels in
human volunteers exposed to intermittent 60 hz magnetic fields. Bioelectromagnetics,
17(4):263–273, 1996.
[10] C Graham, MR Cook, DW Riffle, MM Gerkovich, and HD Cohen. Human melatonin during
continuous magnetic field exposure. Bioelectromagnetics, 18(2):166–171, 1997.
170
[11] C Graham, MR Cook, R Kavet, A Sastre, and DK Smith. Prediction of nocturnal plasma
melatonin from morning urinary measures. Journal of pineal research, 24(4):230–238, 1998.
[12] A Sastre, MR Cook, and C Graham. Nocturnal exposure to intermittent 60 hz magnetic fields
alters human cardiac rhythm. Bioelectromagnetics, 19(2):98–106, 1998.
[13] ML Sait, AW Wood, and HA Sadafi. A study of heart rate and heart rate variability in human
subjects exposed to occupational levels of 50 hz circularly polarised magnetic fields. Medical
Engineering and Physics, 21(5):361–369, 1999.
[14] ML Sait, AW Wood, and HA Sadafi. Human heart rate changes in response to 50 hz sinusoidal
and square waveform magnetic fields: a follow up study. In Electricity and Magnetism in Biology and
Medicine, pages 517–520. Springer, 1999.
[15] C Graham, MR Cook, A Sastre, MM Gerkovich, and R Kavet. Cardiac autonomic control
mechanisms in power-frequency magnetic fields: a multistudy analysis. Environmental health
perspectives, 108(8):737, 2000.
[16] Z Tabor, J Michalski, and E Rokita. Influence of 50 hz magnetic field on human heart rate
variablilty: Linear and nonlinear analysis. Bioelectromagnetics, 25(6):474–480, 2004.
[17] E Baldi, C Baldi, and BJ Lithgow. A pilot investigation of the effect of extremely low frequency
pulsed electromagnetic fields on humans’ heart rate variability. Bioelectromagnetics, 28(1):64–68,
2007.
[18] S Rossi, M Hallett, PM Rossini, and A Pascual-Leone. Safety, ethical considerations, and
application guidelines for the use of transcranial magnetic stimulation in clinical practice and
research. Clinical neurophysiology, 120(12):2008–2039, 2009.
[19] NI Adochiei, G Dorffner, and V David. Heart rate variability monitoring due to 50 hz
electromagnetic field exposure and statistical processing. In Electrical and Power Engineering
(EPE), 2012 International Conference and Exposition on, pages 610–613. IEEE, 2012.
[20] JJ Niederhauser, R Schmied, M Baudenbacher, C Schindler, B Litt, and M Röösli. Magnetic field
triggered syncope – from jaques-arsène d’arsoval’s first observation in 1896 to car accident injury
statistics after power-line motorway crossings in switzerland. In BioEM2015 conference, number
PA-39, pages 223–227. Bioelectromagnetics Society, jun 2015.
[21] JJ Niederhauser, R Schmied, M Baudenbacher, C Schindler, and M Röösli. Influence of power-
transmission-lines on car accidents. In BioEM2016 conference, number PA-213, pages 564–569.
Bioelectromagnetics Society, jun 2016.
[22] RK Adair, RD Astumian, and JC Weaver. Detection of weak electric fields by sharks, rays, and
skates. Chaos: An Interdisciplinary Journal of Nonlinear Science, 8(3):576–587, 1998.
Acknowledgments
The author thanks A. Legros and S. Villard for showing magnetophosphenes. The author also
thanks A. Wood for discussing his heart-rate studies.
Figures

171
Figure 1. Human magnetic field exposure studies since 1896 showing effect (check) or no
effect (cross).

Figure 2. Result of simple converging neural network model simulated for a 50 Hz period of 20
ms. Neural network converging from 1000 to 100 to 10 to 1 is shown on the left. The firing
probability density functions (pdf) for modulation factor 3 and 100 and the corresponding
coincidence counts are shown on the right. Modulation factor 3 produced 9 firing events.
Modulation factor 100 produced 2 firing events. There are less firing events at the output for
the larger modulation factor of 100.

172
 Figure 3. Result of simple converging neural network model. For modulation below the
sensitivity threshold there are only one or two random firing events at the output of the
converging network. For modulation factor 3 the firing rises up to 8 or 9 events. For larger
modulation the number of firing events decreases to one or two.

S09-2 [11:45]
A molecular sensor for extremely low frequency magnetic fields: Stromal interaction
molecules (STIM1 and STIM2) as candidate(s) for intracellular sensing of ELF-MF
Myrtill Simko1 & Mats-Olof Mattsson1
1SciProof International AB, Östersund, Sweden, 83132
Keywords: Mechanistic/Theoretical, ELF/LF, Completed (unpublished)
Presented by: Myrtill Simko
Our hypothesis is that ELF-MF induce strongly localized temperature increases or electric fields
within cells and activates STIM1, leading to Ca2+ concentration changes and/or ROS production.
The alternative hypothesis is that ELF-MF activate STIM1 by oxidative stress or electric fields
leading to the activation of downstream effects.
There is evidence that extremely low frequency magnetic fields (ELF-MF) cause biological
responses such as the moderate release of reactive oxygen species (ROS). However, the lack of
mechanistic understanding, and thus also the question of if there are real non-artefactual effects at
low ELF-MF levels, is problematic. It precludes to a high degree research that can unequivocally
determine if there are any longer lasting or even permanent effects of ELF-MF exposure, and also if
such effects have specific relevance for any health condition. The non-understanding of basic
underlying mechanisms thus promotes “effect finding” research where outcomes are dependent on
variables, which are not fully under control by the scientists. This has led to the present situation
where there is a huge body of literature with conflicting outcomes.
It is not clear how the first interaction between ELF-MF and cells takes place although there are
several hypotheses available (see e.g. (Valberg, Kavet and Rafferty, 1997). As one possible

173
explanation for such effects intracellular local temperature changes have been suggested (Simkó,
2004). We hypothesize here that the first interaction takes place at the cell membrane, and/or at an
organelle´s membrane (on its inner or outer side) such as the membranes of mitochondria or the
endoplasmic reticulum (ER). Since it is known that ELF-MF has too low energy to break molecules,
the interaction can possibly be via heat development in the form of intracellular “hot spots”. Other
potential and suggested interaction mechanisms include that he MF causes induced electric fields
that changes the conformations of molecules such as receptors and ion channels. Alternatively, the
interaction could be mediated by oxidative stress related to a radical pair mechanism, or by a
combination of these or other mechanisms, which can induce cell responses. Very fundamental
processes that are affected immediately by any external stimulation include
2+
intracellular/mitochondrial Ca concentrations and redox homeostasis.
Stromal interaction molecules (STIM1 and STIM2) are critical components of the endoplasmic
reticulum Ca2+ channel called store-operated Ca2+ entry (SOCE) and can be activated by
temperature, electric fields, and ROS. Thus, STIM1 is a relevant sensor for the proposed ELF-MF
effects such as local temperature increase and also for ROS production. Therefore our hypothesis is
that ELF-MF exposure causes the activation of the membrane protein STIM1/Orai1, either via local
temperature increase, induced electric fields, or via oxidative stress, which can lead to downstream
effects by activation of signal transduction processes and changes in gene expression. Local
changes in intracellular temperature occur constantly due to normal physiological processes (Okabe
et al., 2012, Tanimoto et al., 2016), and thus STIM1 functions as a ubiquitous sensor and transducer
of temperature-linked events. These processes are fully understood today and are seen as a central
tenet of SOCE, a mechanism by which receptor-mediated depletion of ER Ca2+ stores leads to the
entry of Ca2+ across the plasma membrane.
Thus, our hypothesis is that ELF-MF induce strongly localized temperature increases or electric
fields within cells and so activates STIM1, leading to Ca2+ concentration changes and/or ROS
production. The alternative hypothesis is that ELF-MF activate STIM1 by oxidative stress or electric
fields leading to the activation of downstream effects.
References:
Okabe et al. (2012) Intracellular temperature mapping with a fluorescent polymeric thermometer and
fluorescence lifetime imaging microscopy, Nature Communications. Nature Publishing Group, 3, p.
705.
Simkó M. (2004) Induction of cell activation processes by low frequency electromagnetic fields,
TheScientificWorldJournal, 4 Suppl 2, pp. 4–22.
Tanimoto et al. (2016) Detection of Temperature Difference in Neuronal Cells, Scientific reports.
Nature Publishing Group, p. 22071.
Valberg et al. (1997) Can Low-Level 50/60 Hz Electric and Magnetic Fields Cause Biological
Effects?, Radiation Research, 148 (1), p. 2.

S09-3 [12:00]
In Silico electromagnetic and electrophysiological modelling of ultra-low field nuclear
magnetic resonance based neuronal current imaging
Antonino Mario Cassara1, Jan Storm2, Nora Hofner2, Esra Neufeld1, Rainer Koerber2 & Niels Kuster1
1IT'IS Foundation, Zurich, Switzerland, 8004
2Physikalish-Technische Bundesanstalt, Berlin, Germany, 10587
Keywords: Mechanistic/Theoretical, ELF/LF, Work in Progress
Presented by: Antonino Mario Cassara
Neuronal current imaging’ (NCI) implemented via ultra-low-field magnetic resonance imaging
technologies is a promising candidate for direct, non-invasive detection and imaging of brain
electrical activity. We address effects related to the practical implementation of NCI techniques, e.g.,

174
spatio-temporal patterns of local neuro-magnetic fields, effects of source location, and the influence
of tissue inhomogeneities, relaxation times, anatomy of the human head, etc., combining multi-
disciplinary data from bioelectromagnetics, MRI, computational neuroscience and experimental
efforts to acquire the missing knowledge and promote technological developments for in silico
design of NCI sequences for future in vivo imaging and clinical trials.
Introduction
Magnetic resonance imaging (MRI) techniques, which enable acquisition of anatomical images of
the brain with sub-millimeter resolution, have long permitted the study of brain function for both
research and medical diagnostics purposes. Some limitations of functional MRI (fMRI) involve those
of metabolic origin (i.e. blood oxygenation level dependent (BOLD) signal), with temporal delays of
ca. 1 s, spatial inaccuracy >2 mm, and a dependence on subject-specific vascular structures. To
overcome these limitations, applications of alternative functional methods, such as ‘neuronal current
imaging’ (NCI) have recently been proposed. NCI aims at the direct measurement and localization
of neuronal activity via influence on nuclear magnetic resonance (NMR) signals from neuronal
magnetic fields (NMF) – the local magnetic fields (MF) generated by neuronal currents. While
various methods have been tried for implementation of NCI by means of high-field, i.e., Larmor
fields >1 T, MRI techniques with inconsistent results [Bandettini2005], it appears that the use of
ultra-low-field MRI (ULF-MRI) technologies (i.e. Larmor fields <50 µT) is the most promising.
Between two candidate implementations of ULF-NCI, namely the resonant mechanism
[Cassara2008] and the dc mechanism (dcNCI) [Burghoff2010], for detection of fast (<100 ms) and
slow (>100 ms) components of the brain activities, respectively, dcNCI detection – based on long-
lived neuronal activity – is the more promising on technological grounds [Koerber2013]. However,
many aspects of the dcNCI technique - including the spatio-temporal patterns of local NMFs, effects
of source location, and the influence of tissue inhomogeneities, relaxation times, and anatomy of the
human head, etc. - must be well understood before practical implementations can be developed for
in vivo experiments. We are addressing many of these points by combining a range of multi-
disciplinary data, including bioelectromagnetics, MRI, and computational neurosciences via
simulations and experiments to acquire the missing knowledge for in silico design of dcNCI
sequences for future in vivo imaging and clinical trials.
Methods
NMR solver and electromagnetic models of MRI hardware
In silico simulations of (available and future) dcNCI experiments are created by combining validated
EM models of the superconducting quantum interference device (SQUID) based ULF MR hardware
installed at the Physikalische-Technische Bundesanstalt (PTB, Berlin, Germany) implemented in
Sim4Life (ZurichMedTech AG, Zurich, Switzerland) – see Fig. 1a) with various head phantoms
including modelled current dipole sources that mimic localized neuronal activities and the resulting
MF. For this purpose a flexible NMR solver integrating the Bloch equations is used to calculate the
time domain signal measured by the SQUID sensor, with the MF distributions imported from
validated MRI coil models in presence or absence of the MFs from current sources. This is followed
by typical dcNCI NMR spectrum analysis, permitting the experimental MR signal contributions (e.g.
field in-homogeneities of the MRI coils, pulse sequences used, etc.) to be separated from those
related to the spatiotemporal characteristics of the NMF.
Modeling distributed neuroelectric sources:
We used MF distributions measured by magnetoenephalography (MEG) at PTB and corresponding
equivalent current dipole (ECD) positions and intensities of N20 evoked responses (and subsequent
long-lasting activity) on adult subjects from median nerve stimulation to develop hypothesis for
phantoms geometries and models of neuronal sources (Fig.2) in the anatomical human head
models. However, MEG measurements supply information only on far-field NMF distributions, and
no imaging technique provide information about the EM structure of extended neuronal sources in
vivo. Therefore, to gather insights simultaneously about both the near- and far- NMF distribution of
extended sources (the latter detectable by MEG), we implemented large scale computational

175
investigations by recreating patches of electrically active cortical columns based on realistic
electrophysiological models of independent pyramidal neurons that are thought to be responsible for
the MEG signals. We modelled the patches either as array of aligned neurons for testing purposes
(see figure 3a) as well as physiologically realistic portions of the cortex with neurons in realistic
positions and orientation within the ‘Duke’ anatomical head model (Virtual Population, IT’IS
Foundation, [Gosselin2014], see Fig. 3b) to model hypothetical columnar neuroelectric activity and
calculating its overall ECD. We further calculated the MF generated by single and multiple segment
of currents with the same ECD positioned at the center of the neuronal distributions, to compare the
two field distributions and identify the regions separating the near- and the far-fields and, in the
future, to understand the effects of realistic waveforms of neuronal activities and source positions on
the extent and distributions on the NMR signals. The NMF were calculated using C++ dedicated
libraries interfaced with Sim4Life, the MEG data analysis was performed using the FieldTrip toolbox.
Validation of Source Reconstruction
In addition to calculating the NMF in the near field (i.e. in close proximity to the neuronal sources),
we are calculating the MF generated by realistic and simplified sources within anatomical head
models at the coordinates of 304-multichannel PTB MEG sensor (Fig 3c), to compare measured
field intensities and distributions. Via inverse source reconstruction techniques, the position,
intensities and orientations of the ECD of the modelled sources will be extrapolated to counter-
validate our modelling approach with available MEG measurements.
Validation of the NMR solver
We have further addressed the issue of validation of the NMR solver by modelling and simulating
the pure NMR nuclear procession signals generated by a spherical phantom filled with tissue
simulating liquid (TSL) having long (>1s) T1 and T2 relaxation times in presence and absence of a
gradient field to test the in-silico reproducibility of the effect of MF in-homogeneities on the NMR
signals.
Effect of body anatomy and tissue properties
We are experimentally investigating the dcNCI signal produced by dipolar current sources within
simplified spherical head phantoms with TSL and comparing the results with simulations to quantify
the technological requirements of the technique (e.g. resolution limits, space for technological
improvement in SQUID-based technology, etc.). Although the NMR signals are affected by the NMF
and the constant magnetic permeability of tissues does not distort the MFs, different proton
densities and relaxation times (e.g. T1CSF=344ms, T1GM=103.5 at ULF regimes) in heterogeneous
tissues affect the NMR signals due to partial volume effects. The modelling and experimental
measurement of the dcNCI signals generated by extended sources (from multiple dipoles) within
spherical phantoms are currently in progress. A new phantom miming the head with multiple dipolar
sources is currently being constructed at PTB, while the prediction of the dcNCI signals produced by
electrophysiological neuronal sources is expected in the next months. To calculate the effect of
anatomy and tissues heterogeneity into the dcNCI signals we are working to model different sources
in multi-compartmental spherical head models (featuring skin, scalp CSF and brain tissue) and
within a MRI-based computational anatomical human head model.
Results
The satisfying validation results of the computational EM model of ULF-MRI setup permitted its use
for further accurate investigations. As the validation of the NMR solver is still in progress, we have
so far concentrated on the realistic simulations of extended neuroelectric sources via
electrophysiological models of cortical neurons and the calculation of the NMF in the near and far-
field. Considering that peak ECD of the simulated pyramidal neurons is in the order of ca. 300fAm
and that peak ECD the long-lasting activity is ca 15nAm, we estimated that an extension of the
corresponding source over a diameter of 1-6.4mm, the large variability being due to uncertainties in
the neuronal densities of L5 neurons in different cortical areas, potential effect of misalignment
between neurons due to cortical folding, variability of single-neuron equivalent ECD. From
preliminary data, we observed for both small and large extensions, the far-field NMF is dipolar in

176
agreement with measurements, while the NMF inside the source and within a few mms is very
complex. We are investigating the effects on the NMR signals of either the NMF within the source,
as in its proximity. No standard techniques exists for measuring localized NMF intracortically,
therefore an experimental validation of near-field distribution is difficult. First in situ magnetic fields
measurement of NMF within active cortical layers in cats have been reported in [Caruso2017] using
miniaturized magnetic sensors based on Giant Magneto Resistance (GMR) technologies, and we
aim to compare our estimations with this and similar measurements. The construction of realistic
phantoms with multiple dipoles to mimic the multi-dipolar spatial sources is in progress, as is the
comparison between measured and simulated MEG MF distributions in the 304 MEG channel
system at PTB.
Conclusions
Our study highlights the complexity of developing validated extended electric source models and
phantoms that realistically mimic brain activities for which, for the purpose of dcNCI investigations,
the MEG-based single-dipolar source representation is not sufficient in the near field (where NMFs
may interfere with the NMR signals). The effect of anatomy, e.g. the folding of cortical structure and
the precise localization of the neuronal sources in the patient’s cortex would provide additional
important information to model with improved accuracy the distributed sources together with
information coming from understanding the partial volume effects. This key element will be a
fundamental ingredient to complete plan and execute in silico experiments for the implementation of
the nc-DCI technology. The computational framework we are developing, although unfinished,
permits all aspects related to the implementation of neuro-functional imaging techniques based on
MRI to be modelled. In the future, the framework can be applied to advanced studies related to
biolectromagnetism per se and to validate hypothesis related to the consequences of
neurostimulation e.g. by transcranial electric (TES) or magnetic (TMS) stimulation.
References
Bandettini PA et al. (2005). Direct detection of neuronal activity with MRI: fantasy, possibility, or
reality? App Magn Reson 29:65-88
Burghoff M et al. (2010). On the feasibility of neurocurrent imaging by low-field nuclear magnetic
resonance. App Phys Lett 96: 233701
Cassara AM et al. (2008). Microscopic investigation of the resonant mechanism for the
implementation of nc-MRI at ultra-low field MRI. Neuroimage 41(4): 1228-41
Gosselin M, et al. (2014) Development of a new generation of high resolution anatomical models for
medical device evaluation: the Virtual Population 3.0. Phys Med Biol 59(18), 5287-5303
Caruso L et al. (2017). In vivo magnetic recording of neuronal activity. Neuron, 95(6): 1283-1291
Körber R et al. (2015) An advanced phantom study assessing the feasibility of neuronal current
imaging by ultra-low-field NMR. J Mag Reson 237:182-190.
Figures

177
Figure 1. Left) ULF-MRI hardware at PTB and right) its corresponding EM computational
model.

Figure 2. Measurements of N20 evoked response and subsequent long-lasting activity after
median nerve stimulation in healthy subjects. Left) time course of the MEG signal measured at
one sensor location, and right) distribution of the measured MEG field by the 304 multichannel
system at PTB. The estimation of the neuronal source, provided by inverse source
reconstruction modelling, identifies the source at a depth of 3.5cm below the MEG sensor with
a ECD ca. 50nAm.

178
 Figure 3. A) Computational models of a pyramidal neuron with included anatomical and
biophysical properties. B) model of an extended, planar cortical neuronal source made of
replicas of single neurons, with illustrated membrane potential along the neuronal bodies and
(on the background) the resulting calculated NMF; c) modelling of in vivo neuronal sources
using neuro-functionalized portions of the cortex (gray matter) of the Duke ViP model with
electrophysiological models of neurons in realistic neuronal orientations and positions. D)
Complete modeled MEG-MRI setup with the anatomical head model with sources distributed
in the first somatosensory area, used to combine electrophysiological modelling with MEG
signals and influence of neuronal magnetic fields on NMR signals. The MEG sensor is shown
only for modeling purposes (not part of the same hardware). E) Reconstructed MEG field
distribution from the simulated neuronal sources in one test sensor.

S09-4 [12:15]
o2S2PARC – Open Online Simulations for Stimulating Peripheral Activity to Relieve
Conditions
Esra Neufeld1, Nicolas Chavannes1, Antonino Mario Cassara1, Bryn Lloyd1, Pedro Crespo-Valero1, Manuel
Guidon1, Odei Maiz1, Tobias Oetiker1 & Niels Kuster1, 2
1IT'IS Foundation for Research on Information Technologies in Society, Zürich, Switzerland, 8004
2Swiss Federal Institute of Technology (ETHZ), Zürich, Switzerland, 8092
Keywords: Mechanistic/Theoretical, All Frequencies, Work in Progress
Presented by: Niels Kuster

179
The NIH SPARC initiative aims to understand nerve-organ interactions and “advance the
neuromodulation field towards precise treatment of diseases”. The aim of o2S2PARC is to establish
an innovative, open online platform to host, connect, and execute computational models. It is
centered on enhanceable, neuro-, physiology-, and data-functionalized computational anatomical
models and uses a mechanistic, electrophysiological perspective to investigate PNS influence on
organ physiology and device-driven neuromodulation. The final software will be a crucial
contribution towards open source simulation platforms for the investigation of electroceuticals,
neuroprosthetics research, and other neuromodulation-based therapies.
Introduction
In recent years, an increased understanding of the nervous system and human electrophysiology,
as well as advances in neuro-engineering, have opened up exciting new avenues for medical
devices that actively influence neural function. The human nervous systemconsists of two
components: the central nervous system (CNS) and the peripheral nervous system (PNS). The PNS
is divided into the somatic nervous system, which primarily handles sensory input and controls
voluntary body movements, and the autonomic nervous system, which influences inner organ
physiology. While the somatic nervous system is targeted to restore lost sensory and/or motor
function (usually referred to as neuroprosthetics), the autonomic nervous system is modulated to
affect organ function and relieve conditions (usually referred to as bioelectronics medicine or
electroceuticals). While electromagnetic (EM) exposure is the most frequent stimulation mechanism,
low intensity focused ultrasound, optical, and thermal exposure are also employed. For example,
tiny implants are envisioned which can be applied to treat chronic diseases such as diabetes,
asthma, hypertension, gastrointestinal, or even immune-response related diseases. In fact, several
devices for vagus nerve stimulation (VNS) have already been approved by regulators for the
treatment of depression. VNS is further investigated for the treatment of obesity, Alzheimer’s,
migraine, addiction, and heart failure, illustrating the vast potential of electroceuticals, but also the
difficulty of ascertaining effective therapeutic impact while avoiding unwanted side-effects. This is
partly due to the complexity of the nervous system, its network structure, the non-linearity of
responses, the multi-scale sophistication of the body internal environment, and a frequently limited
understanding of interaction mechanisms and organ physiology.
An improved understanding of neural dynamics, information encoding, nerve mapping, organ
electrophysiology across a wide range of scales, interaction mechanisms, and the design of safe,
effective, selective, and miniaturized implantable devices is crucial so that electroceuticals can
unfold their full potential. Two important efforts towards achieving this goal emerged in the last
years: (i) the rapid progress in computational life sciences (CLS) which enables increased model
realism and level of detail through the steep increase in available computational power/resources
and methodological advances (e.g., in simulation technology, high performance computing (HPC),
and medical imaging), and (ii)the SPARC initiative, funded by the National Institutes of Health (NIH),
which aims to “transform the actual understanding of nerve-organ interactions” and “advance the
neuromodulation field towards precise treatment of diseases and conditions for which conventional
therapies fall short”.
The multi-disciplinary SPARC initiative involves (i) anatomical and functional mapping of internal
organ innervation, (ii) development of next generation tools and technologies (experimental,
computational, devices…), (iii) support of translational work, and (iv) the creation of a Data and
Resource Center (DRC) to coordinate the infrastructure (storage, computational), the modeling
across SPARC, as well as the organization and mapping of all SPARC-created data and models. In-
silico tools will support the investigation of all above-mentioned factors and, in particular, the study
of interaction mechanisms, device/targeting/selectivity optimization, assessment and minimization of
side-effects and risks, treatment personalization, and closed loop control. One key component of the
SPARC DRC is a free, online accessible platform developed by the IT’IS Foundation, named
o2S2PARC, to support modeling-related activities across the SPARC community. Through the
platform, researchers will be able to study the interaction of physical stimuli with the human body

180
and its physiology via multi-physics, multi-scale simulations involving detailed anatomical human
body models that have been functionalized with dynamic nerve, tissue, and organ models.
Methods
o2S2PARC is an online framework to create, host, and connect computational models of organs and
tissues, as well as solvers to execute multi-physics simulations centered on the anatomical models,
and analysis modules. The modules are partly provided by IT’IS (e.g., an EM solver, a coupled
neuron electrophysiology solver, functionality for uncertainty assessment and propagation) or other
SPARC teams (e.g., a cardiac electrophysiology model), as well as 3rd parties. The platform is
developed open source, while providers of modules can control their licensing conditions. Modules
will run within micro-services (Docker containers enhanced, e.g., with communication infrastructure)
locally or in the cloud under the coordination of ‘Director’ services, and will be accessed through a
user-friendly web-based graphical user interface (GUI) with powerful visualization capabilities. At the
center of o2S2PARC are detailed, PNS-functionalized human and animal anatomical models
(NEUROCOUPLE and NEUROFAUNA) which will also serve as integration centers (spatial
localization and environment/context for physical interaction modeling) for computational models
and measurement data from the SPARC community and 3rd parties. The modeling and model
coupling will be primarily originating from a mechanistic electrophysiology perspective (which does
not exclude, e.g., machine learning-based modules). The platform will also include support for
image-based modeling and meta-modeling (e.g., inverse problem solving or optimization). Quality
assurance is ascertained through various measures, such as ensuring complete chain of custody,
traceability, reproducibility (for example by storing not only the simulation setup and parameters, but
also images of the employed simulator micro-services), transparent verification/validation status,
and support for certification. Central to this is the integration with the DRC DAT-CORE, which
develops the database infrastructure and through which data will typically be routed before entering
the SIM-CORE. HPC-modules will be supported.
Results
The requirements of SPARC users have been collected and used as a basis for detailed
specifications. Prototypes of the o2S2PARC platform, covering the full stack (front-end, back-end,
compute services, communication), have been realized to identify technological challenges and
compare alternative implementation approaches and technologies. Critical solvers (e.g., for
electromagnetic and coupled neuronal dynamics modeling) have been developed, carefully verified
against reference models, and successfully applied within the Sim4Life platform (ZMT Zurich
MedTech AG) to study a range of therapeutic applications. Functionalization of detailed anatomical
model with peripheral nerves is progressing well, and specialized software has been developed to
support cryo-section image-based nerve tracing. Corresponding anatomical images of a human and
a monkey have been acquired. First organ physiology models from two SPARC teams could be
executed within containerized services and the use of Sim4Life as a service has been explored.
Conclusions
The aim of o2S2PARC, part of the NIH SPRAC initiative, is to establish an innovative, open online
platform to host, connect, and execute computational models. It is centered on enhanceable, neuro-,
physiology-, and data-functionalized computational anatomical models and uses a mechanistic,
electrophysiological perspective to investigate PNS influence on organ physiology and device-driven
neuromodulation for therapeutic purposes. The evaluated technologies demonstrate that the
realization of this ambitious vision is feasible. The final software will be a crucial contribution towards
open source simulation platforms for the investigation of electroceuticals, neuroprosthetics research,
and other neuromodulation-based therapies. In addition, it opens the path for online tools for many
other applications of CLS.

181
 Session: S10
Human studies II
Thursday June 28, 2018 • 11:00 - 12:30
Europa C
Chairs: Alexandre Legros & Heidi Danker-Hopfe

S10-1 [11:00]
STUDENT PAPER
Investigating the determinants of IEI-EMF: Is the nocebo effect a normal response?
Adam Verrender1, 2, Sarah Loughran1, 2, 3, Anna Dalecki1, 2, 3, Frederik Freudenstein1, 2, 3 & Rodney Croft1,
2, 3
1Australian Centre for Electromagnetic Bioeffects Research, University of Wollongong, Wollongong, Australia
2School of Psychology, Illawarra Health & Medical Research Institute, University of Wollongong, Wollongong,
Australia
3Population Health Research on Electromagnetic Energy, Monash University, Melbourne, Australia
Keywords: Human, RF/Microwaves, Completed (unpublished)
Presented by: Adam Verrender
While there has been consistent evidence that the symptoms experienced by IEI-EMF sufferers are
likely the result of a nocebo response, much remains to be understood. The present study
investigated whether perceived EMF exposure would elicit symptoms in a healthy population, and
assessed whether messages that emphasize ‘adverse health effects of EMF exposure’ can alter this
nocebo response. The results revealed 1) the crucial role of awareness and belief of exposure in the
presentation of symptoms during perceived exposure to EMF; 2) that healthy participants exhibit a
similar nocebo response to those who suffer from IEI-EMF; and 3) that alarmist media reports raise
concerns about EMF and may contribute to a symptomatic nocebo response.
Introduction
Although there has been consistent evidence that the wide range of non-specific symptoms reported
by individuals who suffer from Idiopathic Environmental Intolerance attributed to Electromagnetic
Fields (IEI-EMF) are more closely associated with a nocebo response [1-3], whether this response
is ‘normal’ and the factors which contribute to this response remain unclear. A nocebo response
generally occurs when conscious or subconscious negative expectations trigger or exacerbate
adverse symptoms in response to an exposure that is not known to otherwise cause those effects
[4, 5]. While there has been some suggestion that mainstream media reports [6] and science
communications [7, 8] negatively influence people’s beliefs about EMF exposure, only a limited
number of studies have demonstrated that this may contribute to the presentation of symptoms via a
nocebo response. For instance, Witthöft and Rubin [6] found that viewing an sensationalised media
report about the adverse effects of Wi-Fi exposure increased the likelihood of participants
experiencing symptoms following a sham exposure. However, the effect was only found for those
with high pre-existing levels of state anxiety. This may be because the study lacked a verified non-
exposure condition, potentially resulting in insufficient statistical power to detect effects in non-
anxious individuals. In support of this, Bräscher, et al. [9] found that participants who watched an
alarmist film perceived tactile electrical stimuli as more intense during a cued Wi-Fi (sham) exposure
compared to a cued no Wi-Fi condition (a non-exposure condition), and that the effect was not
mediated by anxiety. However, as the latter study assessed somatosensory perception rather than
symptom perception, it remains uncertain as to whether negative beliefs induced by information
about EMF exposure can result in a symptomatic nocebo response following a perceived exposure
to EMF.
To address this uncertainty, the present study was designed to investigate whether perceived EMF
exposure would elicit symptoms in a healthy population, and to assess whether messages that
emphasize ‘adverse health effects of EMF exposure’ can induce a nocebo response. In the
182
provocation trials, both active and sham EMF exposures were first demonstrated to participants in
an initial non-blinded, open-label trial, before a series of double-blind, randomized, counterbalanced
trials were conducted. By demonstrating that the experiment contained a ‘no exposure’ condition,
the present study was able to more clearly determine whether any observed increases in symptoms
were the result of a nocebo effect.
Materials & Methods
Participants
Forty-four healthy participants aged 18 – 30 years (M = 21.92, SD = 4.88; half male) participated in
the study.
Radiofrequency Exposure
RF exposure was generated using a portable, self-contained, battery-operated device (Two Fields
Consulting, St Kilda, Australia). The exposure was an RF signal in the 902-928 MHz ISM band
which was digitally modulated in a similar manner to signals from Wi-Fi and 3G/4G mobile phones.
The signal was generated by a commercial RF modem which emits a frequency hopping spread
spectrum signal with an average radiated power output of 1 W (estimated power density at the
participant is 0.3 W/m2, during the RF-ON trials) or was completely EMF silent (RF-OFF trials &
sham trials).
Questionnaires
Participants were asked to indicate belief of exposure and symptom severity via pen and paper
100mm visual analogue scales (SESS). For belief of exposure, participants were asked “how sure
are you of the current exposure status right now?” anchored with the terms ‘Definitely Off’ and
‘Definitely On’. To assess symptoms, a modified version of the 34 item checklist for symptoms in
daily life was used, anchored with the terms ‘Barely Detectable’ and ‘Maximum Severity’. The 40
item version of the State Trait Anxiety Index (STAI) was used to assess participant’s state and trait
anxiety. Risk perception was assessed by a 4 item questionnaire (RPQ).
Design and Procedure
Participants were seated comfortably inside a Faraday cage, where they completed baseline (Time
1 (T1)) measures of the STAI and RPQ. Each participant then watched either an alarmist or control
video (randomly assigned group, 22 participants per group, half male). Participants in the alarmist
video group viewed a 3 min video appeal to the United Nations from a concerned scientist which
included statements about the potential health risks from mobile phone and Wi-Fi signals and
sensationalized images of exposure scenarios [10]. Participants in the control group viewed a 3 min
segment of a documentary on gravity [11] which contained no health-related content. After watching
the video, participants completed Time 2 (T2) measures of the STAI and RPQ. Following this,
participants completed a series of 14 sham and active provocation trials where their symptoms and
belief of exposure status were measured. The first two trials were non-blinded, open-label trials (1
RF-OFF, 1 RF-ON) where the exposure conditions were demonstrated to the participants using a
dosimeter. These trials were followed by a series of 12 double-blind, randomized, counterbalanced
trials (6 sham, 6 RF-ON). Each trial began with a 5-minute baseline interval, followed by a 10-minute
exposure interval and concluded with a 5-minute post-trial interval. The SESS was completed 2.5
mins into the baseline interval and again 7 mins into the exposure interval.
Statistical Analysis
The primary dependent variables for belief of exposure status and symptoms were calculated as
difference scores between the baseline and exposure questionnaires. Parametric, and where
appropriate non-parametric, statistics were used to determine whether there was: (1) a main effect
of exposure on symptoms in the open-label trials; (2) main effects of exposure on belief of exposure
or symptoms in the double-blind trials; (3) an interaction between video group and either belief of
exposure or symptoms in the open-label or double-blind trials; (4) a relationship between belief of
exposure and symptoms in the double-blind trials (irrespective of actual exposure condition); and (5)
a difference in STAI and RPQ score between the two video groups. A number of other preliminary

183
and exploratory analyses were also conducted, and will be discussed further in the presentation.
Results
As shown in Figure 1a, symptoms were higher in the open-label RF-ON than RF-OFF trial (p < .001,
Effect Size (ES) = 0.71). No difference in either symptoms (p = .183 ES = 0.14) or belief of exposure
(p = .144, ES = 0.17) was observed between the two conditions in the double-blind trials.
Participants in the alarmist video group had a significant increase in symptoms (p = .041, ES = 0.26)
(see Figure 1b), state anxiety (p <.01, ES = 0.38) and risk perception (p < .001, ES = 0.60) relative
to the control video group. Spearman’s rho measure of association showed that the relationship
between belief of exposure and symptoms was highly variable between participants; ranging from
-.276 to .882 in the control video group and -.675 to .852 in the alarmist video group. A one-sample t
test found that the transformed rho values for the combined groups were significantly greater than 0
(p < .001, ES = 1.03).
Discussion and Conclusions
In the open-label trials, participants reported higher symptom scores in the RF-ON trial compared to
the RF-OFF trial. In the subsequent double-blind trials, no difference in either belief of exposure or
symptoms was found between the RF-ON and sham conditions. This demonstrates that knowledge
and/or awareness of the exposure condition was essential for producing an effect. In addition to this,
belief of exposure was found to be positively associated with higher symptom scores in the double-
blind trials, giving further indication that a nocebo effect, rather than EMF exposure itself, was
responsible for the increase in symptoms. Participants who viewed the alarmist video also had
higher symptom scores in the open-label trials and a greater increase in state anxiety and risk
perception than those who viewed the control video.
Overall, the results of the present study show that healthy participants exhibit a similar response to
exposure to those who experience IEI-EMF [3], which suggests that the nocebo response displayed
by IEI-EMF sufferers may be a normal human response. Further support was also found for the
crucial role of awareness and/or belief in the presentation of symptoms during exposure to EMF. In
addition, the present study demonstrated that sensationalist media reports about perceived
environmental hazards not only raise concerns and negative beliefs about EMF, but may also be
contributing to a symptomatic nocebo response. These results have implications for the
development of appropriate treatments for those who experience IEI-EMF, as well as for the way in
which scientific evidence is communicated in the community.
Acknowledgments
This research was conducted with the support of the Australian Government Research Training
Scholarship and the National Health and Medical Research Council of Australia [Grant 1042464].
References
[1] Rubin, G.J., Nieto-Hernandez, R., Wessely, S., Idiopathic environmental intolerance attributed to
electromagnetic fields (formerly 'electromagnetic hypersensitivity'): An updated systematic review of
provocation studies, Bioelectromagnetics 31 (2010) 1-11.
[2] van Moorselaar, I., Slottje, P., Heller, P., van Strien, R., Kromhout, H., Murbach, M., Kuster, N.,
Vermeulen, R., Huss, A., Effects of personalised exposure on self-rated electromagnetic
hypersensitivity and sensibility – A double-blind randomised controlled trial, Environment
International 99 (2017) 255-262.
[3] Verrender, A., Loughran, S.P., Anderson, V., Hillert, L., Rubin, G.J., Oftedal, G., Croft, R.J., IEI-
EMF provocation case studies: A novel approach to testing sensitive individuals,
Bioelectromagnetics 39 (2018) 132-143.
[4] Hahn, R.A., The nocebo phenomenon: Concept, evidence, and implications for public health,
Preventive Medicine 26(5) (1997) 607-611.
[5] Bräscher, A.-K., Kleinböhl, D., Hölzl, R., Becker, S., Differential classical conditioning of the

184
nocebo effect: Increasing heat-pain perception without verbal suggestions, Frontiers in Psychology
8(2163) (2017).
[6] Witthöft, M., Rubin, G.J., Are media warnings about the adverse health effects of modern life
self-fulfilling? An experimental study on idiopathic environmental intolerance attributed to
electromagnetic fields (IEI-EMF), Journal of Psychosomatic Research 74 (2013) 206-212.
[7] Wiedemann, P.M., Schuetz, H., Boerner, F., Clauberg, M., Croft, R., Shukla, R., Kikkawa, T.,
Kemp, R., Gutteling, J.M., de Villiers, B., da Silva Medeiros, F.N., Barnett, J., When precaution
creates misunderstandings: The unintended effects of precautionary information on perceived risks,
the EMF case, Risk Analysis 33(10) (2013) 1788-1801.
[8] Wiedemann, P.M., Boerner, F.U., Repacholi, M.H., Do people understand IARC's 2B
categorization of RF fields from cell phones?, Bioelectromagnetics 35(5) (2014) 373-378.
[9] Bräscher, A.-K., Raymaekers, K., Van den Bergh, O., Witthöft, M., Are media reports able to
cause somatic symptoms attributed to WiFi radiation? An experimental test of the negative
expectation hypothesis, Environmental Research 156 (2017) 265-271.
[10] Blank, M., International Scientist Appeal on Risks of Electromagnetic Fields Presented to the
United Nations, Available online: https://vimeo.com/123468632, 2015.
[11] Cox, B., Size Matters, in: A. Cohen (Ed.) Wonders of Life, British Broadcasting Corporation,
2013.
Figures

185
Figure 1. The mean (dot), median (line), interquartile range (box) and range (whiskers) are
shown for a) the SESS symptoms scores for the RF-OFF and RF-ON open-label trials; and b)
the difference in SESS symptom score (RF-ON – RF-OFF) as a function of video group.

186
S10-2 [11:15]
Extremely low frequency electric and magnetic stimulations on the vestibular performance in
humans
Sebastien Villard1, 2, Nicolas Bouisset1, 3, Daniel Goulet4, Michel Plante4, Martine Souques5, François
Deschamps6, Genevieve Ostiguy4, Jacques Lambrozo5 & Alexandre Legros1, 2, 3, 7, 8
1Human Threshold Research Group, Lawson Health Research Institute, London, Ontario, Canada
2Medical Biophysics, Western University, London, Ontario, Canada
3Kinesiology, Western University, London, Ontario, Canada
4Hydro-Québec, Montréal, QC, Canada
5Service des Études Médicales, EDF, Paris, France
6Service Environnement Réseaux, RTE, Paris, France
7Medical Imaging, Western University, London, ON, Canada
8EuroMov, Université de Montpellier, Montpellier, France
Keywords: Behavioural, ELF/LF, Work in Progress
Presented by: Sebastien Villard
The study of magnetophenes suggests that sensory systems interact with induced electric fields
generated by time-varying magnetic stimulation. Postural control literature, also suggests that the
vestibular system is sensitive to induced electric fields. Yet, postural control is an indirect indicator of
the vestibular performance. We are proposing to probe the vestibular function through the
measurement of the Subjective Visual Vertical (SVV). Participants were exposed to 4 frequencies of
stimulation (20, 60, 120 and 160 Hz) at constant maximum dB/dt of 14.2 T.s-1 in the Magnetic Field
condition and 2 mA in the electric condition. No effects of the stimulations on SVV are reported so
far.
Introduction
Knowledge of acute behavioral modulations resulting from extremely low frequency magnetic field
(ELF MF) is crucial to assure the safety of power line workers and general public in close vicinity to
any electrical sources. To date, international organizations1–3, establish their guidelines on the only
reliable behavioral acute response to ELF MF: the perception of flickering lights in the visual field,
named magnetophosphene. The ability of the graded potential cells of the retina, to pick up small
amount of information, and transduce it continuously into electric signal, makes them the likely
target of the effect of time-varying MF4. Graded potential cells are also found in other sensory
systems such as the vestibular system. The vestibular hair cells are the graded potential cells
transducing the mechanical outcome of the acceleration of the head into electric signal carried by
the vestibular nerve and interpreted by the brain. The vestibular system has been pointed out as a
potential sensitive structure to magnetic stimulation because of the impact of such stimulation on
postural control5–7. Yet, postural control is achieved through the integration of multiple arrays of
information and it is often difficult to single out the specific influence of the vestibular function.
Two vestibular subsystems, the otolith organs and the semicircular canals, are respectively sensitive
to linear and rotational accelerations of the head. In humans, the judgment of verticality, is greatly
impacted by the perception of linear accelerations provided by the otolith organs. Therefore, the
subjective visual vertical (SVV), a psychophysical measure of the angle between perceptual vertical
and “true” (gravitational) vertical, can be used as a more specific marker of the vestibular function8.
Moreover, it has been clearly described that the vestibular system can be stimulated via electrical
stimulation through either direct or alternative current, also known as galvanic vestibular stimulation
or GVS9. Zink et al. have shown that the otolith function can be measured at current lower than 3
mA10. This suggest that the SVV is an appropriate functional marker to assess the performance of
the vestibular function at low intensity of electric stimulation. Previous studies assessing the
perception of verticality while stimulated with direct current showed a misperception in the direction

187
of the anodal stimulation11–13. Therefore, we are proposing to evaluate the impact of time-varying
MF below 300 Hz on the vestibular performance by assessing modulations in the SVV in healthy
adults.
Furthermore, the work of Lövsund14has shown that the change in induced electrical field is
proportional to the change in frequency for an alternative magnetic stimulation but not for an
electrical stimulation source. We attend to address the effect of frequency changes by exploring 4
frequencies of stimulation for both galvanic vestibular stimulation and ELF MF stimulation. To do so,
we propose to change the frequency of the ELF MF stimulation while adapting the magnetic flux
density in order to always deliver the same variation of flux density over time (dB/dt) and keep the
induced electrical filed similar for all frequencies.
Methods
13 participants (6 males, 25.2 ± 6 years, 72.9 ± 17 kg, 178.1 ± 10 cm – the final protocol will include
60 participants) took part of the experiment so far. They were asked to stay sited on an stool during
the time of the experiment. MF exposure was delivered using a single coil at the level of the mastoid
process on the left side of the head. Binaural bipolar electric stimulations were delivered on the
mastoid processes (Anode-Left, Cathode-Right). Participants were exposed to: i) A SHAM
stimulation, where no stimulation was delivered. ii) A Direct Current Stimulation (DC) at 2 mA
(positive control). iii) An Alternating Current Stimulation (AC) at 2 mA given at 4 frequencies (20 Hz,
60 Hz, 120 Hz, and 160 Hz). iv) A sinusoidal MF applied via a custom coil at 80 mT, 26.67 mT, 13.33
mT, and 10 mT (estimated at the level of the vestibular system) for respectively 20 Hz, 60 Hz, 120
Hz, and 160 Hz. Flux densities were chosen to match the same dB/dt for the different frequency
conditions: a constant maximum dB/dt of 14.2 T.s-1.
Experimental exposure conditions were administered for 25 s after 90 s of rest. During the
stimulation participants were asked to look through a cone to a monitor displaying a dotted line over
a black background oriented along 8 specific angles: ±4º, ±3º, ±2º, and ±1º. The line was visible for
250 ms, and participants were asked to follow these instructions:
“You will see in front of you a line tilted toward the right or the left. When you see the line, press the
right button if the line is tilted toward the right side, or press the left button if the line is tilted toward
the left. When you are not sure, you must choose between right or left”
These 8 angles were randomly distributed. Each condition of stimulation was repeated 4 times. All
conditions were presented in a pseudo-randomized order, where higher flux density conditions (i.e.
20Hz – 80 mT) were distributed at equal time interval from each other.
MF exposure was delivered via a custom exposure system consisting of a 176-turn coil (16 layers of
11 turns each, 6 cm inner diameter and 22 cm outer diameter) made of hollow square copper wire
cooled by circulating water and powered by a MTS™ Magnetic Resonance Imaging (MRI) gradient
amplifier array capable of delivering up to 200 Arms at ± 345V (MTS Automation, Horsham, PA,
USA). The DC- and AC-GVS were delivered using a Starstim system (StarStim, Neuroelectrics,
Spain).
The data collected were Button Presses (0 for left and 1 for right) for each angle of tilted lines (figure
1, left panels, black dots). For each participant and each experimental condition, button presses
were modeled into a psychometric function (logistic regression) to obtain the probability of pressing
the right button for each angle (figure 1, left panels blue curve). From this model we extracted the
theoretical angle for which there is 50% chance to press the right button, this value (Bias) was used
to determine the SVV. We also extracted the slope of the model (Scale) giving us information about
the difficulty to discriminate left from right. We also gathered the reaction time between the
appearance of the line on the screen and the button presses as well as the number of errors made
for each trials (i.e, simultaneous left and right press, several presses for one line, or no press).
Results
Neither the Bias (mean: -0.37º±0.98), nor the Scale (mean: 1.68±0.49) showed any significant

188
difference between the different stimulation (respectively, F3,26=0.91, p=0.44, and F3,36=1.41,
p=0.25). The number of errors did not show any significant differences between stimulations
(F4,376=1.25, p=0.29). However, the reaction time showed significance difference between
stimulations (F4,376=2.64, p<0.05, partial eta2=0.45)
Discussion
In the current state of advancement, it is precarious to discuss the results over only 13 participants.
Yet, the findings presented here cannot discriminate the effect of the DC stimulationon the SVV. A
preliminary pilot study over 10 participants suggested a SVV towards the DC anodal side as
expected from previous research, but a thorough investigation of our current findings is necessary.
Few points are important to consider. First, all participants stated to have perceived the DC
stimulation and exhibited expected postural tilt towards the anodal side during familiarization
(standing up with eyes closed prior to the experiment), which confirmed the postural effect of the DC
stimulation. Secondly, the Reaction time results suggest that the DC stimulation is indeed perceived
has more difficult to perform the verticality assessment.
The methodology of 2 Alternative Forced Choice (2AFC) presented here to quantify the SVV was
preferred for it has been shown to be reliable in its results and avoid experimental bias from other
methodology like the “bucket test”. However, besides encouraging pilot results it is possible that the
2AFC method is not sensitive enough to discriminate the DC effect on SVV presented in other
studies.
Finally, results showed that the Biases are always directed towards the left side in the SH
conditions. The same results have been reported in Santos et al.12. In both Santos et al. work and
our study participants were almost exclusively right-handed (respectively 100% and 93% of the total
participants). The literature suggest that vestibular information is mainly processed by the
nondominant hemisphere15,16which could explain this asymmetry in our findings. Left-handed
participants should be included to test this hypothesis and handedness added as a covariate in our
statistical analyses.
Results over 20 participants will be presented at the conference and any acute effects would be
discussed in regard to the location of the different stimulations (more diffused for the case of electric
stimulations) and the amplitude of the produced electric field at the level of the vestibular system.
Reference
1. International Commission on Non-Ionizing Radiation Protection. Guidelines for limiting exposure
to time-varying electric and magnetic fields (1 Hz to 100 kHz). Health Phys. 99, 818–36 (2010).
2. IEEE. C95.6. IEEE standard for safety levels with respect to human exposure to electromagnetic
fields, 0-3kHz. (The institute of Electrical and Electronics Engineers, Inc., 2002).
3. World Health Organization. in World Health 543 (2007).
4. Attwell, D. Interaction of low frequency electric fields with the nervous system: the retina as a
model system. Radiat. Prot. Dosimetry 106, 341–8 (2003).
5. Glover, P. M., Cavin, I., Qian, W., Bowtell, R. & Gowland, P. A. Magnetic-field-induced vertigo: A
theoretical and experimental investigation. Bioelectromagnetics 28, 349–361 (2007).
6. van Nierop, L. E., Slottje, P., Kingma, H. & Kromhout, H. MRI-related static magnetic stray fields
and postural body sway: A double-blind randomized crossover study. Magn. Reson. Med. 70,
232–240 (2013).
7. van Nierop, L. E., Slottje, P., van Zandvoort, M. & Kromhout, H. Simultaneous exposure to MRI-
related static and low-frequency movement-induced time-varying magnetic fields affects
neurocognitive performance: A double-blind randomized crossover study. Magn. Reson. Med. 74,
840–849 (2015).
8. Akin, F. & Murnane, O. Subjective Visual Vertical Test. Semin. Hear. 30, 281–286 (2009).

189
9. Fitzpatrick, R. C. & Day, B. L. Probing the human vestibular system with galvanic stimulation. J.
Appl. Physiol. 96, 2301–16 (2004).
10. Zink, R., Bucher, S. F., Weiss, A., Brandt, T. & Dieterich, M. Effects of galvanic vestibular
stimulation on otolithic and semicircular canal eye movements and perceived vertical.
Electroencephalogr. Clin. Neurophysiol. 107, 200–205 (1998).
11. Mars, F., Popov, K. & Vercher, J. L. Supramodal effects of galvanic vestibular stimulation on the
subjective vertical. Neuroreport 12, 2991–2994 (2001).
12. Santos-Pontelli, T. E. G. et al. Polarity-dependent misperception of subjective visual vertical
during and after transcranial direct current stimulation (tDCS). PLoS One 11, 1–14 (2016).
13. Volkening, K. et al. Verticality perception during and after galvanic vestibular stimulation.
Neurosci. Lett. 581, 75–79 (2014).
14. Lövsund, P., Öberg, P. Å. & Nilsson, S. E. G. Magneto- and electrophosphenes: A comparative
study. Med. Biol. Eng. Comput. 18, 758–764 (1980).
15. Dieterich, M. & Brandt, T. Functional brain imaging of peripheral and central vestibular disorders.
Brain 131, 2538–2552 (2008).
16. Dieterich, M. & Brandt, T. Imaging cortical activity after vestibular lesions. Restor. Neurol.
Neurosci. 28, 47–56 (2010)
Figures

190
Figure 1. Probability of Right button press over the angle of the presented dotted line. The left
panels present the difference between stimulation (SH for Sham, DC for Direct Current AC for
Alternative Current, and MF for Magnetic field) The size of black dots represents the count (n)
of each response per angle (1 being Right, 0 being Left). The right panels present the
difference in Frequency of stimulation for AC (top) and MF (bottom)

191
 Figure 2. Mean Reaction Time and standard error of the mean (error bars) between line
appearance and button press for each conditions. Tuckey post-hoc analysis show that DC
tend to be greater than SH and MF, but do not reach significant level

S10-3 [11:30]
Collaborative development of an innovative provocation protocol in studying
electrohypersensitivity
Maryse Ledent1, Maël Dieudonné2, Jimmy Bordarie3, Willy Pirard4, Benjamin Vatovez4, Christophe
Geuzaine5, Veronique Beauvois5 & Luc Verschaeve1
1Risk Assessment, Scientific Institute of Public Health (WIV-ISP), Brussels, Belgium, 1050
2Centre Max Weber (CMW), Lyon, France, 69007
3Université Sorbonne Nouvelle, Paris, France, 75231
4Institut Scientifique de Service Public (ISSeP), Liège, Belgium, 4000
5Applied and Computational Electromagnetics (ACE), Université de Liège, Liège, Belgium, 4000
Keywords: Human, All Frequencies, Work in Progress
Presented by: Maryse Ledent
Electrohypersensitivity syndrome (EHS) is characterised by a variety of non-specific symptoms that
are attributed to electromagnetic fields (EMF) by electrohypersensitive people (EHP). However,
despite the great distress of certain individuals, symptoms cannot be objectively attributed to EMF
so far: results of provocation studies converge to the conclusion that EHP do not perceive and do
not respond physiologically to EMF. However, a critical review of these studies reveals that several
parameters have not been controlled or considered. An original provocation test is being developed.
To meet acceptability criteria an iterative co-creation process involving EHP, researchers and

192
technical experts is implemented.
Introduction
Electrohypersensitivity syndrome (EHS) constitutes a relatively new public health issue. It is
characterised by a variety of non-specific symptoms that are attributed to electromagnetic fields
(EMF) by electrohypersensitive people (EHP). However, despite the great distress of certain
individuals, symptoms cannot be objectively attributed to EMF so far: epidemiology and provocation
studies have not allowed yet establishing a direct causal link with the appearance of symptoms
(Rubin et al., 2010, 2011).
Some researchers concluded that the nocebo effect is the probable cause of EHS (Regel et al.,
2006; Oftedal et al., 2007; Rubin et al., 2010). In this context, the future of provocation studies is
questioned: could they still provide advances in the knowledge?
Methodological limits are expressed by EHP and associations: inadequate inclusion criteria,
diversity of individual sensitivities, latency in the appearance of symptoms, health status before
exposure… Highlighting several drawbacks, Leszcynski (2018) stated that even an improved
methodology would not allow providing objective and reliable scientific data.
However, EHP remain convinced of the relevance of provocation studies. Their results appear even
more unacceptable to EHP so as their conviction of the electromagnetic origin and the results of
their own experiments (Dieudonné, 2016; Prignot, 2016). Moreover, it has been observed that
symptoms often occur before establishing a link with EMF (Bergqvist & Vogel, 1997; de Graaff &
Bröer, 2012; Dieudonné, 2016). The nocebo effect could therefore explain the persistence or
aggravation of the symptoms, but not their initial appearance (Dieudonné, 2016).
EHP request protocols that would fully consider the specificity of their condition (Prignot, 2016).
Research is continuing, and original methodologies are being developed considering the
observations and suggestions of EHP (Huss et al., 2016; Andrianome, 2017; van Moorselaar et al.,
2017) and under real-life conditions (Baliatsas et al., 2015b; Bolte et al., 2015).
Provocation studies have not exhausted their interest yet. They can contribute to the knowledge of
EHS by exploring unusual exposure modalities, especially since studies in real conditions do not yet
allow a comparable dosimetric precision to be achieved, nor to distinguish as precisely the
contributions of the various factors involved. These innovative studies need to integrate the
criticisms of stakeholders when they are scientifically legitimate.
The ExpoComm project intends to develop an original provocation test to radiofrequency
electromagnetic fields that would meet acceptability criteria for individuals with self-reported EHS.
Phases of the ExpoComm project
The ExpoComm project is divided into two phases. In the first phase, two workshops are planned to
think about the conditions required to meet the needs of EHP. For that purpose, limits of the current
provocation tests will be analysed and solutions will be explored through an iterative co-creation
process involving various actors: EHP, researchers and technical experts. In the second phase,
EHP will have the opportunity to test the protocol.
In this paper, we are focusing on the description of the first phase.
Implementation of the first phase
The process used in this first phase is supported by the Wallonia e-health Living Lab (WeLL,
http://well-livinglab.be/), specialized in the co-creation of technological solutions in the field of health,
involving future users. Their experience in the organisation of co-creation workshops will ensure that
brainstorming sessions prior to the development of the protocol are carried out as efficiently as
possible. Focus is given to a precise definition of needs.
This phase is in progress; workshops are planned in April and May 2018. At the time of BioEM2018,
both workshops will have taken place. Results and challenges will be presented.
Objectives of the workshops

193
A first meeting was organized by the WeLL in January 2018 to define the specifications of the
workshops and the action of the WeLL. All ExpoComm project collaborators attended this meeting.
Limitations related to provocation tests and requirements involving good practices in science were
discussed.
A list of compelled criteria and of those that are open to discussion was drawn. Some of the
compulsory criteria include the need to:
explore links between defined EMF sources and symptoms through individualized double
blinded conditions of exposure, taking care of minimizing potential nocebo responses;
clarify the status of these responses by trying to reduce the anxiogenic character of the tests
and by comparing EHP to a control population sharing similar psychometric characteristics.
Attention will be given to individualize exposure sessions (duration, latency, signal…), to limit
nocebo responses, and to handle the actual sensitivity. These criteria will be targeted in the first
workshop and collectively debated.
An on-going evaluation of the co-creation process will be conducted to explore the relevance of
collaborating with EHP in the development of the study protocol.
The second workshop will be conducted as a usability test aiming to evaluate the acceptability of the
protocol.
Recruitment of volunteers
As early as March 2017, before submitting the ExpoComm project to Anses, we contacted a Belgian
association active in EHS recognition to solicit collaboration. However, the length of time required to
establish contacts did not allow including them.
New contacts were made in October 2017 when the ExpoComm project had been approved by
Anses. Currently, this association does not wish to become involved in the study because of
questions about the direction taken by the study and our financial independence. We hope that the
quality of the work that will be provided in the first phase will help ensuring their interest and
establishing a climate of trust. Their participation in the second stage of the project would be
valuable.
Following discussions with this association and several people, a document explaining our
reflections on EHS, the objectives of the study, its planning and the source of funding was drafted.
This document was sent to various contacts established with other associations and with people met
at public seminars. Anyone wishing further information was also invited to contact us to discuss his
or her questions.
Currently eight EHP have agreed to participate to the workshops. Some of their characteristics are
summarized in Table 1, based on informal interviews. A dedicated questionnaire is developed to
assess the level of sensitivity, the confidence of being sensitive to defined sources, the impact of
this sensitivity in daily life, adaptations made to manage exposure, the general sensitivity and
symptomatology.
Table 1 – Characteristics of the volunteers
Duration of EHS Employment
Sex Age group
(> 7 years) status

EHP1 F > 60y yes Inactive

EHP2M M 40-60y no Active

EHP3 F 40-60y no Active

EHP4 M 20-40y / Active

194
 EHP5 M 40-60y yes Active

EHP6 M 40-60y yes Active

EHP7 F 40-60y (in questioning) Active

EHP8 M > 60y yes Partly active

The varied profiles of participants suggest that the range of contributions will be very broad.
Nevertheless, we would welcome some additional participants to ensure the widest possible
reflection on the issue.
Funding
ExpoComm is funded by The French National Research Program for Environmental and
Occupational Health of Anses (EST/2017/2 RF/19).
References
Andrianome, S. (2017). Intolérance environnementale idiopathique avec attribution aux champs
électromagnétiques (IEI-CEM) : Étude des systèmes endocrinien, nerveux sympathique et
immunitaire, Thèse de doctorat en Biologie et Santé, Amiens, Université de Picardie Jules Verne.
Baliatsas, C., Bolte, J., Yzermans, J., Kelfkens, G., Hooiveld, M., Lebret, E., & van Kamp, I. (2015).
Actual and perceived exposure to electromagnetic fields and non-specific physical symptoms: an
epidemiological study based on self-reported data and electronic medical records. Int J Hyg Environ
Health, 218(3):331-44.
Bergqvist, U. & Vogel, E. (1997). Possible health implications of subjective symptoms and
electromagnetic field. A report prepared by a European group of experts for the European
Commission, DGV. Arbete och Hälsa, 1997:19. Swedish National Institute for Working Life,
Stockholm, Sweden. ISBN 91-7045-438-8.
Bolte, J.F.B., Baliatsas, C., Eikelboom, T., & van Kamp I. (2015). Everyday exposure to power
frequency magnetic fields and associations with non-specific physical symptoms. Environmental
Pollution (Barking, Essex: 1987), 196, 224‑229.
de Graaff, M. B., & Bröer, C. (2012). ‘We are the canary in a coal mine’: establishing a disease
category and a new health risk. Health, Risk & Society, 14(2):129-147.
Dieudonné, M. (2016). Does electromagnetic hypersensitivity originate from nocebo responses?
Indications from a qualitative study. Bioelectromagnetics, 37(1):14-24.
Huss, A, Murbach, M, van Moorselaar, I, Kuster, N, van Strien, R, Kromhout, H, Vermeulen, R, &
Slottje P. (2016). Novel exposure units for at-home personalized testing of electromagnetic
sensibility. Bioelectromagnetics, 37(1):62-8.
Leszcynski,, D. (2018, February 11). Brief report on the EHS provocation studies. Retrieved from
https://betweenrockandhardplace.files.wordpress.com/2018/02/leszczynski-updated-brief-report-on-
ehs2.pdf
Oftedal, G., Straume, A., Johnsson, A., & Stovner, L.J. (2007). Mobile phone headache: a double
blind, sham-controlled provocation study. Cephalalgia, 27(5):447-55.
Prignot N. (2016). L’onde, la preuve et le militant. L’écosophie de Félix Guattari a l’épreuve de
l’électrosensibilité et de la polémique sur les dangers des ondes électromagnétiques, Thèse de
doctorat en philosophie, Bruxelles, Université Libre de Bruxelles, 444 p.
Regel, S.J., Negovetic; S., Röösli; M., Berdiñas; V., Schuderer; J., Huss; A., Lott; U., Kuster; N., &
Achermann, P. (2006). UMTS base station like exposure, well being and cognitive performance.
Environmental Health Perspectives, 114 (8):1270-1275.
Rubin G.J., Nieto-Hernandez R. & Wessely S. (2010). Idiopathic environmental intolerance

195
attributed to electromagnetic fields (formerly “electromagnetic hypersensitivity”): An updated
systematic review of provocation studies. Bioelectromagnetics, 31(1):1-11.
Rubin G.J., Hillert L., Nieto-Hernandez R., van Rongen E. & Oftedal G. (2011). Do people with
idiopathic environmental intolerance attributed to electromagnetic fields display physiological effects
when exposed to electromagnetic fields? A systematic review of provocation studies.
Bioelectromagnetics, 32(8):593-609.
van Moorselaar I, Slottje P, Heller P, van Strien R, Kromhout H, Murbach M, Kuster N, Vermeulen R,
& Huss A. (2017). Effects of personalised exposure on self-rated electromagnetic hypersensitivity
and sensibility - A double-blind randomised controlled trial. Environ Int, 99:255-262.

S10-4 [11:45]
STUDENT PAPER
Modeling of electrocardiogram generation based on electric field analysis with numerical
human model
Tatsuhito Nakane1, Takahiro Ito1, Akimasa Hirata1, Nobuaki Matsuura2 & Hiroyoshi Togo2
1Department of Electrical and Mechanical Engineering, Nagoya, Japan, 466-8555
2Device Innovation Center, NTT, Kanagawa, Japan, 243-0198
Keywords: Human, Static, Work in Progress
Presented by: Tatsuhito Nakane
In this study, we proposed an electrocardiogram (ECG) generation model by electric field analysis to
develop a performance evaluation of wearable devices. The propagation of the electrical excitation
in the conduction pathway of the heart can be approximated by sequentially-placed small electric
dipole. We conducted computer simulations that solve potentials based on the scalar-potential finite-
difference (SPFD) method with one electric dipole placed as a source. Post-processing is then
applied to simulate the ECG waveforms for multiple computations. Our computational results
demonstrated the effectiveness of ECG modeling using electric dipoles.
Introduction
In recent years, the development of medical devices based on bioelectromagnetism has been
rapidly advanced owing to improvement in computer performance. In this area, many wearable
devices have been developed for the purpose of health management in daily life according to
combining various technologies such as miniaturization technology of electronic parts, smart textiles
and wireless communication technology [1]. In medical and healthcare applications, biological
signals are measured, for example blood pressure, heart rate, and electrocardiogram (ECG). While
wearable devices have become popular, evaluation for the device characteristics have not been
established yet. On the one hand, in recent years, as mathematical modeling of detailed cardiac
function became possible by improving computer performance, numerical simulations have been
conducted to reproduce the electrical phenomena of heart [2]. However, it is not easy to imitate the
structure of the previous study by electric circuits. Therefore, it is important to reproduce ECG with
simple wave sources. In this study, as a preliminary study, we propose an ECG generation model
with small dipoles utilizing a numerical human model.
Materials and Methods
An anatomical human body model of Japanese adult male developed at National Information and
Communication Technology was used in our analysis [3]. The human body model consists of 51
tissue types, and electric conductivity was assigned to each tissue [4][5]. We assume that the
propagation of the electrical excitation in the conduction pathway of the heart can be approximated
by sequentially-placed small electric dipole. The electric dipole was placed as shown in Fig. 1(a)
according to the conduction system as shown in Fig. 1(b). The propagation time of electrical
excitation was calculated using its propagation velocity [6] and the distance between the dipoles.
Furthermore, we decided six measurement points (V1-V6), corresponding to typical electrode

196
positions, on the chest of the human model. Figure 2 shows the chest lead measurement points in
this study. We then conducted computer simulations for each electric dipole position, and solve
potentials in analysis area based on the scalar-potential finite-difference (SPFD) method [7]. After
that, we calculated the average potential of both wrists and left ankle as the reference potential, and
V1-V6 were calculated as the difference between the potentials under the measurement points and
reference potential.
Results
Figure 3 shows the voltages at the chest lead position obtained by the simulations. As shown in Fig.
3, the P wave, which is the excitement of the atrium, is well simulated at all the lead positions. On
the other hand, focusing on the QRS complex, although the lead position V4-V6 changes relatively
large which is a characteristic of the QRS complex, it cannot be confirmed at other lead position V1-
V3. This result suggests that the locations of dipoles placed on the conduction pathway in the
ventricle is imperfection considering that the QRS complex. The conduction system of the ventricle
consists of the left and right bundle branches, and then the electric excitation propagates to the
ventricular muscle more complicatedly via Purkinje fibers [6]. In our computation, we assumed the
bundle branches as one path and Purkinje fibers as two paths as a preliminary study. We think that
the obtained simulation results had some limitations.
Conclusions
In this study, we proposed an ECG generation model using electric dipoles. From the result, we
confirmed that the voltage change at the chest lead position can be simulated in the time domain for
the first time. Our future studies are to improve the accuracy of waveform reproduction by adjusting
the dipoles location on the conduction pathway, and to compare with clinical measurements.
References
[1] A. Pantelopoulos and N. G. Bourbakis, “A Survey on Wearable Sensor-Based Systems for
Health Monitoring and Prognosis,” IEEE Trans. Syst. Man, Cybern. Part C (Applications Rev)., vol.
40, no. 1, pp. 1–12, Jan. 2010.
[2] E. Schenone, A. Collin, and J.-F. Gerbeau, “Numerical simulation of electrocardiograms for full
cardiac cycles in healthy and pathological conditions,” Int. j. numer. method. biomed. eng., vol. 32,
no. 5, p. e02744, May 2016.
[3] T. Nagaoka et al., “Development of realistic high-resolution whole-body voxel models of
Japanese adult males and females of average height and weight, and application of models to
radio-frequency electromagnetic-field dosimetry,” Phys. Med. Biol., vol. 49, no. 1, pp. 1–15, Dec.
2004.
[4] S. Gabriel, R. W. Lau, and C. Gabriel, “The dielectric properties of biological tissues: III.
Parametric models for the dielectric spectrum of tissues,” Phys. Med. Biol., vol. 41, no. 11, pp.
2271–2293, Nov. 1996.
[5] V. De Santis, X. L. Chen, I. Laakso, and A. Hirata, “An equivalent skin conductivity model for low-
frequency magnetic field dosimetry,” Biomed. Phys. Eng. Express, vol. 1, no. 1, p. 15201, Jun. 2015
[6] J. Malmivuo and R. Plonsey, Bioelectromagnetism - principles and applications of bioelectric
andbiomagnetic fields. New York: Oxford University Press, 1995.
[7] T. W. Dawson and M. A. Stuchly, “Analytic Validation of a Three-Dimensional Scalar-Potential
Finite-Difference Code for Low-Frequency Magnetic Induction,” J. Appl. Comput. Electromagn. Soc.,
vol. 11, pp. 72–81, 1996.
Figures

197
Figure 1. (a) The assigned positions of the electric dipole in the heart of TARO model,and (b)
The conduction system of the heart (edited from the figure in [6]).

Figure 2. The chest lead measurement points.

198
 Figure 3. The voltages at the chest lead position obtained by calculation.

S10-5 [12:00]
Probability of Injury from Radio Frequency Exposure (PIRE): A program to summarize and
expand our knowledge of potential damage produced by RF exposure
Jeffrey Whitmore1 & Jason Payne1
1Radio Frequency Bioeffects Branch, United States Air Force, Fort Sam Houston, Texas, USA, 78232
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Jeffrey Whitmore
The potential for humans to receive a high power RF exposure is increasing as RF source
technology continues to advance. Currently there is no comprehensive model relating the
parameters of a RF exposure (e.g., power density, exposure duration, and frequency) to the nature
and severity of an injury. This work intends to establish a model driven, empirically- validated
description of thermal damage produced in tissue by RF exposure.
Background
Clinical exposure to high power radiofrequency (RF) energy has been studied for its ability to both
warm and destroy tissue though heat deposition (e.g., He et al., 2004). Additionally, the potential for
humans to receive a high power RF exposure outside medical environments is increasing as RF
source technology continues to advance. Currently there is no comprehensive model relating the
parameters of a RF exposure (e.g., power density, exposure duration, and frequency) to the nature
and severity of an injury. Metrics that combine time and temperature information and relate those to

199
the severity of injury have been developed (e.g., CEM43°C introduced by Sapareto & Dewey, 1984)
and applied by researchers to determine the sensitivity of different tissues to thermal insult (see the
reviews of Dewhirst & Sim, 1984; Yarmolenko et al., 2011).
Proposed Solution
Establish a model driven, empirically- validated description of thermal damage produced in tissue by
RF exposure. A comprehensive solution is a very large effort and early advances will relate modeled
thermal insults to general metrics of thermal damage. As data is gathered and modeling information
is produced, refinements will be made enabling a range of damage predictions for specific tissues.
Additionally, given the complexity of RF exposures, including the dynamics of a moving target and
source, probabilistic estimates will emerge to provide some relative measure of confidence for a
given exposure to produce injury.
Method
Perform high-fidelity numerical solutions in realistic anatomy models, use physics solvers to produce
detailed thermal data, and relate those insults to metrics of thermal damage and direct observations
of thermal-induced empirical changes.
Results/Status
A range of RF exposure parameters have been modeled to produce thermal profiles that were
summarized into CEM43°C values. These values were then related to the existing IEEE and ICNIRP
safety standards. See Figure 1 for an example.
Discussion
Work has begun under the PIRE program to produce a comprehensive model of the injury likelihood
resulting from RF exposures. Initial modeling efforts are underway relating generalized RF
exposures to metrics of tissue damage. A research program for collecting supporting empirical
information has also been outlined and the first studies initiated. Performing injury analyses of RF
exposure conditions provides useful information beyond what the current safety standards can
provide. Predicting tissue damage for an RF exposure of interest allows for an associated risk
evaluation, enabling workers and other decision makers to determine whether an exposure is
acceptable.
References
Dewhirst, M. W. & Sim, D. A. (1984). The utility of thermal dose as a predictor of tumor and normal
tissue responses to combined radiation and hyperthermia. Cancer Res, 44(10 Suppl), 4772s–4780s.
He, X., McGee, S., Coad, J. E., Schmidlin, F., Iaizzo, P. A., Swanlund, D. J., Kluge, S., et al. (2004).
Investigation of the thermal and tissue injury behaviour in microwave thermal therapy using a
porcine kidney model. Int J Hyperthermia, 20(6), 567–93.
Sapareto, S. A. & Dewey, W. C. (1984). Thermal dose determination in cancer therapy. Int J Radiat
Oncol Biol Phys, 10(6), 787–800.
Yarmolenko, P. S., Moon, E. J., Landon, C., Manzoor, A., Hochman, D. W., Viglianti, B. L. &
Dewhirst, M. W. (2011). Thresholds for thermal damage to normal tissues: an update. Int J
Hyperthermia, 27(4), 320–43.
Figures

200
 Figure 1. Calculated exposure values in relation to the IEEE C95.1 safety limit.

S10-6 [12:15]
Optimization of an on-skin lens-into-the-body for implantable medical devices
Christopher Trampel1
1Department of Engineering, Weber State University, Ogden, UT, USA, 84408-1803
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Christopher Trampel
This abstract describes the optimization of a novel wearable metasurface sensor, the lens-into-the-
body, for implantable medical devices (IMDs). The lens-into-the-body is an on-skin device designed
to focus the electromagnetic fields from an RFID reader on an implant. The function of the lens is to
increase the read-range of an RFID communication system for medical implants. Full-wave
simulations of the optimized lens are discussed.
I. Introduction
Implantable devices incorporated into most Wireless Body Area Networks (WBANs) communicate
via a small radiofrequency (RF) telemetry system. The power for the transmission is provided by a
battery, the so-called active communication approach. However, transmitting through lossy body
tissue rapidly drains the battery and limits the range of the communication link. [1] In addition, the
small electrical size of current implantable antennas makes them inefficient radiators. The
performance of implantable transmitters is further limited by regulations governing specific
absorption rates (SAR) in the body. The passive backscatter communication approach employed by
Radio Frequency Identification (RFID) systems is an attractive alternative that would increase the
lifetime of medical implants such as subcutaneous glucose sensors, neural interfaces, and retinal

201
and cochlear prosthetics. This abstract describes a novel wearable metasurface sensor, the lens-
into-the-body, optimized for a medical RFID system. Such a passive RFID system avoids the
problems associated with active transmission through the body. The passive approach promises to
reduce the energy budget for the IMD thereby increasing its lifetime.
This abstract describes progress to date and is organized as follows. In Section II, results of the
optimization of the on-skin lens are presented and discussed. Section III includes conclusions and
future work.
II. Results and Discussion
The lens was designed via full-wave simulations from CST Studio Suite. The model of the human
body consisted of three layers: skin (0.062 mm thick), fat (8 mm thick), and muscle (30 mm thick).
Absorbing boundary conditions were used to truncate the domain of the frequency-domain finite-
element simulations. The RFID reader antenna is a loop 122.4 mm in circumference and is
positioned 27 mm above the skin. The reader antenna is fed by a discrete port labeled port 1. The
RFID tag antenna is a loop 31.4 mm in circumference and is located just inside the muscle layer. It
is similarly connect to a discrete port, port 2.
The on-skin lens-into-the-body is shown in Figure 1. The lens consists of two layers of perfectly
conducting metal rings separated by a 0.5 mm thick polyimide dielectric layer. The rings are 0.5 mm
wide and 0.01 mm thick. A multivariable optimization was performed over the radii of the rings using
CST's Covariance Matrix Adaptation (CMA) evolution strategy. The upper and lower surfaces each
have eight rings. Thus, the CMA algorithm was given control over the sixteen variables consisting of
the radii of the rings. The power transmitted to the RFID tag may be studied via the S-parameters of
the two-port network consisting of ports 1 and 2. The goal of the optimization was to maximize |S21|
at 2.45 GHz; one may expect the read-range to be increased by maximizing |S21|. The radii of the
loops shown in Figure 1 resulted in the largest |S21| over the duration of the optimization.
The lens-into-the-body is a Huygens' metasurface [2] that improves the power delivered to the tag.
By varying the density of the wires in the radial direction one is able to modulate the conductivity on
the upper and lower surfaces. The wires are spaced much less than one wavelength apart in the
longitudinal direction, creating a discontinuity in the transverse components of the magnetic field.
The metasurface adjusts the surface impedance such that reflections are reduced and the fields are
concentrated on the implant. The effect of the lens may be observed by comparing the tangential
magnetic field without the lens (Figure 2a) to the field in the presence of the lens (Figure 2b). The
magnetic field is widely distributed in the radial direction throughout the fat and muscle when no lens
is present. The lens focuses the magnetic field on the RFID tag. Figure 3 shows that the optimized
on-skin lens increases |S21| by approximately 4 dB in the vicinity of 2.45 GHz. It should be noted
that the Industrial, Scientific, and Medical (ISM) band from 2.4 - 2.5 GHz is available for WBAN
communication.
III. Conclusions and Future Work
Simulations indicate that the lens may improve read-range over a band centered on 2.45 GHz so
long as the reader and tag coils are coaxial. However, the reader and tag antennas will likely not be
aligned during normal operation. Therefore, future work must address the need for a lens that
improves read-range for noncoaxial as well as coaxial geometries. When radial symmetry is broken
a more complicated metasurface may be necessary. A read-range of one meter for a passive,
implantable RFID tag would be sufficient for a reader incorporated into a mobile device. In order to
difinitively determine the read-range for likely tag-reader configurations the lens needs to be tested
with a commercial RFID system.
IV. References
[1] C.M. Furse and A. Chrysler, "A history & future of implantable antennas," 2014 IEEE Antennas
and Propagation Society International Symposium (APSURSI), Memphis, TN, 2014, pp. 527-528.
[2] C. Pfeiffer and A. Grbic, "Metamaterial Huygens' Surfaces: Tailoring Wave Fronts with

202
Reflectionless Sheets," Phys. Rev. Lett., vol. 110, no. 19,197401, 2013.
Figures

Figure 1. On-skin lens-into-the-body.

Figure 2. Tangential magnetic field at 2.45 GHz without (a) and with (b) the lens-into-the-body.

203
 Figure 3. |S21| with and without on-skin lens.

Session: S11
Dosimetry II - Measurements
Thursday June 28, 2018 • 11:00 - 12:30
Europa D
Chairs: Wout Joseph & Sam Aerts

S11-1 [11:00]
Personal exposure to radio-frequency electromagnetic fields in Europe: is there a generation
gap?
Marloes Eeftens1, 2, Benjamin Struchen1, 2, Laura Birks3, 4, 5, Elisabeth Cardis3, 4, 5, Peter Gajšek6, Anke
Huss7, Leeka Kheifets8, Inger Kristine Meder9, Jorn Olsen10, Tomaž Trček6, Blaž Valič6, Roel Vermeulen7,
11, Martine Vrijheid3, 4, 5, Luuk van Wel7, Mònica Guxens3, 4, 5, 12 & Martin Röösli1, 2
1Swiss Tropical & Public Health Institute, Basel, Switzerland
2University of Basel, Basel, Switzerland
3IS Global, Barcelona, Spain
4Pompeu Fabra University, Barcelona, Spain
5Spanish Consortium for Research on Epidemiology and Public Health (CIBERESP), Institute of Health
Carlos III, Madrid, Spain
6Institute of Non-ionizing Radiation (INIS), Ljubljana, Slovenia
7Institute for Risk Assessment Sciences (IRAS), Utrecht, the Netherlands
8Danish of Epidemiology, University of California, Los Angeles, USA
9Danish National Birth Cohort, Statens Serum Institute, Copenhagen, Denmark

204
10Department of Public Health, Aarhus University, Aarhus, Denmark
11JuliusCenter for Health Sciences and Primary Care, University Medical Center Utrecht, Utrecht, the
Netherlands
12Department of Child and Adolescent Psychiatry/Psychology, Erasmus University Medical Centre–Sophia
Children's Hospital, Rotterdam, the Netherlands
Keywords: Behavioural, Static, Completed (published)
Presented by: Marloes Eeftens
We conducted a personal measurement study of radiofrequency electromagnetic field exposure
among 294 parent child pairs from 5 different European countries, and found that parents had a
slightly higher exposure than their children, especially for uplink and DECT. Exposure was
correlated between parents and their children, especially for exposure resulting from fixed site
transmitters (e.g. broadcast and downlink). Correlations between family members are also higher
during the night and for the time spent at home.
Background
Exposure to radiofrequency electromagnetic fields (RF-EMF) from mobile communication
technologies is changing rapidly. To characterize sources and associated variability, we studied the
differences and correlations in exposure patterns between children aged 8 to 18 and their parents,
over the course of the day, by age, by activity, and for different metrics of exposure.
Methods
Using portable RF-EMF measurement devices, we collected real-time personal measurements of
RF-EMF over 24 to 72 hours in 294 parent-child pairs from Denmark, the Netherlands, Slovenia,
Switzerland, and Spain. The meters measured the power flux density (mW/m2) of 16 different
frequency bands every 4 seconds, and activity diaries kept by the participants were used to collect
activity information in real-time. We analyzed their exposures by activity, for the different sources of
exposure: downlink (from mobile phone base stations), uplink (from mobile phones), broadcast
(from radio or TV antennas), DECT (from cordless phones) and Wi-Fi. We looked at the correlations
between parents and children overall, during day (06:00-22.00) and night (22:00-06:00) and while
spending time at home.
Results
Mean (median) exposure was 0.16 (0.068) mW/m² for children and 0.15 (0.085) mW/m² for parents,
predominantly originating from downlink sources (47% for children and 45% for parents), followed
by uplink (18% and 27% respectively) and broadcast (25% and 19%) (Figure 1). On average,
exposure for downlink and uplink were highest during the day, and for Wi-Fi and DECT during the
evening. Exposure during activities where most of the time is spent (home, school and work) is
relatively low, whereas exposure during travel and outside activities is higher. Exposure to uplink
increased with age among the children in the study population, while DECT decreased slightly.
Exposure to downlink, broadcast, and Wi-Fi show no obvious trend with age. We found that
exposure to total RF-EMF is correlated among children and their parents (Rspearman= = 0.45),
especially while at home (0.62) and during the night (0.60) (see Figure 2). Correlations were higher
for exposure from fixed site transmitters such as downlink (0.57) and broadcast (0.62) than for
mobile device related exposures such as uplink (0.29).
Discussion
The generation gap between children and their parents is mostly evident in uplink exposure, due to
more and longer mobile phone and cordless phone calls among parents, and their tendency to
spend slightly more time in activities with higher environmental RF-EMF exposure, such as travel.
Despite these differences in personal behavior, exposure to RF-EMF is correlated between children
and their parents, especially exposures resulting from fixed site transmitters (such as downlink,
broadcast).

205
Figures

Figure 1. Percentile distribution (boxplot) and mean (diamond) exposure to downlink, uplink,
broadcast, DECT, Wi-Fi and total RF-EMF per person for children and parents. The
percentage indicates the average contribution of each specific band to the total exposure. The
gray violins portray the overall distribution of all participants.

206
Figure 2. Pearson (r[p]) and Spearman (r[s]) correlations for sources broadcast, downlink,
uplink, DECT, Wi-Fi, and total RF-EMF between parents and children for each country and for
the daytime, nighttime, time spent at home and overall.

207
S11-2 [11:15]
STUDENT PAPER
Concepts for innovative measuring methods to monitor EMF exposure evolution
Sascha Schiessl1, Thomas Kopacz1, Christian Bornkessel2, Matthias Hein2 & Dirk Heberling1
1Institute of High Frequency Technology, RWTH Aachen University, Aachen, Germany
2RF & Microwave Research Laboratory, Thuringian Center of Innovation in Mobility, TU Ilmenau, Ilmenau,
Germany
Keywords: Dosimetry (measurements), RF/Microwaves, Concept
Presented by: Sascha Schiessl
Monitoring the evolution of exposure to EMF caused by mobile communication systems is a
continuous requirement for a responsible operation of these networks. However, measurements
performed by qualified personnel with professional equipment are expensive and time-consuming
and can only be carried out for a limited number of discrete points in space and time. In this paper,
we suggest two innovative concepts for exposure monitoring. The first is based on a software-
defined dosimeter whose parameters can be easily adapted to the needs of future network
standards. The second employs a crowd sourcing approach using usual mobile phones and their
on-board capability of monitoring relevant network parameters for generating immission-related
data.
Introduction & Motivation
In the context of responsible technological impact management, a continuous monitoring of the
evolution of the exposure to electromagnetic fields is mandatory. The classic approach to fulfill this
task is to perform measurements by qualified personnel using professional measurement
equipment. This method is expensive and time-consuming and at the same time possible only for a
limited number of discrete points in space and time. Though stationary and automated long-term
measurement systems may record immission data continuously, these are quite inflexible and can
only be applied at few locations where they are protected against vandalism. This leads to
measurement results for positions where people in general do not reside rendering them less
relevant.
To overcome these problems, we suggest two innovative concepts for exposure monitoring. The first
one involves a software-defined dosimeter whose parameters can be easily adapted to the needs of
future cell-based communication standards. The second concept makes use of crowd sourcing,
employing standard mobile phones and their on-board capability of monitoring relevant network
parameters for generating immission-related data.
In the following sections, after defining some general requirements for a monitoring concept, the
concepts and their advantages compared to classic methods are described in greater detail.
Requirements for a monitoring concept
Currently, three fundamentally different methods exist for the measurement of the exposure to radio
frequency electromagnetic fields: Broadband, frequency selective, and code-selective. While the
comparatively insensitive broadband measurements are particularly suitable for an overview and
occupational safety measurements, spectral measurements display the highest accuracy. Code-
selective measurements are mandatory, if – in case of UMTS or LTE – immission related to
individual base stations shall be measured, or an extrapolation to full site-utilization shall be
performed, as it is required for regulatory purposes in Germany. For the purpose of risk
communication, on the other hand, the current or temporally and locally average immissions have to
be assessed also. This can be achieved by sampling a spatial volume and subsequent averaging,
but also by “walking” or “driving” along defined measurement tracks. Another approach for risk
communication is based on long-term measuring systems [1], where the immission of individual
radio services is measured stationary over an extended period of time. In addition to stationary long-

208
term measuring systems, portable dosimeters enable the recording of non-stationary and time-
variant immission profiles.
The technological developments in the field of mobile communications, driven in particular by the
upcoming 5G standard, impact the exposure measurement technology. The extension of radio
spectrum towards higher frequency ranges as well as the increasing bandwidths exert serious
demands on measuring devices. Practice in recent years has shown that the development of
immission measurement techniques lags behind the development of new mobile radio technologies.
In addition, the exposure measurement technology represents just a niche product for the
manufacturers of measurement instrumentation.
Considering developments expected by 2025, an adequate monitoring concept for informing the
population would ideally offer the following features:
All radio services are measured continuously and simultaneously in order to capture even short-
term fluctuations of immission correctly.
The monitoring adapts to an ever-increasing heterogeneity of radio services.
The monitoring includes a flexible or even automated adaptation of the measuring instruments
with regard to new frequency ranges and measuring parameters. This should also ensure a
correct assessment of radio services in frequency bands newly “released” (e.g., LTE), that were
occupied by other radio services until recently (e.g., DVB-T).
Two potential powerful approaches to realize a monitoring of electromagnetic fields are software-
defined dosimeters and crowd sourcing-based techniques.
Concepts for EMF monitoring
Software-Defined Dosimeter
The concept of software-defined dosimeters (SDD) is based on the simplicity of measurements with
conventional dosimeters combined with the flexibility offered by spectrum analyzers in terms of
setting measurement parameters. The underlying idea is that all measurement parameters, which
can be adjusted by software, can be modified by the user for an adaptation to the respective
landscape of radio services. This overcomes the problem of present dosimeters with fixed radio
service frequency allocation which become useless upon changes of the frequency allocation.
The software-based adaptation includes the following parameters:
Frequency bands:
Upper and lower limits of frequency bands of radio services should be adjustable within the
entire measuring range. This allows to account for future developments such as the newly
added frequency bands for mobile communications.
Detector type:
Depending on the radio service operating in the respective frequency band, a peak or RMS
detector can be set.
Resolution bandwidth (RBW):
By adjusting the RBW and enabling spectrally sequential-band power measurements with
subsequent summation, it should be possible to measure also newly added radio services and
signals with very wide bandwidth.
Measuring range:
Conventional dosimeters do not have the functionality of adjusting the input level range with
regard to the measured radio service. With the SDD, it should be possible to use software-
controlled attenuators for the input signal. This is particularly advantageous for quasi-
simultaneous uplink and downlink measurements where different maximum levels may occur.
With such a flexible and sustainable solution, the same hardware could be used on the long term,
despite the rapid developments in the field of mobile radio networks. Due to its portability, the SDD
is especially suited for use in a location-variable approach in combination with localization data. The

209
measurement instrumentation may be worn by people and/or mounted on vehicles of private
individuals or public transport. Due to such a spatial variability, a much lower critical mass of the
necessary number of measuring devices is required to reach statistical significance, compared to
the approach by stationary measuring devices.
As part of our further work, the framework and operational conditions for an SDD in terms of
justifiable hardware and software expenditure will be investigated. Further studies will address the
construction of antennas that are capable of coping with the shadowing effects from the user’s
bodies.
Crowd Sourcing-based Smartphone Application
Crowd sourcing is young but already established method in social sciences. The term means
outsourcing the work of a few but specially trained professionals to a large collective of non-experts
(the "crowd"). The special feature is that as a result of crowd sourcing, despite the use of the "non-
specialists", a data quality can be generated equal to that of the professionals. One recent
prominent example of crowd sourcing is the generation of navigation data from "OpenStreetMap"
(www.osm.org).
An implementation of crowd sourcing in the field of immission monitoring could offer the following
features: Instead of equipping a limited number of specially trained measurement teams with
dosimeters or special instruments to take measurements (or attaching dosimeters to static or
moving objects), this measurement task is transferred to a large number of smartphone users using
standard smartphones as a means of gaining immission-related data. Every mobile phone monitors
different network parameters when connecting and staying connected to a base station, to observe
the link quality and convey relevant information for link control and handover decision to the base
station. Some of these parameters are closely related to the exposure caused by the corresponding
networks.
The crowd sourcing idea requires a smartphone application that continuously records the relevant
network parameters in combination with the exact position and a time stamp. From time to time, the
collected data will be submitted to a server for further analysis. Present and future network
performance testing could beneficially be combined with this novel type of immission monitoring.
Key advantages with regard to other monitoring solutions include:
Cheaper than classical measurements, because there is no need for expensive measurement
equipment nor expert measurement teams.
Always up to date, as the smartphones usually support the latest releases of mobile radio
standards with very short delay, while immission measurement techniques often lag behind.
No problem with low sensitivity, especially in contrast to dosimeters, since smartphones
inherently have high sensitivity, even higher than that of conventional spectrum analyzers.
Problems related to the limited access to restricted areas like private apartments or factory
premises can be avoided.
Quasi-continuous measurement become feasible both in time and space, with no problems
about imprecise measurement values due to stationary measurement systems.
Immission is measured where relevant, because mobile phones are in general closely located to
their users. Stationary measurement systems need to be positioned at non-accessible locations.
Despite the numerous advantages, there are technical, social, and legal challenges, e.g., related to
battery life, data ownership, and user acceptance that have to be considered before implementation.
As a first plausibility test, a comparison of measurement values recorded by a selective radiation
meter (Narda SRM-3006) and a smartphone (Galaxy S5 with the application software „QualiPoc“ by
Rohde & Schwarz [2]) was performed in the environment of three different small-cell base stations in
Germany operating in an LTE network at a center frequency of 2650 MHz.
The physical quantity measured by the radiation meter is the electric field strength of the reference
signal (RS) transmitted by the base station. A smartphone determines the reference signal strength
of the serving as well as neighbored base stations periodically. However, due to the unknown
210
antenna factor, the electric field strength cannot be derived from the measured voltage at the
antenna port. Instead, the reference signal receive power (RSRP) can be evaluated.
Figure 1 shows the results of the tests in the surrounding of three different base stations at 15
measuring points, covering a large range of immission values. Ideally, all data points should be
located on a single straight line, proving that the results of both measurement approaches match
perfectly. The measurements show that indeed a considerable correlation between both methods
exists.
Conclusions
In this paper, we suggested and sketched two novel concepts for long-term and comprehensive
exposure monitoring, a software-defined dosimeter and a crowd sourcing-based approach using
smartphones. Both ideas offer a large potential and many advantages compared to conventional
measurement techniques. For the smartphone-based approach, it could be shown by initial studies
that the recorded network parameter RSRP and the electric field strength of the reference signal
measured by a selective radiation meter are strongly correlated. Nevertheless, both conceptual
approaches require further work to prove the feasibility and take the first steps towards
implementation.
Acknowledgment
The funding of the study by the Deutsche Telekom Technik GmbH is greatly acknowledged.
References
[1] pedion24, A project for continuous measurements of the electromagnetic radiation,
www.pedion24.gr, last accessed 2018‑03‑02.
[2] QualiPoc Android, Smartphone-based product for optimizing mobile networks, product brochure

Figures

211
 Figure 1. Comparison between RSRP recorded by a smartphone (x-axis values) and the field
strength of the reference signal measured by a selective radiation meter (y-axis values).

S11-3 [11:30]
Exposure to electromagnetic fields from smart meter technologies in Great Britain (Phase 3):
On-site measurements in homes
Nishtha Chopra1, Darren Addison1, Carolina Calderon1, Myron Maslanyj1 & Azadeh Peyman1
1Centre for Radiation, Chemical and Environmental Hazards , Public Health England, Didcot, United
Kingdom, OX11 0RQ
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)

212
Presented by: Nishtha Chopra
Real-life electromagnetic field measurements were performed for smart meter devices installed in
residential properties across Great Britain, to determine power densities and duty factors associated
with the Home Area Network signals. Of all devices measured, the maximum exposure was less
than 0.003% of the ICNIRP public reference level and the duty factors were less than 1.2%. The
results suggest that smart meters are likely to contribute less to overall exposure than WLAN access
points in the homes. Exposures were also evaluated around a bank of smart meters in a test facility.
The results showed that the total combined exposure was less than 0.0006% of the ICNIRP public
reference level and the combined duty factor was less than 5.2%.
Introduction
A smart meter uses radiofrequency (RF) signals to communicate home readings of gas and
electricity consumption. The data monitored includes location, energy consumption units, time and
frequency of usage. At regular intervals usage data are transmitted from the device to the utility
company, via a Wide Area Network (WAN). The consumer can monitor the usage data via a Home
Area Network (HAN), which is the subject of this study. The HAN is based on ZigBee technology,
operating at 2.4 GHz [1]. Concentrating only on HAN devices, a smart meter network usually
consists of an In Home Display (IHD), an Electricity Meter (EM), a Communication Hub (CH), which
coordinates the HAN communications, and sometimes a Gas Meter (GM). In some cases, the EM
and CH are combined.
Previously published measurements focused on exposures in an ideal controlled laboratory
environment [2]. However, with the full-scale rollout of smart meters across the Great Britain (GB)
underway, it was also considered important to assess exposures incurred in real-life situations in
people’s homes. ICNIRP (International Commission on Non-Ionizing Radiation Protection)
guidelines give basic restrictions defined in terms of Specific Energy Absorption Rate (SAR), which
should be averaged over a period of 6 minutes in order to prevent adverse effects due to excessive
whole and partial body heating. The guidelines also give reference levels, defined in terms of power
density and used for compliance purposes to ensure that the basic restriction is not exceeded. In
addition, for pulsed RF fields above 10 MHz, the peak power density (averaged over the pulse)
should not exceed 10,000 W/m2 [3].
This paper presents empirically determined exposure data based on measurements performed in a
sample of residential properties where smart meter systems were installed, and summarises the
results in terms of percentage of the ICNIRP general public reference levels.
Material and methods
Public Health England (PHE) recruited 20 homes in this study, with the support of volunteers from
Smart Energy GB, the Department for Business, Energy & Industrial Strategy (BEIS) and utility
representatives in GB. The preparatory work included the development of a consistent
measurement protocol, designed to capture the Zigbee signals and characterise exposure for the
smart meter devices and the other various sources of RF electromagnetic field (EMF) devices likely
to be encountered in the homes. Power density measurements (W/m2) were made using a Narda
Selective Radiation Meter (SRM) 3006 [4] connected to a sensor probe with three-axis isotropic
antennas (Figure 1). Smart Meter devices occupy a single fixed frequency Zigbee channel. The
SRM has a calibration traceable to the manufacturer standards with uncertainty quoted as -2.4
dB/+1.9 dB.

213
 F 1. N S R M -3006 3
Power density measurements were made at a distance of 0.5m and 1m from each device using the
level recorder mode of the SRM. The SRM was set to record both the average root mean square
(RMS) power density and the maximum peak power density over the sampling period of 6 minutes,
using two different types of detector modes. For duty factor calculations, Ember Insight Desktop
software (Silicon Labs, Texas, USA) was used with a passive Telegisis ETRX357 sensor module.
The data was exported and analysed with a MATLAB program (MathWorks, Massachusetts, USA).
For the assessment of the duty factors it was important to take into account that the GMs
transmitted every 30 mins, while the other smart meter devices transmitted roughly every 15
seconds. The number and duration of transmissions are important parameters required to quantify
the smart meter device’s active operation.
An additional SRM was also used in safety evaluation mode to assess the exposure across 32
defined channels from other RF sources in the homes, for comparison purposes. Measurements
were performed in the centre of three of the most occupied rooms at each property, as reported by
home occupiers. The centre of the probe was set at 1m from the floor and at least 1 m from any
other RF source. A note was made on any visible RF sources, typically WLAN access points, and
home occupiers’ mobile phones/laptops or nearby mobile phone masts, as well as smart meter
devices.
To assess the exposure from combinations of smart meter devices, a series of measurements were
made around a bank of combined EMs and CHs, installed in a utility company’s test facility (Figure.
2). The bank of meters consisted of 16 SMETS1 (Smart Metering Equipment Technical
Specifications) systems arranged across three rows. Here, the EM and CH were physically
combined, and as the CH is the coordinator of the network, only transmissions from the CH were
present. The SRM was used in safety evaluation mode to measure power density across all 16
Zigbee channel frequencies, with the probe operating in isotropic mode. Measurements were carried
out in front and at the back of the bank of meters at distances of 0.5 and 1 m.
For duty factor measurements, the utility test house provided channel numbers and PAN ID
(Personal Area Network Identifier) information for the bank of meters. The Ember sniffing software

214
was used to identify the PAN ID and, in case of more than one PAN ID per channel (owing to
channel reuse), the duty factors were combined.

F 2. E
( ) (0.5 1 ) SRM

Results and discussion

The number of devices under test, transmission time and exposure level were analysed using a
MATLAB statistical toolbox. The maximum peak power density values at a distance of 0.5 m for 19
IHDs, 20 CHs (out of which 6 EMs were connected together) and 16 GMs, show an aproximate log-
normal distribution (Figure 3). The calculated geometric mean value of power density for all smart
meter devices was found to be 2.02 mW/m2. The distribution for the average power density was
more irregular due to differences in duty factor across type of smart meter devices. Differences in
power densities for IHDs, CHs and GMs were not statistically significant (p > 0.39) at both measured
distances.
The maximum recorded RMS power density value out of all the smart meter devices at a distance of
0.5 m was 0.262 mW/m2, which constitute as less than 0.003% of the ICNIRP reference levels for
general public.
The duty factor values were found to vary depending on the devices’ activity at any given time. As
expected, GMs transmitted significantly less (p-value < 10-4), than IHDs and CHs, due to the
reduced transmission sequences of these devices. The maximum recorded duty factor was 1.2%,
arising from an IHD device. The highest duty factor observed across all CHs (0.9%) was measured
in the same home where the largest IHD duty factor was found.

215
F 3. P 0.5

The results of measurements for other RF sources in each home have been collated into the
following groups: mobile phones, cordless phones (DECT), WLAN (2 GHz and 5 GHz), and others.
The exposure quotients for these groups, as well as total exposure across all bands, are also
displayed in a boxplot form in Figure 4.
The WLAN 2 GHz band would contain contributions from both WLAN and Zigbee devices present in
the environment. However, in only two of the 61 rooms measured, was a smart meter device (IHD)
in the same room as the WLAN access point. Additional measurements showed that the WLAN
signal was at least four times higher than the Zigbee, in the centre of the room.

F 4. E ( - ) ICNIRP
RF
For the bank of meters, the maximum measured power density for an individual channel (ch.11)
measured in front of the bank at a distance of 0.5 m was 3.59 mW/m2.
The duty factors for all active channels were combined and found to be less than 5.2%. Applying

216
this duty factor to the combined total RMS power density exposure for all active channels would give
a value of 59.8 µW/m2 at a distance of 0.5 m at the front of the bank of meters.
Conclusion
RF exposure levels were assessed around a selection of smart meter (HAN) devices (20 CHs, 20
IHDs, 16 GMs and 6 EMs) installed in a sample of residential properties in GB. Results show that
maximum recorded RMS power density out of all the smart meter devices at a distance of 0.5 m
was 0.262 mW/m2. This exposure level corresponds to less than 0.003% of the ICNIRP reference
level for general public, even for the combined exposure from EM and CHs. Duty factors for various
smart meter devices were all found to be less than 1.2%.
In the case of the bank of meters, the combined total RMS power density exposure with duty factors
applied was 59.8 µW/m2, which is less than 0.0006% of the ICNIRP reference level for the general
public. Duty factors for all active channels, when combined were less than 5.2%.
RF exposure in the centre of the most commonly used rooms suggests that smart meters are likely
to contribute less to overall exposure than WLAN access points in homes.
Acknowledgements
This work was supported financially by Smart Energy GB.
References
[1] ZigBee Specification 053474r17 . www.ZigBee.org
[2] Peyman, A., Addison, D., Mee, T., Goiceanu, C., Maslanyj, M., & Mann, S. (2017). Exposure to
electromagnetic fields from smart utility meters in GB; part I) laboratory measurements.
Bioelectromagnetics, 38(4), 280-294. doi:10.1002/bem.22044
[3] ICNIRP (1998). Guidelines for limiting exposure to time-varying electric, magnetic, and
electromagnetic fields (up to 300 GHz). International Commission on Non-Ionizing Radiation
Protection. Health Phys, 74(4), 494-522.
[4] Narda (2010). SRM3006: Selective Radiation Meter (Operating Manual)(03/04.2010). Retrieved
from http://www.narda-sts.us/pdf_files/OperatingManuals/SRM3006_Manual.pdf

S11-4 [11:45]
Statistical RF exposure analysis for multiple massive MIMO antennas
Maarouf Al Hajj1, Yuanyuan Huang1 & Joe Wiart1
1Chaire C2M, Télécom ParisTech, Paris, France, 75013
Keywords: Dosimetry (computational), RF/Microwaves, Completed (unpublished)
Presented by: Joe Wiart
This paper presents a statistical analysis of the human exposure linked to multiple M-MIMO
antennas emissions. The aim of this study is to better estimate the compliance boundary around 5G
base stations utilizing M-MIMO antennas without complex estimation methods for different
transmission and traffic characteristics, and for different power ratios between the antennas. The
statistical estimation is obtained by representing the exposure as a beta distribution with parameters
depending on each configuration. The exposure induced by different scenarios of multiple
transmitters is statistically performed. Based on that, the ratios of the quantile of total exposure to
the theoretical maximum are studied and presented.
The combined emissions produce a CDF bounded by the other two. And since the two emissions
are assumed to have the same maximum transmit power, the transmitted power at higher
percentiles is almost always bigger than the powers at the same percentiles for both the two
emissions. This might not be the case for different maximum power ratios between the transmitters.

217
I. Introduction
Radio frequency compliance assessments are conducted to ensure that the radio emitters comply
with the regulations on human exposure. The purpose of the assessments is to define a volume
boundary outside of which, the RF exposure is below the limits. Exposure assessments have been
developed and generalized for previous technologies. These assessments are conducted assuming
the theoretical maximum power transmitted in all directions. While this approach proved useful for
previous mobile technologies, it is problematic for assessments conducted for massive MIMO
antennas used in 5G systems.
Massive MIMO will allow, for a short duration, the concentration of the antenna’s power in a single
high gain beam, towards a specific user. The classical assessment methods based on the envelop
of the maximums when used for massive MIMO antennas will produce a nonrealistic overly
conservative compliance boundary. In order to assess more realistically these later statistical
approaches are considered for designing the compliance boundaries for massive MIMO antennas.
II. Emission estimation method
Multiple statistical approaches have been developed to evaluate the actual transmitted power in
multiple scenarios e.g. in [1] and [2]. Since the approach in [2] makes simplified assumptions
regarding beamforming design and traffic conditions, and it is based on modeling the expectation of
the total power emitted, in this analysis, we build on the results obtained in [1].
The principle of this approach is to derive the cumulative distribution function of the normalized
power transmitted by the BSs. Two scenarios were the main focus in this approach: Urban macro
and urban micro with differences in deployment heights and power emitted. To study scenarios
where the RF exposure can be a real issue, a full buffer model is assumed by simulating a
sequence of UE drops with ‘D’ duration, and K active users per cell.
The results presented in fig. 1 show that even when we have even just K=1 UE active per BS and
D=60s, the actual transmit power is 32% in UMa and 27% in UMi of the transmitted power
calculated using traditional methods.
III. Effect of multiple transmitters
It is of interest to investigate the effect of multiple transmitters on the compliance boundary since
base station sharing between operators and sites containing transmitters of different technologies
are very common. The data consisting of the cumulative distribution functions of the normalized
transmitted power obtained from [1] is analyzed for different configurations of: emitted power, traffic
conditions, and number of users in a cell. To study the effect of combining the emissions, the data
must be represented by a statistical distribution. The beta distribution, having a positive PDF, and a
finite support [0 1] is chosen since we are dealing with a normalized transmitted power, and having a
non-limited distribution may produce unwanted results [3].
The parameters governing the shape of the beta distribution are determined by curve fitting for each
data set. In Fig. 1 the CDF of the normalized power for the antenna configuration: UMi K=1, D=60s
is represented alongside the CDF of the combination of two of these emissions.
The combination is considered as the sum of the independent random variables following the beta
distribution, and constituting the two emissions.

As formulated in eq. 1 the random variable Y, representing the combined emissions, is the sum of
the N independent random variables. Each of these random variables has a CDF given by eq. 2

Fig. 1 shows that for the higher percentiles, the normalized actual transmitted power gets smaller
when combining multiple transmitters than when considering a single transmitter. The notion of

218
power combination factor is introduced to represent the ratio between the exposure assessed at the
given quantile and the one at maximum maximorum. It is applicable to the actual transmitted power
when combining multiple transmitters. The PCF is calculated for a one antenna configuration
reference. Table I contains the values of the PCF for the combination of multiple identical antennas.
TABLE I
POWER COMBINATION FACTORS FOR MULTIPLE ANTENNAS SHARING A SINGLE SITE FOR
UMi K=1, D=60s
Percentiles 20% 40% 50% 60% 95% 99%

1 Tx (Ref) 1.00 1.00 1.00 1.00 1.00 1.00

2 Tx 1.284 1.113 1.056 1.007 0.833 0.785

3 Tx 1.411 1.156 1.074 1.004 0.765 0.701

4 Tx 1.488 1.181 1.083 1.001 0.724 0.652

5 Tx 1.539 1.196 1.088 0.998 0.697 0.619

A. Antennas with the same maximum power
The transmit power of the antenna represents its weight in the statistical combination. Antennas with
equal power on the same transceiver are due to multiple operators sharing the station for same-
technology services. While the antennas are assumed to have equal power, the
configuration/technologies can differ, producing different distributions. In Fig. 2 the result of the
combination of two emissions with the same statistical weight, i.e. maximum power, is presented.
The combined emissions produce a CDF bounded by the other two. And since the two emissions
are assumed to have the same maximum transmit power, the transmitted power at higher
percentiles is almost always bigger than the powers at the same percentiles for both the two
emissions. This might not be the case for different maximum power ratios between the transmitters.
B. Antennas with different powers
To investigate the case of combining two transmitters with different power ratios, the data extracted
from the UMi K=1 D=60s is used. Also, a scaled emission is introduced. The combination was
investigated for varying ratios ALPHA between the emissions.
The results of the power combination factor (ratio between the exposure at the given quantile and
the one at maximum maximorum) at 95th and 99th percentiles are presented in the table below.
TABLE II
PCF FOR TWO ANTENNAS WITH DIFFERENT POWER RATIOS

Ratio 95th Percentile 99th Percentile

0.1 0.959 0.947

0.2 0.916 0.897

0.3 0.895 0.870

0.4 0.874 0.843

0.5 0.857 0.821

219
 0.6 0.846 0.803

0.7 0.841 0.802

0.8 0.838 0.790

0.9 0.835 0.786

1 0.833 0.785

IV. Conclusions
Beta distribution can accurately represent the emissions of the massive MIMO antennas, and it can
be easily manipulated to simulate the different scenarios of installation and configuration of sites.
The analyzed data shows that the combination of two emissions will always result in a lower actual
normalized power transmitted value at higher percentiles. The results of the combination of multiple
emissions with different configurations in the number of users, traffic conditions and power ratio,
enables us to determine the scaling of the compliance boundary when multiple antennas are
present.
V. Acknowledgments
The authors thank Christophe Grangeat from Nokia and Christer Törnevik from Ericsson for the
fruitful discussion. The authors thank also Christophe Grangeat for the data provided.
VI. References
[1] Paolo Baracca, Andreas Weber, Thorsten Wild, Christophe Grangeat, “A Statistical Approach for
RF Exposure Compliance Boundary Assessment in Massive MIMO Systems”, arXiv:1801.08351.
[2] Björn Thors, Anders Furuskär, Davide Colombi, Christer Törnevik, “Time-Averaged Realistic
Maximum Power Levels for the Assessment of Radio Frequency Exposure for 5G Radio Base
Stations Using Massive MIMO”, IEEE Access, Vol. 5, 18 September 2017.
[3] Joe Wiart, “Statistical analysis of the Radio Frequency electromagnetic fields exposure induced
by base stations with multiple Massive MIMO transmitters”, 2nd URSI AT-RASC, Gran Canaria, 28
May – 1 June 2018.

Figures

220
Figure 1. CDF of the UMi antenna with K=1 and D=60s alongside the CDF of the combination
of two of these antennas.

Figure 2. Combination of two power distributions with different configurations and same
maximum transmit power

221
S11-5 [12:00]
Exposure assessment of 60 GHz communication antenna and 79 GHz automotive radar
Gunter Vermeeren1, Sven Kuehn2, Björn Debaillie3, Guy Torfs4, Niels Kuster2, Piet Demeester4, Wim Van
Thillo3, Luc Martens1 & Wout Joseph5
1imec - WAVES, Department of Information Technology at Ghent University, Ghent, Belgium, B-9052
2IT’IS Foundation, Zürich, Switzerland, 8004
3imec - IoT, Leuven, Belgium, B-3001
4IDlab, Department of Information Technology at Ghent University - imec, Ghent, Belgium, B-9052
5WAVES, Department of Information Technology at Ghent University - imec, Ghent, Belgium, B-9052
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (published)
Presented by: Gunter Vermeeren
We measured the exposure, in terms of spatial-averaged power density, of a 60 GHz wireless
communication module and 79 GHz automotive radar at distances less than 30 mm of the antenna.
We assessed the upper limit on the exposure by transmitting maximum power and at 100 % duty-
cycle and compared this upper limit to the ICNIRP and FCC basic restrictions. The upper limit
complied with ICNIRP basic restrictions, but exceeded FCC basic restrictions at short distances of
the antennas. This difference in compliance is mainly due to the difference in averaging area and
limits specified by both guidelines.
Introduction
Two emerging applications of millimeter-wave electromagnetic fields are (automotive) radar and
(short-range) wireless communications. The increasing demand for high – Gbit – data rates and
mobility drives advances in wireless communications. The next step in the evolution of wireless
communication technologies is 5G. To cope with the demand for high data rates, millimeter-wave
communication bands will be introduced with 5G communication technologies; targeted millimeter-
wave frequencies are 28 GHz and 60 GHz. Automotive radar is another application that utilizes
millimeter-wave frequencies. Advances in these radar applications are driven by the development of
self-driving vehicles, such as, cars, which rely strongly on various sensing technologies, amongst
them automotive radars. In both cases, antennas can be positioned in close proximity – within a few
centimeters – of the human body. An important aspect in the development of these devices is the
assessment of the localized exposure.
The objective of this study was to assess experimentally an upper limit – for maximum output power
and a 100% duty cycle – on the exposure of a 60 GHz communication module and a 79 GHz
automotive radar module and evaluate the measured exposure, in terms of spatial-averaged
incident power density, against current ICNIPR and FCC guidelines.
Materials and methods
We measured the near fields in the proximity of a radiating 60 GHz communication module (60 GHz
16-elements phased array beam steering antenna for high data rate WiFi and 5G small cell
backhaul [3]) and a 79 GHz automotive radar antenna [4]. To assess an upper limit on the exposure,
we applied maximum power and a 100% duty-cycle for both antennas. Figure 1 shows the
measurement setup consisting of a EUmmWV2 near-field probe (SPEAG, Zürich, Switzerland)
connected to a DASY6 (SPEAG, Zürich, Switzerland) and a module under test. The EUmmWV2 is
designed for near field measurements in the millimeter-wave range up to a frequency of 110 GHz
with a dynamic range of <50 – 3000 V/m, and a linearity error of less than ±0.2 dB. We positioned
the antennas horizontally resulting in an upwards radiation and measured the exposure in horizontal
planes at distances of 3 mm, 5mm, 20 mm, and 30 mm above the antenna surface. We compared
the measured spatial-averaged power density, in W/m2, with ICNIRP [1] and FCC guidelines [2].
ICNIPR specifies two limits: 10 W/m2 averaged over 20 cm2 and 200 W/m2 averaged over 1 cm2;
FCC specifies a limit of 10 W/m2 averaged over 1cm2 at these frequencies.

222
Results
Figure 2 shows the distribution of the power density (S) of the 60 GHz communication antenna and
the 79 GHz automotive radar antenna; Figure 3 shows the spatial-averaged power density (Savg)
over 1 cm2 and 20 cm2 with the height above the antenna for both millimeter-wave antennas. As we
opted to determine an upper limit on the exposure by both devices, the devices were transmitting at
maximum power with a duty-cycle of 100 %. These configuration settings will allow us to easily
rescale exposure values and evaluate compliance in case of realistic exposure scenarios.
We observed that the upper limit on the exposure complied current ICNIRP basic restrictions at all
the measured distances from the antenna surface, but it did not comply with FCC basic restrictions
at short distances of the antenna surface. We extrapolated the compliance distance: the radar
antenna complied FCC basic restrictions for distances larger than 23 mm; the communication
antenna complied FCC basic restrictions for distances larger than 50 mm. As the antenna
configurations were set to provide an upper limit the exposure, the presented exposure values will
overestimate realistic exposure conditions. The different compliance outcome when compared with
ICNIRP and FCC basic restrictions are due to the difference in the specified averaging area and
limits by these guidelines.
Conclusions
We measured experimentally the exposure of a 60 GHz millimeter-wave communication module and
a 79 GHz automotive radar antenna. We assessed the exposure for maximum power and 100 %
duty cycle. A large difference in compliance distance between FCC and ICNIRP guidelines is
observed for millimeter-wave electromagnetic field exposure. This is due to the difference in
averaging area and limits specified by both guidelines.
References
[1] International Commission on Non-Ionizing Radiation Protection, 1998. Guidelines for limiting
exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). Health
physics, 74(4), pp.494–522.
[2] Evaluating Compliance with FCC Guidelines for Human Exposure to Radiofrequency
Electromagnetic Fields, Federal Communication Commission, OET Bulletin 65, Edition 97-01.
[3] G. Mangraviti, K. Khalaf; Q. Shi, K. Vaesen, D. Guermandi, V. Giannini, S. Brebels, F. Frazzica,
A. Bourdoux, C. Soens, W. Van Thillo; P. Wambacq, "13.5 A 4-antenna-path beamforming
transceiver for 60GHz multi-Gb/s communication in 28nm CMOS," in IEEE International Solid-State
Circuits Conference (ISSCC), Feb. 2016.
[4] D. Guermandi, Q. Shi, A. Dewilde, V. Derudder, U. Ahmad, A. Spagnolo, I. Ocket, A. Bourdoux,
P. Wambacq, J. Craninckx, W. Van Thillo, “A 79-GHz 2 × 2 MIMO PMCW Radar SoC in 28-nm
CMOS,” Journal of Solid-State Circuits, Vol. 52, Issue 10, Oct. 2017
Figures

223
Figure 1. The measurement setup consisting of the precision millimeter-wave near field probe
connected to a DASY6 measurement system (SPEAG, Zürich, Switzerland) and the
millimeter-wave communication device.

Figure 2. Power density distribution measured in a horizontal plane at 5 mm above the surface
of (a) the communication module antenna at 60 GHz and (b) the automotive radar antenna at
79 GHz.

224
 Figure 3. Comparison of the measured spatial-averaged power density from the 60 GHz
communication antenna and 79 GHz radar antenna with current ICNIRP and FCC basic
restrictions. The shown spatial-averaged power densities provide an upper limit on the
exposure as maximum power and 100% duty cycle was used. Hence, the presented exposure
values overestimate realistic exposure conditions.

S11-6 [12:15]
Microwave hyperthermia treatments of palms: a promising tool of control against
Rhynchophorus ferrugineus (Coleoptera: Curculionidae)
Rita Massa1, Daniele Pinchera2, Marco Donald Migliore2, Fulvio Schettino2, Emilio Caprio3, Raffaele Griffo4,
Manuela Martano5, Paola Maiolino5 & Gaetano Panariello2
1Physics Department “Ettore Pancini”, University of Naples Federico II, Naples, Italy, 80126
2Department of Electrical and Information Engineering "Maurizio Scarano" , University of Cassino and
Southern Lazio, Cassino, Italy, 03043
3Department of Agricultural Sciences, University of Naples Federico II, Naples, Italy, 80100
4Plant Protection Service, Campania Region, Naples, Italy, 80100
5Department of Veterinary Medicine and Animal Production, University of Naples Federico II, Naples, Italy,
80100
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Rita Massa
This work is aimed at demonstrating the feasibility of the microwave heating as a tool to contrast the
Red Palm Weevil (RPW) Rhynchophorus ferrugineus pest, which is currently affecting many palm
species all over the world. Microwave technology is effective and eco-compatible, and suitable for
use within an Integrated Pest Management (IPM) strategy. In this work we demonstrate the
225
microwave heating allows to reach insect lethal temperature, without affecting the internal palm
tissues, when an electromagnetic-thermal protocol specially developed is applied.
Introduction
The Red Palm Weevil (RPW) Rhynchophorus ferrugineus is the major destructive insect pest of a
broad range of palm species. In the Mediterranean area it spread slowly during the mid-1990s (in
1993 it was detected in Spain), then it expanded quickly during the years between 2004-2009 [1],
reaching USA (California) in 2010. RPW is a serious pest of coconut, oil palm, sago palm, date palm
and in the Mediterranean Basin it affects mainly Phoenix canariensis Hort. ex Chabaud. A rapid
death of thousands of these ornamental plants resulted, involving monumental specimens in historic
sites, some of which included in the UNESCO World Heritage Site list. In addition to the important
economic consequences, big concern arises for both landscape preservation and public safety due
to the abrupt crash of the collapsed plant. The RPW is currently managed employing an Integrated
Pest Management (IPM) strategy and, in this context, there is a considerable interest toward
solutions able to control the pest with a minimum impact on the environment. Microwave heating of
infested palms is a promising eco-compatible technology to contrast the pest diffusion. In this work
we present the results of palm treatments with a microwave ring applicator. We demonstrate that
insect lethal temperatures can be reached, without affecting the internal palm tissues, and that
microwaves can influence the physiology of the survived weevils. In addition, a better control of the
temperature profile, inside microwave treated palm tissues, can be achieved when an
electromagnetic-thermal protocol specially developed is applied.
Methods
The feasibility of microwave application on infested palm to contrast RPW has been carried out with
both laboratory and semi-field tests. In order to identify the duration of the microwave treatment to
reach a desired temperature in the palm, both electromagnetic and thermal parameters of the
tissues involved were measured, as well as the lethal dose temperature for the insect in the different
stages.
Permittivity measurements were carried out by means of the truncated coaxial cable. Broad-band
electromagnetic characterization was performed in the 0.4 -18 GHz frequency range of both healthy
and damaged tissues of the P. canariensis palm, with different moisture content, as well as of the
RPW in different stages (larva, pupa and adult) [2]. The thermal conductivity and the specific heat
capacity of the palm tissues were estimated during both the heating and the cooling process of a
microwave treated palm in a rigorous and controlled way with a WR340 waveguide opportunely fed
by a magnetron (1 kW incident power). Temperature measurements were conducted with non-
perturbative probes placed at different depths along the radius of the palm [3].
The thermal death kinetics of RPW, for the estimation of lethal temperature and time combinations,
was obtained with the application of hot air to increase the ambient temperature to get a preliminary
indication of the thermal limits of survival of the pest [1]. In addition, in laboratory test, we monitored,
in weevils exposed in a waveguide applicator [4], the relative increase of surface temperature (by an
infrared thermocamera), the survival days and the histological alterations of ovaries and testes.
Finally, the electromagnetic-thermal model was performed by means of an FDTD approach and
validated calorimetrically by non-perturbative measurements (infrared thermocamera or fiber optic
thermometer). The temperature increase was recorded when the palms were treated with a
commercial ring microwave applicator (EcoPalm, patented by Bi.Elle s.r.l.). The applicator consists
in 12 open waveguides fed by 12 magnetrons (2.45 GHz, 1 kW nominal power) and arranged in a
ring that can be closed around the palm (Fig.1) in order to surround a section of the trunk (typically
the crown were the insect prefers to stay).
Results
The thermal kinetics study indicated that a temperature of 55 °C for 30 min assures a probability of
mortality higher than 90% for all stages of the weevil. Thus the electromagnetic and thermal model
was developed in order to reach this temperature in the area were the pest is supposed to remain.

226
In addition we found that high power (2.45 GHz, 5.4 W/cm2) short exposures (5, 15, 30 sec) can
affect the survival time and reduce or remove completely the reproductive capacity of insects (both
male and female) [5]. The effects raise with the increase of irradiation time and the relative
temperature rise. In particular, results indicate that, even though the maximum temperature
recorded was less than 50 °C (30 s treatments), the survival time decreased with respect to that of
control. In addition an impact of radiation on ovary and testis was found. The former could lead to an
incomplete mating due to non generation of eggs, or sterility of eggs, while the latter could induce
loss of germinal cells, which progressively could induce a total lack of germ cells, with consequent
arrest of spermatogenesis. In Fig. 2 a,b are reported, respectively, the average survival days
observed at the different exposure duration (0, 5, 15, 30 sec) and the score assigned for each
samples: 0- no pathological alterations; 1 mild pathological alterations; 2- moderate pathological
alterations; 3- severe pathological alterations.
Concerning the palm treatments, the broad band permittivity measurements showed a low
penetration depth and a low thermal conductivity. The measurements on the RPW surface indicate
that the conductivity of larvae is higher than that of the adult and the penetration depth in the
chamber is of many centimeters thus allowing the microwaves to reach the larva/adult depending on
the stage of the metamorphosis. We have experimentally verified that only the external region of the
palm is deeply involved in the microwave heating [3]. In this portion eggs, small larvae, cocoons and
adults are directly reached by the electromagnetic radiation and a rapid increase of the temperature
is expected due to both the insect and palm tissues heating.
All these results hint that microwave applications can be effective if the annular area, where the
insects are foreseen to stay, reaches a temperature of 55°C for at least 30 min. However shorter
interactions with the RPW adults can affect the weevils due to the limited physiological capacity to
regulate their body temperature.
On the basis of these data, we developed a simple mono-dimensional electromagnetic-thermal code
[3] able to predict the thickness of the annular area in “living” palms wherein the temperature
reaches the lethal temperature for RPW, and how long it should stay in order to have a high
probability that the treatment is successful without damaging the palm core that is the most relevant
area for the life of the palm itself. Assuming that during irradiation the evaporating water is rapidly
substituted by other water drained by the plant, that temperature dependence of thermal and
electromagnetic parameters can be neglected, a cylindrical symmetry illumination around the trunk
of the palm and a longitudinal invariance, all physical quantities in both thermal and electromagnetic
models only depend on the radial coordinate and time. Thus the analysis has been easily carried out
by means of an FDTD, being the thermal and electromagnetic models uncoupled. A very good
agreement was found when we compared the results obtained by the simplified model with both
commercial CAD (Ansoft) and measurements on treated (both young and older) P. canariensis palm
with Ecopalm microwave applicator.
Conclusions
Our work has shown that microwave heating is an eco-compatible way to contrast the RPW pest,
the economic importance of which, as a destructive pest for palm species was recognized since the
1800’s [6]. The results show that microwaves could induce a thermal increase capable of killing or
affecting the insects without damaging the palm tissues. We are planning additional palm treatments
in order to confirm our preliminary observations that palm functions are not damaged, being only the
external layer treated slightly dehydrated. Microwaves can be used as a curative treatment of
infected palm plants, but they can also be employed as a preventive treatment in the production of
healthy plants for planting, as well as for disposal of unrecoverable palm trees that are in advanced
stage of infestation. No re-infestation can be avoided, and the efficacy can be improved by
combining different techniques that need to be chosen and applied according to each particular
case.
Acknowledgements
The project MIPALM was supported in the framework of PSR Campania 2007–2013 Misura 124

227
MIPALM CUP N.B95C12000040004
References
[1] R. Massa, E. Caprio, M. De Sanctis, R. Griffo, M.D. Migliore, G. Panariello D. Pinchera, P.
Spigno, “Microwave Treatment for pest control: the case of Rhynchophorus ferrugineus in Phoenix
canariensis” EPPO Bullettin, vol. 41.2, pp. 128-135, August 2011
[2] R. Massa, M. D. Migliore, G. Panariello, D. Pinchera, F. Schettino, E. Caprio, and R. Griffo “Wide
Band Permittivity Measurements of Palm (Phoenix canariensis) and Rhynchophorus ferrugineus
(Coleoptera Curculionidae) for RF Pest Control.” Journal of Microwave Power and Electromagnetic
Energy, vol. 48, issue 3, pp. 158-169, 2014
[3] R. Massa, G. Panariello, D. Pinchera, F. Schettino, E. Caprio,R Griffo, MD Migliore,
“Experimental and numerical evaluations on palm microwave heating for Red Palm Weevil pest
control”, Nature Scientific Reports, doi:10.1038/srep45299, 2017
[4] S.Romeo, C. D’Avino, D.Pinchera, O.Zeni, MR Scarfì, R. Massa, “A waveguide applicator for in
vitro exposures to single or multiple ICT frequencies.” IEEE Trans Microw Theory Tech.,
61(5):1994-2004, 2013
[5] M. Martano, R. Massa, I. Pagano, A. Greco, G. De Leva, A. Linguadoca, P. Maiolino “Histological
Findings of the male and female reproductive system of the RPW (Rhynchophorus ferrugineus)
irradiated with microwaves”, IX International Symposium on Wild Fauna, Slovakia, 2015
[6] El-Mergawy and A. M. Al-Ajlan, “Red Palm Weevil, Rhynchophorus ferrugineus (Olivier):
Economic Importance, Biology, Biogeography and Integrated Pest Management”, Journal of
Agricultural Science and Technology, vol. A1, pp. 1-23, 2011
Figures

Figure 1. Ecopalm Ring applicator patented by BIELLE surrounding the palm.

228
Figure 2. RPW adult survival days (a) and score assigned (b) after 5, 15, 30 sec microwave
exposure (0- no pathological alterations; 1- mild pathological alterations; 2- moderate
pathological alterations; 3- severe pathological alterations).

Figure 3. Thermogram acquisition of a cut palm during just after treatment

229
 Session: S12
Nano pulses
Thursday June 28, 2018 • 13:30 - 15:00
Europa C
Chairs: P. Thomas Vernier & Francesca Apollonio

S12-1 [13:30]
Nanosecond Pulsed Stimulation study of intracellular organelle changes in liver tumor cells
in vivo using electron microscopy
Richard Nuccitelli1, Snjezana Anand1, Holly Hartman1 & Darrin Uecker1
1Biology Group, Pulse Biosciences, Hayward, CA, USA, 94545
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Richard Nuccitelli
The intracellular effects of Nano-Pulse Stimulation (NPS) on rat liver tumors cells directly observed
by EM shows that NPS treatment doubles the size of mitochondria and induces swelling and
fragmentation of the endoplasmic reticulum. The Golgi apparatus vesiculates at 5 min after NPS
treatment and is no longer visible at later time points. The nucleus shrinks and the nuclear envelope
appears less distinct by 2h after NPS treatment.
Introduction
When cells or tissues are exposed to an electric field, the ions within the cells and organelles
normally move along the field lines until they abut the surrounding membrane and charge the
membrane capacitance. This ion redistribution generates an equal and opposite electric field so that
the net field within the cells and organelles is zero. However, this redistribution of ions occurs over
about 1 µs. When the rise time of the imposed field is less than a microsecond, the imposed field
penetrates both the cell and organelles. Nano-Pulse Stimulation (NPS) applies pulsed electric fields
with pulse durations of 100-800 ns to cells and tissues. When the imposed field is sufficiently large
to generate 500 mV across the organelle membranes, water dipoles are be forced into the lipid
bilayer, resulting in the formation of transient nanopores through which water and small molecules
such as ions can flow. This movement of ions and water molecules exerts changes on intracellular
organelles. This study demonstrated the distinct changes to the intracellular organelles of rat liver
tumor cells caused by NPS exposure in vivo using electron microscopy.
Methods
One week after injecting syngeneic McA-RH7777 liver tumor cells into one lobe of a Buffalo rat liver,
a 5 mm-wide tumor formed. After surgically exposing the liver and applying 300 ns-wide pulses
delivering a total of 6 joules to the tumors, the tumors were removed for fixation at either 5 min, 30
min, 2h or 4 h post treatment. Both control and treated samples were fixed in 2% glutaraldehyde,
1% paraformaldehyde in 0.1M sodium cacodylate buffer pH 7.4, post fixed in 2% osmium tetroxide
in the same buffer, en bloc stained with 2% aqueous uranyl acetate, dehydrated in acetone,
infiltrated, and embedded in LX-112 resin (Ladd Research Industries, Burlington, VT). Samples were
ultrathin sectioned (65 nm) on a Reichert Ultracut S ultramicrotome and counter stained with 0.8%
lead citrate. Grids were examined on a JEOL JEM-1230 transmission electron microscope (JEOL
USA, Inc., Peabody, MA) and photographed with the Gatan Ultrascan 1000 digital camera (Gatan
Inc., Warrendale, PA). This work was conducted by the Gladstone Institute of the University of
California, San Francisco.
Results
The mitochondria and endoplasmic reticulum (ER) exhibited the greatest changes compared to the
other organelles of treated cells. The mitochondria swelled following NPS treatment, increasing from
a mean diameter of 0.4 mm to 0.8 mm in 2 h (fig. 1 and 2). Over the same time period the inner
mitochondrial membranes appeared less distinct and vesicles began to appear within the
230
mitochondria.
The ER cisternae swelled within 5 min and subsequently fragmented within 2 h following NPS
treatment. Ribosomes dissociated from the ER in many regions. Clusters of ribosome particles
became observable in some regions.
By 2 h post NPS treatment, the cell nucleus had contracted and the nuclear envelope appeared less
distinct.
By 30 min post NPS, the Golgi apparatus was no longer visible and small breaks in the plasma
membrane were observed.
Conclusions
High-resolution EM images of intracellular organelles in NPS-treated liver tumor cells document
rapid and progressive effects over time while preserving broader cellular integrity. Following NPS
treatment, the mitochondria and ER both swell and vesiculate, the nucleus shrinks and the nuclear
envelope becomes less distinct. The Golgi apparatus was vesiculating in the 5 min images and was
not present in the later electron micrograph time points.
Figures

Figure 1. Rat liver tumor cells treated in vivo with 6 joules of 300 ns-long pulses

231
 Figure 2. Average mitochondria size swells after NPS treatment in vivo

Figure 3. By 2 h ER structure is lost, dual membrane becomes single and vesiculated and
some ribosomes dissociate in patches

S12-2 [13:45]
Nanosecond bipolar cancellation of action potentials in nerve fibers
Maura Casciola1 & Andrei Pakhomov1
1Frank Reidy Research Center for Bioelectrics , Norfolk, Virginia, USA, 23508
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Maura Casciola
Advancements in pulsed power technology allow to deeply focus sub nanosecond electric pulses
(nsEP) in tissues, opening opportunities for non-invasive, targeted electrical stimulation. Our recent
publication [1] showed that 12-nsEP can elicit thousands of action potentials in sciatic nerves
without electroporative damage. However, the stimulation threshold for bipolar (BP) pulses, such as
232
the emitted ones, is higher than for unipolar (UP) already in the µs scale [2]. Here, using pulses in
the ns-µs range, we show for the first time that the shorter the pulse the higher the ratio between BP
and UP stimulation threshold, challenging the possibility to use sub-nsEP for remote stimulation.
1. Introduction
Electrical stimulation (ES) is widely used in clinical applications to treat different conditions, such as
refractory pain and Parkinson disease [3, 4]. The use of sub nanosecond electric pulses (nsEP) for
ES is promising due to the possibility to focus them deep into tissues by external devices (e.g.
antennas) [5], allowing for targeted and non-invasive excitation.
However, the ability of nsEP to elicit action potentials (APs) in excitable cells is still a topic of
investigation [6]. The main debate regards whether the excitation is induced “directly” by the action
of the electric field on voltage gated (VG) sodium channels or is a consequence of electroporation
(EP) induced depolarization of the plasma cell membrane. In a recent paper, using peripheral nerve
preparations, we showed that 12-nsEP (3.3–8.8 kV/cm) can elicit thousands of compound action
potentials (CAPs) without damage to fibers [1]. Our results suggest that a pulse as short as 12 ns
allows for the VG sodium channels to respond without EP involved. However, the mechanism
behind this response remains to be clarified.
It is known that the stimulation threshold for BP pulses is higher than for the unipolar (UP) ones
already for pulse durations of tens of µs [2, 7-9]. This phenomenon occurs because when the
polarity of the pulse is reversed the opening of VG channels is inhibited and the AP started by the
first phase is abolished (cancellation effect). To compensate for this reversal, the BP pulse requires
a higher amplitude to achieve the threshold.
Computational works, using µs pulses, predicted that the cancellation effect is more efficient when
the duration of the pulse is shortened [9], suggesting an extremely high stimulation threshold for
sub-ns and nsEP.
Here, for the first time, we studied the effect of BP pulse parameters on the inhibition of the AP
using pulses below the µs duration. Our results showed, indeed, that short pulses abolished more
effectively APs and that consequently the ratio between BP and UP stimulation thresholds increased
with the decrease of the pulse duration. These evidences challenge the use of sub-nsEP for remote
ES, since emitted pulses are BP by nature.
2. Methods
Experiments were performed on isolated sciatic nerve preparations from Rana catesbeiana frogs.
The nerves were isolated, ligated and submerged in a physiological solution (pH 7.3). Before the
experiments, the nerve was moved into a stimulation chamber and covered by sealing paste to
avoid drying.
The proximal end of the nerve, in contact with the stimulation electrodes, was subjected to either
conventional µs pulses (few V) from a 4100 isolated high power stimulator (A-M SYSTEMS,
Sequim, WA) or to ns pulses (up to 1 kV) from a custom made nsEP generator. The pulse amplitude
and shape were continuously monitored with a 5-GHz TDS 3052 oscilloscope (Tektronix, Beaverton,
OR).
The recording electrodes, positioned few cm after the stimulating ones, were connected to a DAM50
amplifier (World Precision Instruments, Sarasota, FL) and the Acqknowledge Software (BIOPAC
Systems, Inc., Goleta, CA). During different sets of experiments the peak of the CAP was
measured. Data are presented in graphs as mean values ± standard error for n independent
experiments. All the exposure conditions were randomized.
3. Results and Discussions
The most accredited hypothesis to explain excitation by pulses shorter than the response of voltage
gated (VG) sodium channels and BP cancellation of AP, is related to the passive membrane
repolarization [2, 7-9]. If the duration of the first stimulus is shorter than the response of VG sodium
channels, at the completion of the pulse the passive repolarization of the plasma membrane begins

233
before the channels open. In this case, the discharging time of the membrane is determined by its
electric properties and this slow passive process eventually allows the VG channel to respond. An
anodic pulse applied between the cessation of the cathodic stimulus and the initiation of the AP
accelerates the repolarization of the membrane and hampers the activation of the VG channel.
Thus, the inhibition of the AP can be considered a two-step process: the first phase initiates the
movement of the voltage sensing module of the channels, while the second one blocks it,
suppressing the AP generation.
In the first set of experiments, we tested how the amplitude of the second phase affects the
amplitude of evoked CAPs. The amplitude of the first phase was set constant at either 40, 60, 80, or
100 V (60 V evoked maximum response when applied as a UP stimulus). The amplitude of the
second phase was varied from 0% (i.e. UP pulse) to 200% of the first phase (Fig. 1, inset). Fig. 1
shows that: (a) increasing the second phase from 0 to 100% suppressed the evoked CAP; (b) the
suppression was stronger when the amplitude of the first phase was lower; (c) when the second
phase became larger than the first one, CAP was progressively restored.
"HERE FIG 1"
The goal of the second set of experiments was to quantify the effect of the duration of the pulse on
the inhibition of the CAP.
The phase duration was set at 200, 300, 700 ns and 10, 20, 50, and 300 µs. For each duration, the
amplitude of the first phase, when the second was the 0% (i.e. UP pulse), was adjusted to evoke
maximal response. The amplitude of the second phase was varied from 0 to 200% of the first phase.
Fig. 2 shows that: (a) decreasing the phase duration suppressed the evoked CAP more efficiently;
(b) 50 µs BP pulses produced a mild reduction of the CAP, that was completely absent for 300 µs.
One possible explanation for this data would be that the shorter the depolarizing pulse the less time
the membrane was charged, facilitating the repolarizing effect of the second phase, and reducing
the probability of channel opening.
"HERE FIG 2"
The effect of the introduction a delay was also investigated. The phase duration of symmetric pulses
was set at 200, 300, and 700 ns. For each phase duration the interpulse interval was varied from 0
to 100 µs. Fig. 3 shows that increasing the delay gradually restored the CAP, reaching its maximum
at 100 µs interpulse intervals, as predicted by [2].
Longer interpulse intervals allowed for more VG channels to respond, thus reducing the hindering
effect of the second phase. When the delay between the two phases overcame the opening time of
the VG channels, the second phase no longer affected the channel dynamics, abolishing the
cancellation phenomenon. From our results the time needed for all the channels to open was about
100 µs.
Fig. 3 also confirms that, for the same interpulse interval, short pulses better abolished the CAP
evoked.
"HERE FIG.3"
To quantify the difference between BP and UP stimulation thresholds we calculated the ratio of the
strength duration curves (S/D) of both stimuli. To build the S/D curves, for each phase duration the
amplitude of the pulse was increased until the minimum detectable response of the nerve was
recorded. Table 1 shows that the ratio BP/UP stimulation threshold increased as the duration of the
pulse decreased. Extrapolating the data from the S/D curves, we estimated that for 12-nsEP the
threshold of BP excitation would be 24-folds higher than the UP 12-nsEP.
Comparison of the BP/UP ratio at 2.5 and 10 µs with previous results [7] showed good agreement
with theoretical predictions.

Ratio BP/UP stimulation Ratio BP/UP from theoretical

Phase duration
threshold work [7]

234
 12 ns 24* -

300 ns 6.6 -

700 ns 4.7 -

2.5 µs 3.3* 2.7

10 µs 1.6 1.7

Tab. 1. Ratio of BP and UP excitation thresholds. The data is reported as mean of n = 5-10
independent measurements. (*) Extrapolated data.
4. Conclusions
In this work for the first time, using pulses below the µs range, we demonstrated that the threshold
for BP stimulation is higher the shorter is the pulse.
The abolition of the AP is due to a two-step process, where the application of the first phase initiates
the opening of the VG channel, while the delivery of a second phase of opposite polarity blocks it
and inhibits the AP generation. To compensate for this effect, higher amplitudes of the BP pulse are
necessary to reach the stimulation threshold.
We also observed that the suppression of the CAP occurs if the second phase is delivered before
the complete opening of the channels, i.e. ~100 µs.
Our results challenge the use of emitted sub-nsEP for non-invasive ES that may probably have an
extremely high stimulation threshold due to their short duration and the intrinsic bipolarity.At the
same time, the high degree of cancellation is most promisingfor remote stimulation by a
“cancellation of cancellation” (CANCAN) paradigm [10].
Acknowledgements
Supported by AFOSR 2015 MURI grant FA9550-15-1-0517 (to A.G.P.).
References
[1] M. Casciola, S. Xiao, and A. G. Pakhomov, "Damage-free peripheral nerve stimulation by 12-ns
pulsed electric field," Sci Rep, vol. 7, p. 10453, Sep 5 2017.
[2] C. van den Honert and J. T. Mortimer, "The response of the myelinated nerve fiber to short
duration biphasic stimulating currents," Ann Biomed Eng, vol. 7, pp. 117-25, 1979.
[3] C. Schlenstedt, A. Shalash, M. Muthuraman, D. Falk, K. Witt, and G. Deuschl, "Effect of high-
frequency subthalamic neurostimulation on gait and freezing of gait in Parkinson's disease: a
systematic review and meta-analysis," Eur J Neurol, vol. 24, pp. 18-26, Jan 2017.
[4] M. Leone and A. Proietti Cecchini, "Advances in the understanding of cluster headache," Expert
Rev Neurother, vol. 17, pp. 165-172, Feb 2017.
[5] R. A. Petrella, K. H. Schoenbach, and S. Xiao, "A Dielectric Rod Antenna Driven by a Pulsed
Power System," Ieee Transactions on Dielectrics and Electrical Insulation, vol. 24, pp. 2157-2163,
Aug 2017.
[6] A. G. Pakhomov, I. Semenov, M. Casciola, and S. Xiao, "Neuronal excitation and
permeabilization by 200-ns pulsed electric field: An optical membrane potential study with FluoVolt
dye," Biochim Biophys Acta, vol. 1859, pp. 1273-1281, Jul 2017.
[7] J. P. Reilly, V. T. Freeman, and W. D. Larkin, "Sensory effects of transient electrical stimulation--
evaluation with a neuroelectric model," IEEE Trans Biomed Eng, vol. 32, pp. 1001-11, Dec 1985.

235
[8] P. H. Gorman and J. T. Mortimer, "The effect of stimulus parameters on the recruitment
characteristics of direct nerve stimulation," IEEE Trans Biomed Eng, vol. 30, pp. 407-14, Jul 1983.
[9] R. Butikofer and P. D. Lawrence, "Electrocutaneous nerve stimulation-II: stimulus waveform
selection," IEEE Trans Biomed Eng, vol. 26, pp. 69-75, Feb 1979.
[10] A. G. Pakhomov et al., "Going bipolar with nanosecond pulses:How to control bioeffects with
the pulse shape," 2nd World Congress on Electroporation, September 24-48, 2017, Norfolk, VA,
USA.
Figures

Figure 1. Degree of cancellation as a function of the repolarizing phase amplitude. The phase
duration was fixed to 700 ns. The amplitude of the cathodic phase was 40, 60, 80 and 100 V.
The amplitude of CAP evoked by a UP pulse of indicated amplitude was taken as 100%. The
data is reported as mean and standard error of n = 6 independent measurements. Inset,
representative traces of a 700-ns UP and BP pulse.

236
Figure 2. Shorter pulses produced a higher degree of cancellation. The following phase
durations were applied: 200, 300, and 700 ns, 10, 20 50 and 300 µs. The amplitude of CAP
evoked by a UP pulse of indicated durations was taken as 100%. The data is reported as
mean and standard error of n = 4-5 independent measurements.

Figure 3. Increase of the interpulse interval reduced the cancellation until its complete abolition
at 100 µs. The following phase durations were applied: 200, 300, and 700 ns. The amplitude
of CAP evoked by a UP pulse of indicated durations was taken as 100%. The data is reported
as mean and standard error of n = 2-8 independent measurements.

237
S12-3 [14:00]
Detecting the structural changes of planar lipid bilayer exposing to nanosecond electric field
Peter Kramar1, Damijan Miklavčič1 & P. Thomas Vernier2
1University of Ljubljana, Ljubljana, Slovenia
2Old Dominion University, Norfolk, Virginia, USA
Keywords: Electrochemistry, Static, Work in Progress
Presented by: Peter Kramar
Planar lipid bilayers are often used to study electroporation phenomena. In this study, preliminary
experiments were carried out to measure changes in planar lipid bilayer resistance and capacitance
resulting from exposure to nanosecond pulsed electric fields. We also looked at the effects of
calcium and magnesium ions on bilayer electrical properties. The results raise several new
questions that challenge our basic understanding of membrane responses to nanosecond pulsed
electric fields.
The application of pulsed electric fields to biological cells has emerged as a powerful technique for
manipulation of cell membrane permeability, with wide-ranging use in biomedical and
biotechnological applications [1,2]. Pulsed electric field (i.e. applied electric pulses), induces local
distortions and structural rearrangements of lipid molecules in the cell membrane. As a result, the
membrane becomes transiently permeable even after the external field is removed, allowing
otherwise impermeant molecules to cross the membrane and reach the cytosol. This phenomenon
is often referred to as electroporation or electropermeabilization [3-5].
Although electroporation protocols are common in biomedicine and biotechnology [1,2], including
food processing [6,7], the fundamental biological and physical processes and molecular-scale
mechanisms involved in electroporation — the dynamics of electroporation-mediated molecular
transport and the associated membrane structural rearrangements — are still not completely
understood. Molecular dynamics (MD) simulations of small patches (100 nm2) of planar lipid bilayers
containing hundreds of lipids and associated interfacial and bulk water establish a voltage across
the lipid bilayer [8-10] that induces the formation of a lipid pore, which can be large and hydrophilic
[11,12] or small and hydrophobic [13,14]. Pore shape and size is dependent on lipid composition,
ion concentrations, and temperature [15].
Studies of electroporation (cell, tissue, and organ) are often categorized on the basis of pulse
duration, microsecond or nanosecond, even though the fundamental mechanisms are believed to be
similar. Microsecond electroporation utilizes pulsed electric fields less than 1 MV/m lasting for
microseconds. Nanosecond electroporation protocols call for fields greater than 1 MV/m with
durations less than 1 µs [16]. Furthermore, even shorter, sub-nanosecond pulses become an
interesting stimulating signal for studying electroporation phenomenon at cell respects to these
pulses [16-18]. For now, microsecond electroporation is far more studied than nanosecond or even
sub-nanosecond electroporation. To bridge this gap and to explain the dynamics of the events on
the cell membrane, a simple model was studied.
Planar lipid bilayers as a simple lipid membrane structures can be considered as a small patch of a
cell membrane and it are often used to study phenomena on electroporation in the experimental
approaches. Planar lipid bilayer can be electrically accessed from the both sides and represented as
electrical model of resistor and capacitor wired in parallel. However planar lipid bilayers are typically
built in specially designed chambers, filled with buffer solution, and as such cannot be observed
optically using microscope. Therefore all its properties and changes have to be measured
electrically. Measurement of voltage and current within the short intervals of applied nanosecond
electric pulses is not practical. The relatively high resistance and capacitance of the planar lipid
bilayer, buffer solution, and chamber reduce the dynamics of circuit response. Therefore, voltage
and current on planar lipid bilayers cannot be measured with the required high sampling frequency
of 3 GHz. To bypass this bottleneck a new measurement system was designed and constructed. In

238
this work, an LCR meter was used to measure capacitance and resistance.
In preliminary experiments, impedance changes due to planar lipid bilayer exposure to nanosecond
electric field were measured. To assess the system’s capability for the detection of changes in the
planar lipid bilayer, additional experiments were performed with the divalent ions Ca2+ and Mg2+ in
the surrounding buffer solution.
The electrical apparatus includes a nanosecond pulse generator, power supply, precision LCR
meter, analog switch for electrodes, and digital delay generator. The nanosecond pulse generator
and power supply were developed at the University of Ljubljana [19]. The pulse generator uses a
blumlein technique to generate pulses of duration 7 ns and variable amplitude from 0 to 60 V. An
analog switch was developed to connect electrodes either to the pulse generator or to a
measurement device. Four Ag-AgCl electrodes, two on each side, are immersed in buffer solution in
a two-compartment Teflon chamber divided by a partition with an aperture for bilayer formation. A
precision LCR meter Agilent E4980A, 20 Hz - 300 kHz, USA is used for measuring bilayer
capacitance and resistance in frequency and time regime. The LCR meter uses the bridge method
where the device under test is exposed to a sine wave of selected low voltage of 20 mV and
frequency from 20 Hz up to 300 kHz. From the current amplitude and the phase shift between
voltage and current, the LCR meter calculates capacitance Cp and resistance Rp. A Digital Delay
Generator DG645 (Stanford Research Systems, USA) is used to trigger switching the electrodes
between impedance measurement and pulse application, and the pulse generator trigger. The
complete system is controlled using customized scripts with Matlab 2017a and the Instrumentation
Control Toolbox.
Planar lipid bilayers were prepared from POPC (1-pamitoyl 2-oleoyl phosphatidylcholine) in powder
form (Avanti Polar Lipids Inc. USA). Lipids were dissolved in hexane:ethanol (9:1). A mixture of
hexadecane and pentane (3:7) was used for bilayer formation, with an aqueous buffer of 0.1 M KCl
and 0.01 M HEPES, adjusted to pH 7.4. Bilayers were formed on a 25 µm thick Teflon sheet with an
aperture 106 µm in diameter, which was placed in between the two compartments of the Teflon
Chamber. Each compartment was filled with 5 ml of the buffer solution. The folding method for
bilayer formation was used [20]. For some measurements, 5 µL aliquots of 1 M CaCl2 or MgCl2
were added to the buffer solution on one side of the bilayer. Each addition corresponds of 0.5 mM
CaCl2 or MgCl2 respectively.
The measurement protocol consisted of building the lipid bilayer and measuring its capacitance and
resistance first in the frequency regime (FM), then in the time regime (TM) for 400 seconds. In some
cases the Ca2+ or Mg2+ solutions were added or nanosecond pulse(s) were triggered. At the end of
each complete measurement sequence, the lipid bilayer was broken down by applying a 1.5 V DC
bias voltage, and the chamber electrical properties were measured without the bilayer in place.
We anticipated changes in resistance or capacitance caused by transmembrane current, charging of
the membrane interface, or bilayer thickness increase or decrease. In preliminary experiments
where the planar lipid bilayer was exposed to nanosecond electric pulses, however, no change in
measured resistance and capacitance was observed. We did observe changes after adding Ca2+ or
Mg2+, ions which are known to interact with phospholipid bilayers, with Ca2+-lipid interactions being
stronger.
Results of the measured planar lipid bilayer resistance and capacitance obtained after exposure to
nanosecond pulse rises several new questions on basic understanding of membrane mechanism
under nanosecond electric field, which requires additional work to be performed.
Acknowledgement:
This research was in part supported by Slovenian Research Agency. Research was conducted in
the scope of the EBAM European Associated Laboratory (LEA). Part of experiments were
performed within Infrastructure Programme: Network of research infrastructure centers at University
of Ljubljana (MRIC UL IP-0510). Experiments were performed during four month visit of Frank Reidy

239
Research Center for Bioelectrics, Old Dominion University, Norfolk, USA supported by Fulbright
scholarship, received by Peter Kramar for traveling and living in USA.The authors would like to
thank Matej Reberšek for dedicated nanosecond pulse generators and power supply development,
to Chunqi Jiang for lending LCR meter during the visit in USA.
References:
[1] Yarmush ML et al., Electroporation-based technologies for medicine: principles, applications, and
challenges. Annu. Rev. Biomed. Eng. 16: 295-320, 2014.
[2] Kotnik T et al., Electroporation-based applications in biotechnology. Trends Biotechnol. 33:
480-488, 2015.
[3] Neumann E et al., Permeability changes induced by electric impulses in vesicular membranes. J.
Membr. Biol. 10: 279–290, 1972.
[4] Weaver JC et al., Theory of electroporation: a review. Bioelectrochem. Bioenerg. 41: 135–160,
1996.
[5] Kotnik T et al., Tarek M. Cell membrane electroporation – Part 1: The phenomenon. IEEE Electr.
Insul. M. 28(5): 14-23, 2012.
[6] Mahnič-Kalamiza S et al., Electroporation in food processing and biorefinery. J. Membrane Biol.
247: 1279-1304, 2014.
[7] Puértolas E et al., Improving mass transfer to soften tissues by pulsed electric fields:
fundamentals and applications. Annu Rev Food Sci Technol 3: 263–82, 2012.
[8] Tieleman, D. et al., Simulation of Pore Formation in Lipid Bilayers by Mechanical Stress and
Electric Fields. J. Am. Chem. Soc. 125: 6382–6383 2003.
[9] Tarek M. et al., Membrane Electroporation: A Molecular Dynamics Simulation. Biophysical
Journal 88(5): 4045-4053, 2005.
[10] Tokman M et al., Lee JH, Levine ZA, Ho M-CC, Colvin ME, Vernier PT. Electric field-driven
water dipoles: nanoscale architecture of electroporation. PLoS ONE 8, e61111, 2013.
[11] Delemotte L et al., Modeling membranes under a transmembrane potential. J Phys Chem B
112: 5547–50, 2008.
[12] Kramar P et al., Molecular-level characterization of lipid membrane electroporation using
linearly rising current. J. Membrane Biol. 245: 651-659, 2012.
[13] Dehez F et al., Evidence of conducting hydrophobic nanopores across membranes in response
to an electric field. J. Phys. Chem. C 118: 6752-6757, 2014.
[14] Vernier PT et al., Nanopore Formation and Phosphatidylserine Externalization in a Phospholipid
Bilayer at High Transmembrane Potential. J. Am. Chem. Soc. 128, 6288-6289, 2006.
[15] Maček Lebar A et al., Internal configuration and electric potential in planar negatively charged
lipid head group region in contact with ionic solution. Bioelectrochemistry 111: 49-56, 2016.
[16] Batista Napotnik T et al., Effects of high voltage nanosecond electric pulses on eukaryotic cells
(in vitro): A systematic review. Bioelectrochemistry 110: 1-12, 2016.
[17] Vernier PT et al., Picosecond and Terahertz Perturbation of Interfacial Water and
Electropermeabilization of Biological Membranes. J. Membr. Biol. 248, 837–47, 2015.
[18] Semenov I et al., Electroporation by subnanosecond pulses. Biochem Biophysics Reports 6,
253–259, 2016.
[19] Reberšek M et al., Modeling and optimization of Blumlein nanosecond pulse generator for
experiments on planar lipid bilayers, in 2017 14th International Conference on Synthesis, Modeling,
Analysis and Simulation Methods and Applications to Circuit Design (SMACD), 2017.
[20] Montal M et al., Mueller P. Formation of Bimolecular Membranes from Lipid Monolayers and a
Study of Their Electrical Properties. Proceedings of the National Academy of Sciences 69:
3561–3566, 1972.

240
S12-4 [14:15]
STUDENT PAPER
Ion models and ion transport in molecular simulations of electrically stressed biological
membranes
Federica Castellani1, 2 & P. Thomas Vernier2
1Biomedical Engineering, Frank Batten College of Engineering and Technology, Old Dominion University,
Norfolk, Virginia, USA, 23508
2Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, Virginia, USA, 23509
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Federica Castellani
Standard ion models in the widely used CHARMM36 force field for molecular dynamics simulations
are not properly optimized for systems representative of biological membranes. For simulations of
Ca2+ interactions with lipid bilayers we have implemented ion parameters that result in more realistic
interactions between Ca2+, water, and phospholipids. With this improved model we have
characterized lipid pore-mediated ion transport in these systems by tracking the electric field- and
diffusion-driven passage of ions and other species through field-stabilized pores over time,
calculating the resulting currents and conductances, and relating these transport properties to the
pore geometry.
Introduction
Molecular dynamics simulations of lipid membranes facilitate the understanding of biological and
biophysical phenomena at the atomic and molecular level that cannot be observed with conventional
experimental methods, including electric-field-induced pore formation in lipid bilayers. Lipid
electropore formation in lipid bilayers can be analyzed under a wide range of conditions: number
and types of lipids, presence or absence of cholesterol and other non-phospholipid components of
the membrane, number and types of ions, physical conditions, thermodynamic ensembles, barostats
and thermostats, etc.
Interactions among atoms and molecules in molecular dynamics simulations are governed by sets
of properties, functions, and parameters called force fields, which express the potential energy of a
system of particles. The force field, a compromise between physical accuracy and computational
efficiency, provides a way to represent realistic atomic and molecular interactions by approximating
the more precise, but also much more computationally demanding, calculations of quantum
mechanics.
For our studies of membranes in electric fields, to better understand how the physical and
mechanical properties of the membrane constituents and the interactions among them are
influenced by the force field, we have compared material properties such as area per lipid, lipid
order parameter, ion coordination number, and ion binding for the older GROMOS-OPLS and the
newer CHARMM36 force fields. Close correspondence between the behaviors of model and
experimental systems in these properties and others increases our confidence in the overall
reliability of the simulations. We found serious discrepancies between the behavior of the
CHARMM36 calcium ion in aqueous solution and the experimentally determined properties of the
hydrated ion, and we report here the results of simulations that use an improved CHARMM36
calcium ion model.
Methods
Molecular dynamics simulations were performed using Gromacs 4.6.6, in an NPT ensemble, at 310
K (v-rescale thermostat) and 1 bar (Berendsen barostat). Periodic boundary conditions were applied
in all directions. The CHARMM36 force field was used for the lipids and the TIP3P model for water.
Calcium and chloride ions were added to a fully hydrated bilayer, containing 128 POPC (1-palmitoyl-
2-oleoyl-sn-glycero-3-phosphatidycholine), corresponding to an initial box of about 7 x 7 x 11 nm.
241
Different CaCl2 concentrations and different ion models were compared. A system containing KCl
was studied as a reference. All systems were equilibrated until the area per lipid stabilized. An
electric field (250 MV/m) was then applied along the z-axis to each system. Pore formation was
observed within a few ns. Once the pore expanded to a diameter of about 2 nm, the electric field
was decreased to different values to obtain stable pores of different radii. Pore radius and calcium
conductance were then calculated as previously described in (M. L. Fernández et al., 2012). At least
five trials were run for each electric field strength for each system. Fig. 1 shows the ion clustering
observed with the standard CHARMM36 calcium ion model.
Results
Calcium ions from the standard CHARMM36 force field distribution do not interact properly with
water and chloride ions. We improved the standard CHARMM36 Ca2+ model and tested the
behavior of Ca2+ in lipid bilayer systems. The standard CHARMM36 Ca2+ ion equilibrates with a
POPC bilayer faster than the improved model. Both models show approximately the same final
value for the area per lipid (0.62 nm2 for the standard model and 0.60 nm2 for the improved model),
as shown in Fig. 2. The two models have similar radial distributions of water molecules around the
ion. The improved Ca2+ model shows a CN (6) closer to the experimental value (6-8). It also takes
10 μs to reach binding equilibrium with a POPC bilayer, compared to 200 ns for the standard model.
With the improved Ca2+ model the dependence of pore radius on applied electric field is nonlinear.
The pore radius is stable only in a very narrow range of transmembrane potential (applied field), as
Fig 3 shows. Hypothesis: Ca2+ binding to POPC alters the free energy of electric-field-driven lipid
pores, resulting in a nonlinear pore energy profile and an apparent quasi-stable state, a local energy
minimum.
Acknowledgments
All simulations were run on the Old Dominion University High Performance Computing (HPC) Turing
cluster. Support for this work was provided by Air Force Office of Scientific Research
(FA9550-15-1-0517, FA9550-14-1-0123). FC received Batten College of Engineering and
Technology Graduate Fellowship from the Old Dominion University.
References
D. P. Tieleman, H. Leontiadou, A. E. Mark, and S.-J. Marrink, Simulation of Pore Formation in Lipid
Bilayers by Mechanical Stress and Electric Fields, J. Am. Chem. Soc., vol. 125, no. 21, pp.
6382–6383, 2003.
D. P. Tieleman, The molecular basis of electroporation, BMC Biochem., vol. 5, no. 1, p. 10, 2004.
M. Tarek, Membrane electroporation: a molecular dynamics simulation, Biophys. J., vol. 88, no. 6,
pp. 4045–4053, Jun. 2005.
Z. A. Levine and P. T. Vernier, Life cycle of an electropore: field-dependent and field-independent
steps in pore creation and annihilation, J. Membr. Biol., vol. 236, no. 1, pp. 27–36, 2010.
M. Kohagen, P.E. Mason, P. Jungwirth, Accurate description of calcium solvation in concentrated
aqueous solutions, J. Phys. Chem. B 118:7902-7909, 2014.
P. T. Vernier, M. J. Ziegler, R. Dimova, Calcium binding and head group dipole angle in
phosphatidylserine-phosphatidylcholine bilayers, Langmuir 25:1020-1027, 2009.
Z. A. Levine, P. T. Vernier, Calcium and phosphatidylserine inhibit lipid electropore formation and
reduce pore lifetime, J. Membr. Biol., 245 (10), 599-610, 2012
M. C. Ho, M. Casciola, Z. A. Levine, P. T. Vernier, Molecular Dynamics Simulations of Ion
Conductance in Field-Stabilized Nanoscale Lipid Electropores, J. Phys. Chem. B 117:11633-11640,
2013.
M. L. Fernández, M. Risk, R. Reigada, and P. T. Vernier, Size-controlled nanopores in lipid
membranes with stabilizing electric fields, Biochem. Biophys. Res. Commun., vol. 423, no. 2, pp.
325–330, 2012.
Figures
242
Figure 1. Lipid bilayer system with Ca2+ and Cl-. Cyan – water, gray – lipid hydrocarbon tails,
blue and gold – lipid head group atoms N and P, magenta – Ca2+, green – Cl-.

Figure 2. Area per lipid over time. On the right panel, bound Ca2+ to POPC lipids in purple and
area per lipid moving average in black.

243
 Figure 3. Pore radius versus external electric field.

S12-5 [14:30]
Biocontrol of skin infections causative Pseudomonas aeruginosa using nanosecond pulsed
electric fields: an in vivo study
Vitalij Novickij1, Auksė Zinkevičienė2, Eglė Lastauskienė3, Jurgita Švedienė4, Algimantas Paškevičius4,
Svetlana Markovskaja5, Jurij Novickij1 & Irutė Girkontaitė2
1Institute of High Magnetic Fields, Vilnius Gediminas Technical University, Vilnius, Lithuania, 03227
2Department of Immunology, State Research Institute Centre for Innovative Medicine, Vilnius, Lithuania,
08406
3Institute of Biosciences, Vilnius University, Vilnius, Lithuania, 10257
4Laboratory of Biodeterioration Research, Nature Research Centre, Vilnius, Lithuania, 08412
5Laboratory of Mycology, Nature Research Centre, Vilnius, Lithuania, 08406
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Vitalij Novickij
New in vivo data of successful eradication of skin bacterial contamination using nanosecond range
electroporation (EP) (20 kV/cm x 500 ns x 1000 pulses, 15 kHz) separately and in combination with
1% acetic acid (AA) is presented. Bioluminescent Pseudomonas aeruginosa and BALB/C mice were
used as a model. Combination of EP and AA 1% resulted in full eradication of bacteria in the
contaminated area. In all cases the muscle contractions were minimal due to ultrashort pulsing
protocols. The results are useful for the development of new methods for treatment of extreme
cases of wound infections when the chemical treatment is no longer effective.

244
Introduction
Skin infections, which are caused by drug-resistant pathogens represent a clinical challenge due to
being a frequent cause of morbidity and mortality, while burn wound infections are associated with
even more severe complications [1,2]. Electroporation (EP) may represent an important alternative
to drugs in addressing wound contamination by pathogens, particularly those contaminated with
multidrug-resistant species [3,4]. The proof of concept has been shown in vitro [5,6] and, recently, in
vivo using microsecond range electroporation [3]. We present new in vivo data of successful
eradication of bacterial contamination (Pseudomonas aeruginosa) using high frequency nanosecond
pulsed electric fields (PEF). We have also tested the additive effects of EP with acetic acid, which is
used clinically for treatment of bacteria-contaminated chronic wounds [7].
Methods
The bioluminescent P. aeruginosa ATCC27853 bacterial strain containing plasmid pAKlux2 (a kind
gift from Attila Karsi, Addgene plasmid # 14080) was used [8].In this study BALB/c mice were used.
For experiments, the flanks of the animals were shaved and depilated. Subsequently, suspension of
P. aeruginosa in PBS containing 6 × 109CFU/ml cells, was inoculated onto the surface of skin using
cotton swab. One hour later the mice were imaged with IVIS Spectrum (Calipter LifeSciences) using
Living Image Software to confirm bacterial contamination. The animals with bacterial contamination
were treated with 20kV/cm 500 ns pulses (a total of 1000 pulses) at repetition frequency of 15 kHz.
150 µl of distilled water or acetic acid 1% was used as a skin-electrode interface. The parallel plate
electrodes were applied (gap 1.5 mm). The whole shaved area was treated by adjusting the position
of electrodes. The treatment took less than 2-3 min. After the treatment mice were again analyzed
with IVIS Spectrum.The protocol was approved by the State Food and Veterinary Service and the
study was carried out in strict accordance with the recommendations in the Guide for the Care and
Use of Laboratory Animals. All applicable international, national and/or institutional ethical guidelines
were followed.
Results
All of the treatments were normalized to the untreated control. The exemplary images adequately
representing the treatment efficiency using different protocols are shown in Fig. 1. The
luminescence (representing the extent of bacterial contamination) increased up to 46±34% when the
contaminated area was treated solely by distilled water (Refer to Fig. 1A). However, the 20kV/cm
protocol allowed to achieve almost 16-fold reduction of luminescence (Fig. 1B). Acetic acid 1%
alone was also effective resulting in almost 10-fold reduction of luminescence (Fig. 1C). The
combination of EP and acetic acid 1% treatment was the most successful resulting in full eradication
of the bacteria in the contaminated area (Fig. 1D). In all cases when PEF was used the muscle
contractions were minimal due to high frequency and short duration pulsing protocols.
Discussion
Nanosecond range PEF pulses allow countering multiple limitations, which are experienced in
conventional micro-millisecond range electroporation. Firstly, ultrashort PEF protocols offer better
control of delivered energy and non-thermal treatment [9]. Additionally, non-thermal interactions of
the electric field with subcellular structures is possible [10]. Lastly, the reduced electrochemical
reactions [11] in the electrode-tissue interface and a more patient-friendly treatment with minimal or
no muscle contractions further improve the applicability of the methodology [12]. In our study, we
have shown that nanosecond PEF can be successfully applied separately and in combination with
acetic acid for potential treatment of skin infections. The possible mechanism of the effect (improved
treatment efficiency using PEF and acetic acid) presumably lies within the severe physical damage
of bacterial cells that is caused by electroporation. The exposure to PEF triggers permeabilization of
bacteria plasma membrane and disruption of the cell wall [13], which is the main bacteria’s barrier
against environment [14]. The improved bactericidal properties of treatment can be attributed to the
internal pH shock, which was induced due to the high permeabilization of bacteria cell wall and
uptake of acids [7].The results of our research are useful for the development of new methods for

245
treatment of extreme cases of wound infections, when the chemical treatment is no longer effective.
Combination of the procedure with antimicrobials allows the treatment to become more effective and
less painful and thus create a safety margin in energy control. The non-toxicity of the treatment,
minimal muscle contractions, capability to eradicate biofilms [4], the physical nature of the method
and improved selectivity due to nanosecond range pulsing protocols could be highlighted as the
main advantages of the methodology.
Acknowledgement
The research was funded by Research Council of Lithuania Towards Future Technologies
programme grant Nr. LAT-02/2016. Authors declare no conflict of interest.
References
[1] Y.K. Coban, Infection control in severely burned patients., World J. Crit. Care Med. 1 (2012)
94–101. doi:10.5492/wjccm.v1.i4.94.
[2] T.C. Rice, A.P. Seitz, M.J. Edwards, E. Gulbins, C.C. Caldwell, Frontline Science : Sphingosine
rescues burn-injured mice from pulmonary Pseudomonas aeruginosa infection, 100 (2016)
1233–1237. doi:10.1189/jlb.3HI0416-197R.
[3] A. Golberg, G.F. Broelsch, D. Vecchio, S. Khan, M.R. Hamblin, W.G. Austen, R.L. Sheridan, M.L.
Yarmush, Pulsed electric fields for burn wound disinfection in a murine model., J. Burn Care Res. 36
(2016) 7–13. doi:10.1097/BCR.0000000000000157.
[4] S.I. Khan, G. Blumrosen, D. Vecchio, A. Golberg, M.C. Mccormack, M.L. Yarmush, M.R.
Hamblin, W.G. Austen, Eradication of multidrug-resistant pseudomonas biofilm with pulsed electric
fields, Biotechnol. Bioeng. 113 (2016) 643–650. doi:10.1002/bit.25818.
[5] V. Novickij, J. Svediene, S. Markovskaja, E. Lastauskiene, J. Novickij, I. Girkontaite, Pulsed
electric field-assisted sensitization of multidrug-resistant Candida albicans to antifungal drugs,
Future Microbiol. (2018).
[6] V. Novickij, A. Grainys, J. Svedienė, A. Paškevičius, J. Novickij, Controlled inactivation of
Trichophyton rubrum using shaped electrical pulse bursts: Parametric analysis, Biotechnol. Prog. 32
(2016). doi:10.1002/btpr.2276.
[7] V. Madhusudhan, Efficacy of 1% acetic acid in the treatment of chronic wounds infected with
Pseudomonas aeruginosa: Prospective randomised controlled clinical trial, Int. Wound J. (2015)
1129–1136. doi:10.1111/iwj.12428.
[8] A. Karsi, M.L. Lawrence, Broad host range fluorescence and bioluminescence expression
vectors for Gram-negative bacteria, Plasmid. 57 (2007) 286–295.
doi:10.1016/j.plasmid.2006.11.002.
[9] M.B. Sano, C.B. Arena, M.R. DeWitt, D. Saur, R. V. Davalos, In-vitro bipolar nano- and
microsecond electro-pulse bursts for irreversible electroporation therapies, Bioelectrochemistry. 100
(2014) 69–79. doi:10.1016/j.bioelechem.2014.07.010.
[10] K.H. Schoenbach, B. Hargrave, R.P. Joshi, J.F. Kolb, R. Nuccitelli, C. Osgood, A. Pakhomov, M.
Stacey, R.J. Swanson, J.A. White, S. Xiao, J. Zhang, S.J. Beebe, P.F. Blackmore, E.S. Buescher,
Bioelectric effects of intense nanosecond pulses, IEEE Trans. Dielectr. Electr. Insul. 14 (2007)
1088–1107. doi:10.1109/TDEI.2007.4339468.
[11] D.E. Chafai, A. Mehle, A. Tilmatine, B. Maouche, D. Miklavčič, Assessment of the
electrochemical effects of pulsed electric fields in a biological cell suspension, Bioelectrochemistry.
106 (2015) 249–257. doi:10.1016/j.bioelechem.2015.08.002.
[12] C. Valdez, M.B. Jirjis, C.C. Roth, R.A. Barnes, B.L. Ibey, Nanosecond electric pulses modulate
skeletal muscle calcium dynamics and contraction, Proc. SPIE. 10066 (2017) 100660X.
doi:10.1117/12.2253693.
[13] F. Pillet, C. Formosa-Dague, H. Baaziz, E. Dague, M.-P. Rols, Cell wall as a target for bacteria
inactivation by pulsed electric fields, Sci. Rep. 6 (2016) 19778. doi:10.1038/srep19778.

246
[14] S. Haberl Meglic, T. Marolt, D. Miklavcic, Protein Extraction by Means of Electroporation from E.
coli with Preserved Viability, J. Membr. Biol. 248 (2015) 893–901. doi:10.1007/s00232-015-9824-7.
Figures

Figure 1. The mice before treatment on the top and the same mice after the treatment on the
bottom, where A – treated with distilled water; B – treated with distilled water and
electroporation; C – treated with acetic acid 1%; D – treated with acetic acid 1% and
electroporation. Combined treatment allows full eradication of bacteria.

247
S12-6 [14:45]
Remarkable differences in the response of voltage-gated Na+ current to a single or pair of
high intensity nanosecond electric pulses in adrenal chromaffin cells
Lisha Yang1, Sophia Pierce1, P. Thomas Vernier2, Indira Chatterjee3, Gale Craviso1 & Normand Leblanc1
1Department of Pharmacology, University of Nevada, Reno, Nevada, USA, 89557
2Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, Virginia, USA, 23508
3Department of Electrical and Biomedical Engineering, University of Nevada, Reno, NV, USA, 89557
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Lisha Yang

This study explored the effect of pairs of 5 ns electric pulses (NEP) on Na+ current (INa) in adrenal
chromaffin cells. Cell exposure to two NEP separated by a 1 s interval produced a transient
inhibitory effect on INa (8% at 5 MV/m; 15% at 8 MV/m). The recovery of INa following inhibition was
not seen with a single NEP. Reducing the time interval between the two pulses from 1 s to 0.1 s led
to similar effect on INa. A further reduction of the time interval to 10 ms abolished the inhibition of INa
at 5 MV/m. At 8 MV/m, INa could still be inhibited by a pulse pair but the current fully recovered and
even exceeded baseline. These results suggest that NEPs can modulate INa to allow for fine tuning
of excitability in chromaffin cells.
Keywords
Nanosecond electric pulses; Adrenal chromaffin cells; Whole-cell patch clamp; Voltage-Gated Na+
channel
Introduction
We have been investigating how 5 ns duration electric pulses of high intensity (>1 MV/m) affect
neural cell excitability, using isolated bovine adrenal chromaffin cells as a simple model of neural-
type cells [1-3]. Our research group recently showed that a single 5 ns, 5 MV/m pulse elicited
membrane permeabilization in whole-cell voltage clamped bovine adrenal chromaffin cells,
evidenced by the appearance of an instantaneous inward current permeable to Na+ at a holding
potential of -70 mV [4]. This event was accompanied by an instantaneous decrease (~ 4%) of fast
voltage-dependent Na+ current (INa) [5], which was in agreement with a previous study [6] reporting
that much longer pulses (300 and 600 ns) were able to directly modulate voltage-gated Na+ and
Ca2+ channels in GH3 and NG108 cells. The decrease in INa we observed in chromaffin cells was
associated with a decrease in maximal conductance without a change in its voltage-dependence [5].
Whether the application of a second pulse of identical characteristics leads to additive or different
effects on INa was the purpose of this study. We carried out whole-cell patch clamp experiments in
bovine chromaffin cells to explore the effects of a pair of 5 ns NEPs on fast voltage-dependent Na+
currents recorded under conditions designed to simulate near physiological ionic gradients.
Methods
Adrenal chromaffin cells were isolated by collagenase digestion of the medulla of fresh bovine
adrenal glands obtained from a local abattoir (Wolf Pack Meats, University of Nevada, Reno) and
maintained in suspension culture in Ham’s F-12 medium supplemented with 10% bovine calf serum,
100 U/ml penicillin, 100 µg/ml streptomycin, 0.25 µg/ml fungizone, and 6 µg/ml cytosine arabinoside
at 36.5°C under a humidified atmosphere of 5% CO2 as previously described [1-3]. For experiments,
single, isolated cells were obtained by treatment of cell aggregates with the protease dispase and
attached to fibronectin-coated glass coverslips which was placed in the chamber perfused with a
balanced salt solution (BSS) consisting of 145 mM NaCl, 5 mM KCl, 2 mM CaCl2, 1.2 mM

248
Na2HPO4, 1.3 mM MgCl2, 10 mM glucose and 15 mM Hepes, pH 7.4 at room temperature. Whole-
cell currents were recorded in voltage clamp mode using an Axopatch 200B amplifier and Digidata
1322A data acquisition system (Axon Instruments, Sunnyvale, CA), and pClamp software
(Molecular Devices, Sunnyvale, CA). The internal pipette solution consisted of 10 mM NaCl, 30 mM
KCl, 110 mM K-gluconate, 1 mM MgCl2, 10 mM EGTA, 3 mM MgATP, and 10 mM Hepes, pH 7.2.
A nanosecond pulse generator, designed and fabricated by Transient Plasma Systems, Inc.
(Torrance, CA), generated pulses that were 5 ns in duration and were delivered to a pair of
cylindrical, gold-plated tungsten rod electrodes (100 µm diameter) having a gap of 100 µm between
the electrode tips. After rupturing the plasma membrane to achieve the whole-cell recording mode,
the NEP-delivering electrodes were lowered to a predefined “working” position using the automated
targeting function of a motorized MP-225 micromanipulator (Sutter Instruments, Novato, CA). In this
“working” position, the patched cell was situated midway between the electrode tips, with electrode
tips positioned 40 µm from the bottom of the coverslip. A single pulse was applied to the cells at
E-field amplitudes ranging from 5 to 8 MV/m. The E-field distribution at the location of the target cell
was computed using the commercially available Finite-Difference Time-Domain (FDTD) software
package SEMCAD X (version 14.8.5, SPEAG, Zurich, Switzerland). Whole-cell membrane currents
were continuously recorded, except for the interval between 10 ms before and 8 ms after delivery of
a pulse, using an automated pulse exposure system controlled by a program written in LabVIEW [4].
Results and Discussion
Previous experiments examining the voltage-dependence currents of chromaffin cell established
that the early peak inward current in current-voltage (I-V) relationships was primarily composed of a
fast voltage-dependent INa that activated near –30 mV, peaked around +10 mV and reversed at ~
+56 mV [5]. The half-inactivation potential V0.5 (~ –45 mV) determined from the analysis of the
steady-state inactivation properties of the inward current was also consistent with this current being
predominantly composed of a fast voltage-dependent INa [5]. Then a constant voltage step protocol
consisting of 200 repetitive voltage steps to +10 mV applied every 3 seconds from a HP of -70 mV
for10 min was used to monitor peak INa before and after cell exposure to 5 ns pulse.INawas first
recorded for 60 s prior to NEPs exposure to establish baseline conditions. A single or pair of NEPs
applied at an E-field of either 5 MV/m or 8 MV/m was then delivered between the 20th and 21st
voltage steps of the 200 voltage step train.
A single pulse applied at an E-field amplitude of 5 MV/m resulted in an instantaneous decrease of ~
4% in peak INa that then declined exponentially to a level that was ~ 85% of the initial level after 10
min. The decrease in INa was not due to a change in either the steady-state inactivation or activation
of the Na+ channel but was associated with a decrease in maximal Na+ conductance. Results of
new experiments showed that application of a second 5 ns pulse applied 1 s after the first pulse
amplified the instantaneous inhibitory effect of a single pulse on INa (from ~ 4% to 8%).
Paradoxically, the inhibition of INa then reversed wherein INa recovered toward and even surpassed
the control level after 5 min. This recovery contrasted with the effect produced by a single pulse
alone whereby INa continued to decline after the instantaneous inhibition (INa after 10 min: control
(no pulse) 92 ± 0.2%, n = 16; single pulse 85 ± 0.2%, n = 9; two pulses 96 ± 0.2%, n = 9).
Decreasing the interval between the two pulses from 1 s to 100 ms slightly reduced the
instantaneous inhibition (from ~ 8% to 6%) and increased the recovery of INa (INa after 6 min: 1 s
delay: 96 ± 0.3%, n = 9; 100 ms delay: 98 ± 0.3%, n = 10). Moreover, when the time interval
between the two pulses decreased from 100 ms to 10 ms the instantaneous inhibition was nearly
abolished and INa completely recovered to baseline (INa after 6 min: 10 ms delay: 100 ± 0.3%, n =
4).
To further assess how NEPs affect chromaffin cell excitability, we explored the effects of a second 5
ns pulse on INa at the higher E-field amplitude of 8 MV/m. Our data showed that INa decreased
instantaneously by ~ 4% and then continued to further decline exponentially following a single 5 ns,

249
5 MV/m pulse while at higher E-field of 8 MV/m a single pulse produced instantaneous inhibitory
effect that was around twofold greater with no further decline (from ~ 4% to 9%) over the course of
10 min. The application of a second 5 ns pulse, 1 s after the first pulse, caused a significantly larger
instantaneous inhibitory effect (from ~ 9% to 15%) on peak INa of a single pulse but failed to recover
over a 10 min time course, an observation contrasting with that seen with two NEPs at 5 MV/m.
When decreasing the interval between the two pulses from 1 s to 100 ms the instantaneous
inhibition was reduced from ~ 15% to 11%, yet still higher than the effect of a single pulse on peak
INa. However, INa started to recover with this shorter interval, again contrasting with the effect
produced by either a single pulse, or two pulses with a 1 s delay whereby INa remained unaffected
after the instantaneous inhibition (INa after 6 min: single pulse 92 ± 0.2%, n = 10; two pulses with 1 s
delay 87 ± 3%, n = 9; two pulses with 100 ms delay 96 ± 0.2%, n = 5). Furthermore, when the time
interval between the two pulses decreased from 100 to 10 ms the instantaneous inhibition of INa
was similar but surprisingly the current completely recovered and even surpassed the baseline (INa
after 10 min: 1 s delay 85 ± 0.1%, n = 9; 100 ms delay 95 ± 0.2%, n = 5; 10 ms delay 105 ± 0.2%, n
= 9). Our data highlight remarkable differences in the response of this voltage-dependent channel
when chromaffin cells are exposed to single vs pulse pairs applied at different interpulse intervals
and E-field amplitudes.
Conclusion
The results of this study have shown that INa can be differentially inhibited in a manner that can be
manipulated by altering the time interval between two pulses and E-field amplitude. The application
of a second 5 ns pulse either amplified or cancelled the instantaneous inhibitory effect of the first
pulse on INa depending on the time interval between two pulses. The recovery of the INa was
enhanced with shorter time intervals between the two pulses. These findings could point to the
potential of NEPs to fine tune chromaffin cell excitability. Additional experiments are required to
determine whether NEPs directly affect the Na+ channel protein or indirectly through an alteration of
the lipid environment surrounding the channel.
Acknowledgements
This work is supported by grants from the Air Force Office of Scientific Research (AFOSR)
FA9550-14-1-0018, FA9550-14-1-0123 and MURI FA9550-15-1-0517.
References
[1] G. L. Craviso, P. Chatterjee, G. Maalouf, A. Cerjanic, J. Yoon, I. Chatterjee and P.T. Vernier.
Nanosecond electric pulse-induced increase in intracellular calcium in adrenal chromaffin cells
triggers calcium-dependent catecholamine release. IEEE Trans. Dielectr. Electric. Insul.,
16:1294-1301, 2009.
[2] G. L.Craviso, S. Choe, P. Chatterjee, I. Chatterjee and P.T. Vernier. Nanosecond electric pulses:
a novel stimulus for triggering Ca2+ influx into chromaffin cells via voltage-gated Ca2+ channels.
Cell. Mol. Neurobiol. 30:1259-1265, 2010.
[3] G. L. Craviso, S. Choe, P. Chatterjee, I. Chatterjee and P.T. Vernier. Modulation of intracellular
Ca2+ levels in chromaffin cells by nanoelectropulses. Bioelectrochemistry 87:244-252, 2012.
[4] J. Yoon, N. Leblanc, J. Zaklit, P.T. Vernier, I. Chatterjee and G.L. Craviso. Enhanced monitoring
of nanosecond electric pulse-evoked membrane conductance changes in whole-cell patch clamp
experiments. J Membrane Biol. 249:633–644,2016
[5] L. Yang, G.L. Craviso, P.T. Vernier, I. Chatterjee, N. Leblanc. Nanosecond electric pulses
differentially affect inward and outward currents in patch clamped adrenal chromaffin cells.
PLOSONE 12(7): e0181002, 2017.
[6] Nesin, A.M. Bowman, S. Xiao, A.G. Pakhomov. Cell permeabilization and inhibition of voltage-
gated Ca2+ and Na+ channel currents by nanosecond pulsed electric fields. Bioelectromagnetics

250
33:394-404, 2011.

Session: S13
Public Policy I
Thursday June 28, 2018 • 15:30 - 16:30
Europa C
Chairs: Peter Gajšek & Frederik Freudenstein

S13-1 [15:30]
STUDENT PAPER
Design of a low specific absorption rate microstrip patch antenna for WLAN applications
Niamat Hussain1, Min-Joo Jeong1, Ji Woong Park1, Hanul Bong1, Uktam Azimov1 & Nam Kim1
1Chungbuk National University, Cheongju, Korea, 28644
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Niamat Hussain
We presented the design of a microstrip patch antenna with low Specific Absorption Rate (SAR) for
WLAN applications. The antenna consists of a diagonally truncated patch and a ground plan, which
were patterned on both sides of RO4003 substrate. The antenna provided, good impedance
matching and radiation pattern with stable gain for the bandwidth of 5.1–6.06 GHz. Moreover, the
antenna showed 88% reduction in SAR1g and 75% in SAR10g values without using any extra back
reflector compared to the international SAR standards.
Introduction
Nowadays the wireless communication has become an indispensable part of human life. The
responsible component of wireless communication is an antenna. Thus, a variety of antennas (PIFA,
folded chips, folded loops, microstrip etc) having desired radiation characteristics have been
proposed to meet the user demands in wireless communication. The wireless device manufacturers
prefer low profile antennas with characteristics of high gain, high radiation efficiency, wider
bandwidth with multiband behavior. Microstrip patch antennas have become famous due to their
compact size, low cost, easy integration with the circuit board and improved performances.
Moreover, the World Health Organization has reported that the radiations from antennas especially
those which are used in personal gadgets like cellular phones, could cause cancer. To prevent the
users from electromagnetic exposure safety limits are imposed in terms of specific absorption rate
(SAR). Therefore, antennas with the low SAR is the main challenge while designing antennas for
wireless communication especially cellular phones. In this paper, a microstrip patch antenna with
low SAR that can be operated in WLAN band has been presented.
Model and Methods
The antenna consists of a diagonally truncated patch with a semicircle and a ground plan, which are
patterned on both sides of RO4003 substrate (ɛr = 3.38). The antenna has an overall compact size
of 0.587λo × 0.55λo × 0.058λo, where λo is the free space wavelength of the central frequency of 5.5
GHz. The antenna was optimized and designed for high gain, impedance bandwidth and especially
low SAR in the frequency range of 4.8–6.8 GHz. The thickness of substrate was kept thicker, and
ground plane was etched on the bottom side of the substrate to decrease back radiations. The SAR
simulations were carried out in computational simulator SEMCAD-X using post-processing feature
according to IEEE standard algorithm. For SAR calculation the input power was set at 1W while the
antenna was placed with the ear of the human head, that is the distance between the head and the
antenna was 0 mm.
Results

251
The radiation characteristics and impedance of the proposed antenna were analyzed along with
SAR to find its suitability for WLAN applications. The antenna gave a smooth gain and good
radiation pattern at all the frequencies of interest, while the antenna showed -10 dB impedance
bandwidth from 5.1–6.06 covering the whole WLAN band. The S-parameter, gain and radiation
pattern (5.5 GHz) have been shown in Fig. 1. The antenna presented stable high gain and desired
radiation pattern. The boresight gain checked at 5.5GHz was 5.5 dBi. The antenna also presented
low SAR values ensuring human safety besides of good radiation characteristics. Fig. 2 shows
average SAR value for 1g (SAR1g) and 10g (SARg10) head tissue at the central frequency of 5.5
GHz. The SAR value was found to be 0.185W/Kg for 1g and 0.496W/Kg for 10g, which are under
the international safety SAR standards. The accepted SAR standards are 1.6W/kg and 2W/kg for 1g
and 10g head tissues, respectively. Conventionally, a back reflector is used in antennas to get high
gain and low back radiations, our antenna design achieved low back radiations enabling very low
SAR values without any extra back reflector due to the successful implementation of the ground
plan and diagonally truncated patch on both sides of a thick substrate.
Conclusion
The radiation characteristics, impedance, and SAR of the proposed antenna was analyzed to find its
suitability for WLAN applications.The simulation results show the antenna has -10 dB impedance
bandwidth of 960 MHz (5.1–6.06) covering the WLAN band. Besides of the desired radiation pattern
and stable gain, the proposed design presents the key advantage of considerable very low SAR
values than the international standards in the human head without using any reflector.
Figures

Figure 1. S11, gain and radiation pattern (5.5GHz) of the antenna

252
 Figure 2. 1g and the 10g averaged SAR distribution when the antenna was placed at 0 mm
distance from the human head.

S13-2 [15:45]
Analytical determination of maximal power-density averaging areas for conservative 5G
exposure assessment and computational validation
Esra Neufeld1, Eduardo Carrasco1, 4, Manuel Murbach1, Quirino Balzano1, 3, Andreas Christ1 & Niels
Kuster1, 2
1IT'IS Foundation for Research on Information Technologies in Society, Zürich, Switzerland, 8004
2Swiss Federal Institute of Technology (ETHZ), Zürich, Switzerland, 8092
3Department of Electrical and Computer Engineering, University of Maryland, College Park, Maryland, USA
4Information Processing and Telecommunications Center, Universidad Politecnica de Madrid, Madrid, Spain
Keywords: Standards, RF/Microwaves, Completed (published)
Presented by: Niels Kuster
Exposure thresholds for 5G frequencies are mostly formulated in terms of surface averaged incident
power density. For localized exposure, critical heating can occur when the averaging area is too
large. An analytical model of the antenna distance-, aperture-, and frequency-dependence of the
maximal averaging area was derived and validated against simulations of a wide range of single
element antennas and antenna arrays, resulting in recommendations for standardization.
Background
At 5G communication technology relevant frequencies, exposure thresholds are mostly formulated
in terms of surface averaged incident power density. The safety guidelines (e.g., [1, 2] foresee
thresholds and averaging areas that have been determined with primarily far-field exposure in mind.
However, localized near-field exposure can lead to high temperature increases which can be
hazardous. We have developed an analytical frequency-, distance-, and antenna aperture-
dependent model to determine maximal averaging areas that ensure that tissue-heating remains
below a maximal temperature increase for cases when the surface averaged incident power density
is below the standardized limit of 10W/m^2. While the model needs to be conservative it should, at
the same time, avoid unnecessary constraints on the technological potential of 5G. Validation was
253
performed against computational modeling of skin heating when exposed to a range of relevant and
representative single element antennas (long and short slot, dipole, and patch antennas) and
antenna arrays (with varying arrangement, polarization, and steering).
Methods
The model combines analytical sub-models of 1) effective energy transfer into layered skin,
2)frequency- and distance-dependent beam width, and 3) skin heating resulting from Gaussian-
shaped surface energy deposition. The energy transfer (1) is computed by analytically solving
transmission and reflections of a plane wave of normal incidence to a planar stack of skin tissue
layers (accounting for multiple reflections and standing waves). The dielectric properties of the
different tissue layers (stratum corneum, viable epidermis, dermis, hypodermis, muscle tissue) were
assigned based on the dispersive models from Gabriel et al. [3], and the layer thickness variabilities
were obtained through an extensive literature review. Based on this, frequency dependent worst-
case skin configurations were identified. The beam width model (2) distinguishes the near- and far-
field. The wavelength-normalized near-field dependence of beam width on wavelength-normalized
distance was extracted from simulations of short dipole and slot antennas and verified for a large
range of realistic 5G antenna configurations. The far-field beam width was determined by identifying
the distance- and aperture-dependent worst-case focusing of a Gaussian beam. The skin heating
(3) estimation involves the Pennes Bioheat equation Green’s function for surface heating of a semi-
infinite perfused medium. Based on the combined model, a maximal averaging area can be
determined (analytically) that limits heating to a threshold value for exposure at a given frequency
and distance.
Results
The different analytical sub-models have been validated against electromagnetic and thermal
simulations of layered skin tissue exposed to a wide range of single element antennas and antenna
arrays, demonstrating their suitability and conservativeness for frequencies where energy deposition
can be considered superficial. For frequencies above 30GHz, analytical temperature predictions
always deviate by less than 25% from the simulation results. At distances greater than 2mm from
the source, a maximal averaging area of <2cm^2 (at 10GHz) to 3cm^2 (at 100GHz) is found to limit
temperature increase to below 1K. These values are conservative but not overly so in comparison
with the results of the coupled electromagnetic-thermal modeling.
Conclusions
An analytical model of skin heating due to 5G exposure has been constructed and validated against
computational modeling. From this model, conservative thresholds on maximal averaging areas for
incident power density could be derived that limit temperature increase to a threshold value when
applying the standardized incident power density limits. The model has been applied to derive limits
for exposure at distances greater than 2mm from the source, but can also be used to determine
distance- and frequency-dependent limits.
References
[1] International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines for Limiting
Exposure to Time-Varying Electric, Magnetic and Electromagnetic Fields (up to 300 GHz),” Health
Physics, vol. 74, no. 4, pp. 1–32, Apr. 1998.
[2] “IEEE standard for safety levels with respect to human exposure to radio frequency
electromagnetic fields, 3 kHz to 300 GHz,” IEEE Std C95.1-2005 (Revision of IEEE Std
C95.1-1991), pp. 1–238, April 2006.
[3] Gabriel, S, Lau R.W, & Gabriel C 1996, 'The dielectric properties of biological tissues: III.
Parametricmodels for the dielectric spectrum of tissues.' Physics in Medicine and Biology , 41 (11),
2271.

254
S13-3 [16:00]
Levels of environmental non-ionizing electromagnetic radiation in children´s home with
Acute Leukemia in the Metropolitan Area of Guadalajara and current situation of the
regulatory framework in Mexico
Leonardo Soto-Sumuano1, Emanuel Abundis-Gutierrez1, Roberto Garibaldi-Covarrubias2, Alberto Tlacuilo-
Parra3 & Yuridia Salazar-Galvez3
1Information Systems Department, Universidad de Guadalajara, Guadalajara, Mexico, 44100
2Hematology and Oncology, Pediatric Hospital , Guadalajara, Mexico, 44100
3Research Departament, Pediatric Hospital of IMSS, Guadalajara, Mexico, 44100
Keywords: Public Health Policy, All Frequencies, Work in Progress
Presented by: Leonardo Soto-Sumuano
Exposure to electromagnetic radiation has caused an environmental imbalance both for the
preservation of health and life itself. This article describes original results obtained from the
intradomiciliary measurement of ELF and RF in homes of pediatric patients with acute leukemia in
the metropolitan area of Guadalajara, presents a discussion of the possibility that RF be a risk factor
either in the development of leukemia or in preventing its cure or worsening the patient’s condition,
analyzes international regulation and examines the impact of the lack of regulations in Mexico even
though the current legal framework recognizes the electromagnetic radiation as a pollutant and a
possible threat to human health.
Introduction
Electromagnetic Radiation EMR is classified as ionizing and non-ionizing (Fig. 1). The main
characteristic of the first one is to have enough energy to ionize the matter when placed in a higher
frequency range. For its part, the non-ionizing radiation does not have enough associated energy to
produce the ionization process, so it does not generate ions and its effects are only perceived by the
human being in the microwave range externally through its thermal effects. This does not mean that
non-ionizing radiation can´t have negative effects on the health. In 2002, the International Agency
for Research on Cancer (IARC) evaluated the risks of non-ionizing radiation in the extremely low
frequency range (ELF), classifying it as a possibly carcinogenic agent Group "2B" (IARC, 2002).
Subsequently, in 2011 the same IARC evaluated the risk of cancer associated with exposure to non-
ionizing radiation in the radiofrequency (RF) range.The set of non-ionizing EMF that can affect the
human body either by the telecommunications infrastructure and / or equipment for domestic and
industrial use or by the massive use of new applications has been called "Electromagnetic
Environmental Pollution" or "Electropollution". Mexico, despite of being a heavily industrialized
country, with nearly 100 million of cell phone lines, lacks the regulations that should protect its
citizens from this type of radiation. Only the Ministry of Work and Social Prevention regulates the
occupational exposure whereas the Environmental, Health and Telecommunications authorities
have not yet issued any regulations even though they have the constitutional mandate to do so.
Moreover, a large percentage of the telecommunications infrastructure is irregular since every
county is responsible for regulating its territory but most of them don’t have the resources to enforce
their own regulations or fight back large telecom companies when they intend to close a site.The
absence of local and federal regulations has a huge impact since we ignore the intensity of the fields
we are exposed to everyday. We may have a high expose scenario nationwide without knowing
about it, mainly because neither the local nor the Federal Government know exactly how many
towers, masts and poles are installed, let alone, as mention above, the intensity of this
exposure.The annual incidence of childhood leukemia during 2003 was estimated at 4 / 100,000
inhabitants according to the World Health Organization (WHO). While for 2012, the Global Cancer
Surveillance group (GLOBOCAN) reported a higher incidence in developing countries (147,000
cases / year) than in developed countries (Ferlay et al., 2012).A positive association between
residential exposure to ELF-MF and childhood leukemia has been observed in several
epidemiological studies in different settings at different points in time. Combining those studies in
pooled analysis showed a relative risk of 1.5 - 2 at daily average exposure levels exceeding 0.3/0.4

255
μT [Ahlbom et al., 2000; Kheifets et al., 2010], these elevated levels occur mainly in close vicinity of
high-voltage power lines or other major electrical installations, but exceptionally to indoor wiring.
Such association had been observed with a relative high consistency, in another publication
including 20 epidemiological studies [Schüz, 2011].
Recent age-standardized incidence rates of 56.6 and 53.1 per million children aged 0 - 14 years
have been reported for Germany [Kaatsch et al., 2016] and the US [Howlader et al., 2016], however
incidence rates vary by race/ethnicity, it is well recognized the higher incidence for white Hispanic
children living in US. Moreover, in a recent publication in our country the childhood leukemia
incidence rate was 64.4 per million and appearing in clusters [Tlacuilo-Parra et al., 2017]. This study
aimed to measure the exposure to ELF-RF in the house of children with acute leukemia living in the
metropolitan area of Guadalajara, Mexico.
Objective
Determine the levels of non-ionizing electromagnetic radiation in low frequency (ELF) and
radiofrequency (RF) within the homes of children with acute leukemia in the metropolitan area of
Guadalajara, analyze international regulation and examine the current regulatory framework in
Mexico.
Materials and Methods
Cross-sectional study, patients diagnosed with children's LA attended at the High Specialty Medical
Unit UMAE Hospital de Pediatría HP of the National Medical Center of the West CMNO belonging to
the Mexican Institute of Social Security (IMSS) and the New Civil Hospital of the University of
Guadalajara that live in the metropolitan area of Guadalajara, during the period 2010-2015
The patients included in the study had to meet the following criteria: being under 16 years old, living
in the metropolitan area in the same house for at least one year before diagnosis, the diagnosis with
acute leukemia had to be confirmed by marrow aspiration bone and an immunophenotype
characterization and finally having been diagnosed and treated in any of the hospitals during the
period of the study.
Of the 84 eligible cases, 4% rejected participating in the study whereas 96% joined willingly. We
identified by means of a survey the sources of non-ionizing radiation divided into extra and
intradomiciliary sources, we went to the patient's home to measure the levels of radiation in low
frequency (ELF) and radiofrequency (RF).
The measurements of ELF and RF were made inside the houses of the patients in their bedroom
where they spend 8 continuous hours, for this purpose we utilized: 1) portable equipment HI-3604
ETS-Lindgren meter to measure the magnetic field intensity in μT (microTeslas) in the range of ELF
and 2) HI-2200 portable equipment for RF measurements of electric field strength in Volts / m and
power density in μW / cm2. A probe with a bandwidth from 100 KHz to 5 GHz was used.
Descriptive statistics were used to explain data summaries, like mean and standard deviation. For
the inferential statistics we used the T test in order to compare averages between hospitals and Chi-
square to compare qualitative variables. The research protocol was approved by the IMSS ethics
and research committee and a review committee of the participating hospitals.
International Regulation and Regulatory Framework in México
In 1992, the International Commission on Non-Ionizing Radiation Protection (ICNIRP) [ICNIRP,
1998] was constituted as an independent commission to continue the work of the INIRC
(International Non-Ionizing Radiation Committee) and the IRPA (International Radiation Protection
Association). In the United States, in 1996 the Federal Communications Commission (FCC) [FCC-
OET-Bulletin, 1997] created a regulation of human exposure to electromagnetic fields, whose
maximum level of power density must not exceed 10 W / m2 (1000 μW / cm2), which has been
under revision since 2015 due to the increase of adverse health signals and the exponential growth
registered in the emissions in the last 20 years.
It should be noted that the ICNIRP has not updated its recommendations and this approach has

256
triggered actions in countries that are sensitive to the follow-up of scientific research on the health
effects of this type of radiation and as a consequence have made adjustments to the maximum level
allowed in order to lower the risk and increase caution to its citizens. Countries such as Switzerland
(limit of 5 μW / cm2), Italy (limit of 10 μW / cm2) and others have such differentiated limits with
respect to the United States (limit of 1000 μW / cm2), which usually confuse other countries or
organizations at the time of issuing a report [SwissOrdNIR, 1999].
When analyzing the legal framework prevailing in México, there is a perception that although there
is a legal basis for creating norms and regulations, the various government agencies have ignored
their constitutional mandate, exposing a deep disconnection between the legal reality and the social
reality because it is illogical and incongruous that without having norms for the non-occupational
human exposure to EMF the telecom industry has been able to develop to the degree of having
circa 100 million cell phone lines in a country of nearly 120 million inhabitants.
Currently among all the Ministries, Commissions and Agencies which have the obligation to regulate
human and environmental exposure to EMF, only the Ministry of Work and Social Prevention has
issued an Official Mexican Norm (NOM) that regulates occupational exposure.
Results
A value of p <0.05 was considered significant, the average exposure level measured for low
frequency ELF in the patients' beds was 0.25 ± 0.35 μT (median 0.09 μT), 32% with levels higher
than the limit. Additionally, there were no significant differences between the average and the
median obtained from the houses of the patients in the two hospitals 0.28 ± 0.37 μT (median 0.13
μT) vs 0.19 ± 0.28 μT (median 0.08 μT) (p = 0.251).
However, when categorized into groups for values ≥ 0.3 μT, 32% (26 cases) of the measurements
were values higher than the reference level and even more 5% of the cases obtained values higher
than 1.0 μT. For measurements in the range of radio frequencies (RF) the average exposure level
measured the results were 142.4 ± 50.3 μW/cm2 (median 0.54 μW/cm2), 35% with levels greater
than the cut (0.1 μW/cm2). It is noted that of the total measurements, 13% resulted with high levels
in both frequencies.
Discussion
As described, the scenario of non-ionizing electromagnetic radiation in Mexico is worrisome. Firstly,
the lack of an urban regulation is prevalent. Homes less than 2 meters away from sources of EMR
(high voltage and antennas of mobile phones) have been found, even more, there is
telecommunications infrastructure above the houses. Secondly, the lack of official standards that
protect the citizens and safely control emissions as well as the growth in the demand for services
that can significantly affect the health of the child population by receiving excessive doses of
radiation. Disturbingly, the significant increase in cases of acute leukemia in children in the pediatric
hospitals of the country. The high value of radiation levels measured of the inside of the houses of
the children in the study invites us to reflect on the possible existence of association with the
leukemia.
Conclusions
There is a poor involvement of the state and municipal authorities, in the knowledge of the problems
surrounding EMR and its impact, of current regulations and their application.
One third of children with LA of the ZMG are exposed to high levels of low frequency radiation (ELF)
and radiofrequency (RF) maybe it´s normal Low frequency (LF) electromagnetic fields (EMFs) are
abundantly present in modern society and in the last 20 years the interest about the possible effect
of extremely low frequency (ELF) EMFs on human health has increased progressively.
Nevertheless, the possible association between EMFs and an increased incidence of childhood
leukemia, brain tumors or neurodegenerative diseases was not fully elucidated. The previous
reflections expressed here aim to raise awareness about the importance of the development of
telecommunications, as long as it is sustained without affecting the environment and, consequently,

257
human health. As a general conclusion, the subject consists precisely in finding the formula to
reconcile the growth of telecommunications without affecting human health.
Figures

Figure 1. SERVICES INFRAESTRUCTURE, RANGES OF FREQUENCY, WAVELENGHT

AND RADIATION IN ELECTROMAGNETIC SPECTRUM

258
 Figure 2. Table 1 SUMMARY OF THE CURRENT REGULATORY FRAMEWORK IN MEXICO
REGARDING EMISION AND HUMAN EXPOSURE TO NON- IONIZING EMF

S13-4 [16:15]
Research programme "Radiation Protection in the Process of Power Grid Expansion"
Blanka Pophof1, Cornelia Egblomassé-Roidl1, Dirk Geschwentner1, Jens Kuhne1, Christiane Pölzl-Viol1,
Janine-Alison Schmidt1 & Gunde Ziegelberger1
1Radiation and Health, Federal Office for Radiation Protection, Oberschleißheim/Neuherberg, Germany,
D-85764
Keywords: Public Health Policy, ELF/LF, Concept
Presented by: Blanka Pophof
In Germany, the existing electric power transmission and distribution grids will be expanded. In order
to improve risk assessment and to reduce scientific uncertainties, the Federal Office for Radiation
Protection (BfS) will conduct an accompanying research programme "Radiation Protection in the
Process of Power Grid Expansion". Eight research topic areas will be investigated. The research
programme started in 2017.
Introduction
In order to increase the proportion of renewable energy in power supply in Germany, the German
electric power transmission and distribution grids will be expanded. New high-voltage alternating

259
current (HVAC) power lines and high-voltage direct current (HVDC) power lines will be constructed,
the latter for the first time in Germany. The attitude of a part of the population affected by the
expansion ranges from being critical to disapproving the planned measures. Among the reasons put
forward are concerns about health risks.
The limit values set in the Federal Immission Control Act provide protection against all established
health risks from static and low-frequency electric and magnetic fields generated by electricity lines.
However, there are uncertainties with regard to potential health effects below existing limit values
and other open questions that have to be addressed by further research.
In order to reduce scientific uncertainties, the Federal Office for Radiation Protection (BfS) initiated
an accompanying research programme "Radiation Protection in the Process of Power Grid
Expansion", including more than 35 research projects in eight topic areas. The research programme
started with a public announcement in July 2017 in Berlin. After that, an online consultation was
conducted in order to involve international and national experts, other stakeholders as well as the
general public in designing the research programme. For about two months it was possible to
evaluate the relevance of the research programme as a whole or the thematic fields separately and
to provide suggestions for setting priorities or for the conception of the research programme. A total
of 240 ratings and 163 comments were received. The majority of the evaluations were positive.
The eight topic areas are described in the following:
1. Neurodegenerative diseases
Epidemiological studies found weak associations between exposure to magnetic fields and
amyotrophic lateral sclerosis or Alzheimer's dementia in occupationally highly exposed groups of
individuals (Vergara et al. 2013). For Alzheimer's dementia, one study revealed an additional
indication of an association with residential proximity and duration of residence near to high-voltage
power lines (Huss et al. 2009). Experimental animal studies have not yet been able to confirm the
epidemiological findings; no mechanism of action is known either (Liebl et al. 2015).
One of the first research activities within this topic was an international workshop on
neurodegenerative diseases held in December 2017 in Munich. The workshop summarized the
recent state of knowledge of several aspects of these diseases and identified research gaps. As a
further step, existing epidemiological data will be analyzed by means of a meta-analyses and pooled
analyses. Suitable animal and/or in vitro models will be used to investigate possible action
mechanisms.
2. Perception and effect thresholds
Low-frequency magnetic fields induce electric currents in the body which stimulate nerves and
muscles when limit values are exceeded. Static electric fields from overhead power lines can be
directly perceived at high field strengths. Moreover, metal objects located underneath power lines
can become electrically charged. Spark discharges or contact currents may occur if an individual
comes into contact with the metallic object.
An international workshop on perception thresholds was held in October 2016 in Munich to
summarize the current state of knowledge and to identify knowledge gaps (BfS 2018). Further
projects will focus on more precise definition of perception thresholds with respect to age and
gender dependency and the influence of external factors.
3. Childhood leukemia
Low-frequency magnetic fields are classified as "possibly carcinogenic" by the International Agency
for Research on Cancer (IARC) since 2001. The classification is based on the consistent finding of a
positive association between weak low-frequency magnetic fields and risk of childhood leukemia
consistently observed in epidemiological studies. Since 2008, the BfS has initiated several
international workshops on the causes of childhood leukemia, partly in cooperation with the WHO,
the International Commission on Non-Ionizing Radiation Protection (ICNIRP) and other radiation
protection authorities. As a result, an interdisciplinary research agenda was published (Ziegelberger

260
et al. 2011). On the basis of this agenda, five pilot studies have been performed and the results
published. They are the basis of follow up research projects focusing on epidemiology (meta-
analyses, pooled analyses, participation in a birth cohort study), molecular and genetic background
of leukemia in humans and the influence of magnetic fields and other exposures on the imitation of
leukemia in a newly developed transgenic mouse model.
At first, a project on the immune status of transgenic mice exposed to magnetic fields has started. It
will clarify, whether the magnetic field effects on cytotoxic T-cells observed in pilot studies can be
confirmed.
4. Co-carcinogenicity of magnetic fields
Animal studies and in vitro studies on co-carcinogenicity have been classified as high priority by the
WHO (2007). A recent study (Soffritti et al. 2016) on rats with lifelong exposure to magnetic fields
and an additional exposure to a single dose of ionizing radiation revealed a significantly increased
morbidity for several types of tumors. The possible co-carcinogenic effect of magnetic fields has to
be verified.
5. Magnetic fields and miscarriage
Based on indications concerning occupational and residential exposure to magnetic fields (e.g.
using electric blankets), miscarriage as an endpoint was included in the WHO Research Agenda
(WHO 2007) although the evidence is classified as very weak. If there is access to an already
existing birth cohort with data on residential and/or occupational exposures to magnetic fields the
risk of miscarriage and ELF will be investigated.
6. Investigations of the occurrence, propagation and absorption of corona ions
Due to the corona effect observed at high-voltage power lines, air molecules and particles around
overhead lines become electrically charged. A hypothesis implies that charged molecules and
particles increase the risk of respiratory diseases in residents living near high-voltage power lines
(Henshaw et al. 1996, Fews et al. 1999). An increased health risk from air pollutants due to charged
particles around conventional HVAC lines was considered unlikely by the British radiation protection
authorities (NRPB 2004).
As it is assumed that the charge clouds emitted by HVDC lines exhibit increased stability over time
and are therefore expected to expand further than those generated by HVAC lines, the hypothesis is
gaining new significance. The question to what extent air pollutants emanating from the subsoil
(such as radon and its decay products), industrial processes or traffic are increasingly absorbed by
the body depending on their electric charge state and the implication on health risks, has yet to be
clarified.
As a starting point, a literature review will be performed. The state of scientific knowledge on the
formation, concentration and propagation of ionized air molecules and dust particles will be
determined. Conventional HVAC transmission lines and the planned HVDC transmission lines will
be compared also in view of weather conditions. Potential health risks will be documented,
knowledge gaps identified and research approaches as well as appropriate investigation methods
will be recommended.
7. Exposure analysis, exposure assessment and updated data on the exposure of the general
public
The electric fields and currents induced in the human body during exposure to low-frequency fields
can partially be determined only with large uncertainties. Thus, the independent Scientific
Committee on Emerging and Newly Identified Health Risks (SCENIHR) of the European Union
points out, for example, that data and systematic studies on the dielectric properties of tissues at low
frequencies are lacking. In its Research Agenda for low-frequency fields, the WHO (2007)
recommends improving dosimetric models for tissue and tissue structures which are particularly
sensitive to induced electric fields. Even technical professional associations such as IEEE/ICES
have identified research needs in this context (Reilly & Hirata 2016). To this end, the BfS will

261
contribute to overcome these knowledge gaps.
For precautionary reasons the public exposure should not increase significantly when constructing
and substantially modifying electrical power supply facilities. Investigations of low-frequency
magnetic fields in Germany conducted mainly before 2000, revealed time-averaged exposures of
about 0.1 µT. These data need to be updated. The exposure of the general public will be measured
using personal dosimeters. Measurements in the vicinity of power lines and other appliances are
also planned.
Electrical power supply facilities have to be operated in a way that considerable nuisance due to
contact currents and spark discharges is prevented. The occurrence of these effects will be
investigated more closely.
8. Risk perception and risk communications
In this part of the research programme, the citizens' knowledge and risk perception and the public's
need for information will be determined. On that basis, it will be clarified how information has to be
prepared and how and by whom it should be conveyed in order to reach the widest possible public.
The programme also investigates which factors are decisive for public opinion formation and how
the credibility of the involved authorities and the confidence in the individuals acting on their behalf
can be ensured or enhanced. Representative and regular surveys to determine public concern will
start soon and provide a suitable basis for reacting to public concerns and needs with target-group-
specific information and communication measures.
The projects within the scope of risk communication are intended to counterbalance the citizens'
uncertainty about the health effects of power lines. The aim is to improve the citizens’ ability to form
their own well-founded opinions using the provided information.
In order to discuss the current knowledge from socio-scientific research among interested and
professional members of the public, a workshop on various aspects of communication in the
process of power grid expansion was held in November 2017 in Munich. The main topics were
discourse approach, credibility, confidence, transparency and lessons learned from practical
examples of power grid expansion so far.
References
BfS 2018, Workshop Abstracts, urn:nbn:de:0221-2018011814498
Fews AP et al. 1999. Int J Radiat Biol 75: 1505-1521.
Henshaw DL et al. 1996 Int J Radiat Biol 69: 25-38.
Huss A et al. 2009 Am J Epidemiol 169: 167-175.
Liebl MP et al. 2015. Sci Rep. 5:8585.
Reilly JP & Hirata A 2016 Phys. Med. Biol. 61:R138
Soffritti M 2016 Int J Radiat Biol 92: 202-214.
Vergara X et al. 2013 J Occup Environ Med 55: 135-146.
WHO 2007 Research Agenda for Extremely Low Frequency Fields
Ziegelberger G et al. 2011 Blood Cancer Journal 107: 312 - 314.

262
 Session: S14
Dosimetry III - Measurements
Thursday June 28, 2018 • 16:30 - 18:00
Europa C
Chairs: Luc Martens & Blaž Valič

S14-1 [16:30]
STUDENT PAPER
RF exposure in the vicinity of small cell base stations
Thomas Kopacz1, Sascha Schiessl1, Christian Bornkessel2, Matthias Hein2 & Dirk Heberling1
1Institute
of High Frequency Technology, RWTH Aachen University, Aachen, Germany
2RF & Microwave Research Laboratory, Thuringian Center of Innovation in Mobility, Technische Universität
Ilmenau, Ilmenau, Germany
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Thomas Kopacz
In the course of the evolution of mobile radio networks and the increasing high-capacity needs,
operators have started to deploy small-cell base stations. To account for possible concerns about
the RF exposure, we investigate the exposure for two types of small-cell base stations embedded
into street furniture: 1. A phone box (PB) and 2. a service-area interface (SAI). We have found that
in public areas the highest exposure amounts to 17.9 % of the electrical field reference level (%E)
for the PB and 53.0 %E for the SAI. Downlink exposure and uplink transmit power have been
assessed for typical and worst-case user scenarios in the vicinity of the SAI proving that higher
downlink field strength leads to lower uplink transmit power values.
Introduction
As a transition towards the fifth generation of mobile communication standard, the growing data
traffic especially at hot spots with high user densities has led to the deployment of small cells. Small-
cell base stations are often embedded in street furniture such as phone boxes or lamp posts and are
located closer to the general public compared to conventional macro-cell base stations. These
circumstances give rise to higher downlink exposure despite their lower transmit power. Still,
compliance with RF exposure limits has to be ensured. For fixed transmitters with an equivalent
isotropic radiated power (EIRP) of less than 40 dBm, the reference levels for RF exposure in
Germany are specified in DIN EN 50392 [1] and in line with the ICNIRP exposure guidelines [2]. The
public area is defined for height levels below 2.30 m above ground.
In this paper, the results of a RF exposure compliance assessment in the vicinity of two types of
small-cell base stations embedded into street furniture are presented:
Phone box (PB)
Service-area interface (SAI)
Investigated base stations, measurement equipment and protocol
Small-cell base stations
Both types of small-cell base stations were part of LTE networks in regular operation in the 2.6 GHz
band with a channel bandwidth of 20 MHz. The phone box investigated here was located in
Cologne, Germany, and was equipped with two vertically polarized MIMO antennas at a height of
about 2.50 m above ground. In total, the EIRP was 35.7 dBm. The service-area interface measured
in Saarbrücken, Germany, was equipped with two antennas at a height of approximately 1.50 m
above ground. In total, the EIRP was 39.1 dBm.
Exposure assessment
The downlink field strength was measured using the selective radiation meter SRM‑3006 from
Narda. The sum field strength of the Reference Signal (RS) was assessed by code-selective

263
measurements. Extrapolation to full utilization of the base stations was carried out afterwards by
evaluating the ratio of the total output power to the averaged output power for one RS. For the
small-cell base stations this ratio was determined to be 600 (27.8 dB).
For the assessment of uplink exposure, a Samsung Galaxy S5 smartphone in combination with the
application software “QualiPoc” from Rohde & Schwarz [3] was used. This setup provides
information about the uplink transmit power of the UE. The values recorded for transmit power were
averaged over one minute while staying stationary, according to the procedure described in [4].
Measurement points
For the phone box, the exposure was determined at stationary measurement points along two
horizontal lines from the left side and the back side of the box; the scenario is illustrated in the left-
hand part of Fig. 1. Due to partially restricted access around the PB, the lines covered distances
between 0 m and 2.00 m for the back side and between 0 m and 5.00 m for the left side. Note that
0m distance implies that the probe radome touched the base station enclosure. The linear
arrangements of measurement points were investigated at two different heights, namely at 2.30 m
above ground, which in Germany denotes the boundary between public accessible and restricted
area, and at 2.20 m above ground which marks the limiting height according to IEC62232 [5].
For the service-area interface (see right-hand part of Fig. 1), stationary measurements on the
surface of the enclosure were carried out. Like for the exposure assessment of the phone box, at a
distance of 0m, the radome of the probe touched the SAI enclosure. The measurements in the
vicinity of the SAI took also into account typical user scenarios; consequently, the field strength was
additionally measured at points with typical and worst-case user positions (e.g. leaning against the
SAI enclosure or waiting at a tram station nearby). The probe was swept to assess the local
maximum. Furthermore, the uplink transmit power was determined.
Results
Phone Box
Figure 2 shows the measured exposure to the phone box in percent of the electrical field strength
reference level along the two horizontal lines at 2.30 m and 2.20 m height in dependence of the
distance from the base station. The maximum exposure reached a value of 17.9 %E and was found
on the left side at 2.30 m height and zero-distance. On the back side, the maximum exposure
amounted to 16.9 %E at 2.30 m height and zero-distance. At 2.20 m height, the peak exposure was
markedly lower, reaching 10.9 %E at the back and 8.4 %E at the left side. For larger distances, the
exposure decreased to around 4 %E at distances of 2.50 m and 2.5 %E at 5.00 m.
Service-Area Interface
Figure 3a shows as an example the exposure to the service-area interface measured along a
horizontal and a vertical line along the enclosure. Considering the green-colored line at z = 0 m,
where the antennas are located, the antennas appear not to be arranged centrally with respect to
the x‑axis, but shifted to the right. Accordingly, the maximum exposure of 53 %E was found at x =
0.70 m. Another local maximum was found at x = 0.10 m (43 %E). On the left- and right-hand
boundaries of the front side, the exposure decreased to 7 %E, respectively 23 %E. Starting from the
position of the maximum (at x = 0.70 m), the exposure was also assessed along a vertical line. At a
height of z = –0.20 m, the exposure was already reduced to 12 %E. The measurements revealed
only a slight decrease until the ground was reached at z = –1.40 m.
In Figure 3b, the exposure resulting for typical user scenarios is shown for both the extrapolated
downlink in terms of electrical field strength and uplink in terms of transmit power. The distances of
the measurement points (MP) range from MP1 representing a user leaning against the SAI
enclosure up to MP5 representing a person waiting at a tram station in a distance of 90 m. In
accordance with expectation, the highest exposure to the SAI was found at MP1 with a field strength
of 144 dBµV/m (25.8 %E). At MP2, in a distance of 1.00 m to the SAI, the downlink exposure
already decreased to 130 dBµV/m (5.3 %E). For the further measurement points, the downlink
exposure dropped to 112 dBµV/m (0.6 %E) at MP5. The uplink transmit power of the mobile phone

264
displayed the minimal value of -44 dBm at MP1. According to the decreasing downlink exposure for
the further measurement points, the UE transmit power rose to +4 dBm.
Discussion
Compliance assessment of RF exposure
For all measurement points, compliance with ICNIRP reference levels could be shown. For the
phone box, measurements at a height of 2.30 m yielded higher exposure than at 2.20 m.
Additionally, for the lower height, the peak exposure was found at distances between 0.10 m and
0.30 m. These observations are suspected to reflect the radiation characteristics of the PB
antennas, leading a strong variation with the elevation angle. This is also the reason why the
decrease of the field strength is not exactly inversely proportional to the distance of the phone box.
For the service-area interface, the highest exposure was measured in the top region of the
enclosure, where the antennas are located. Additionally, the peak exposure was not found in the
middle of the SAI but shifted to the right, pointing to a corresponding location of the base station RF
unit including the antennas within the SAI.
The measurements for the phone box and the service-area interface took place partially in the near-
field of the antennas, where the electrical and the magnetic field strengths are not simply related by
the free-space wave impedance Z0. Furthermore, especially for the measurements at zero-distance,
coupling effects between conducting materials of the small-cell base stations and the measurement
probe cannot be ruled out.
Exposure for typical and worst-case user positions
For typical user positions, the exposure to the SAI was markedly lower than the peak exposure
values found during compliance assessment, which reflects the larger distances between the
measurement points and the two types of base stations.
According to the transmit power control in LTE [6], a high downlink field strength leads to a low UE
uplink transmit power. For the same measurement points, the uplink transmit power ranged between
-22 dBm and -15 dBm for the PB, and between -44 dBm and 4 dBm for the SAI. Compared to
transmit power levels typical for macro-cell networks (-2.3 dBm on average, as investigated in [7])
and to the maximum possible transmit power of LTE UEs of 23 dBm [8], the investigated small cell
revealed very low uplink transmit powers. The recorded transmit power as well as the relation to the
downlink field strength agreed very well with the results investigated in another small-cell
environment [4].
Conclusions
For the assessment of RF exposure compliance with ICNIRP reference levels in proximity of two
small-cell base stations types, the following conclusions can be drawn:
Phone box: The maximum exposure was 17.9 % of the electrical field strength reference level
(%E) at 2.30 m height and at a distance of 0 m to the base station enclosure. At 2.20 m height,
the exposure was lower with a maximum of 10.9 %E. The exposure levels differed for the two
PB sides investigated.
Service area interface: The highest exposure of 53.0 %E was found in the upper area of the SAI
with contact to the enclosure, where the RF unit is located. Approaching to the ground, the
exposure decreased steeply.
The assessment of extrapolated downlink exposure in typical and worst-case user scenarios and
the corresponding uplink transmit power can be summarized as follows:
The downlink exposure ranged between 0.6 %E and 25.8 %E for the SAI and was clearly lower
than the maximum exposure.
Due to the high downlink coverage, the corresponding uplink transmit power remained low,
ranging between -44 dBm and +4 dBm for the SAI, confirming the results from another small-cell
environment investigated in [4].

265
Acknowledgment
The funding of the study by Deutsche Telekom Technik GmbH is greatly acknowledged.
References
[1] DIN EN 50392:2004-08, Generic standard to demonstrate the compliance of electronic and
electrical apparatus with the basic restrictions related to human exposure to electromagnetic fields
(0Hz- 300GHz); German version EN50392:2004, 2004.
[2] ICNIRP Guidelines, Guidelines for Limiting Exposure to Time-Varying Electric, Magnetic and
Electromagnetic Fields (up to 300 GHZ), Health Physics, vol. 74 no. 4, p. 494-522, 1998.
[3] SwissQual QualiPoc Android, Smartphone-based product for optimizing mobile networks,
https://www.rohde-schwarz.com/de/produkt/qualipoc_android-produkt-
startseite_63493-55430.html?rusprivacypolicy=0, Rohde & Schwarz GmbH & Co. KG, last accessed
2018-02-28.
[4] T. Kopacz, C. Bornkessel, M. Hein and D. Heberling, Investigation of LTE user equipment
transmit power control and comparison of uplink exposure between small and macro cell
environment, BioEM 2017, Hangzhou, China, 2017.
[5] IEC 62232, Determination of RF field strength, power density and SAR in the vicinity of
radiocommunication base stations for the purpose of evaluating human exposure, International
Electrotechnical Commission (IEC), 2017.
[6] ETSI TS 136 213, LTE; Evolved Universal Terrestrial Radio Access (E-UTRA); Physical layer
procedures (3GPP TS 36.213 version 13.0.0 Release 13), V13.0.0, 2016.
[7] P. Joshi, D. Colombi, B. Thors, L. E. Larsson and C. Törnevik, Output Power Levels of 4G User
Equipment and Implications on Realistic RF EMF Exposure Assessments, IEEE Access, 2017.
[8] ETSI TS 136 101, LTE; Evolved Universal Terrestrial Radio Access (E-UTRA); User Equipment
(UE) radio transmission and reception (3GPP TS 36.101 version 13.3.0 Release 13), V13.3.0, 2016.
Figures

266
Figure 1. Investigated small cell base stations embedded into street furniture: a phone box
(left-hand photograph) and a service-area interface (right-hand photograph), both in regular
operation.

Figure 2. Exposure to the PB in percent of the electrical field strength reference level versus
distance to the small-cell base station along two measurement lines at heights of 2.30 m
(grey- and yellow-colored symbols) and 2.20 m (blue- and orange-colored symbols).

267
Figure 3. a) Exposure to the SAI in percent of the electrical field strength reference level in %E
for the green-colored measurement points (horizontal, along the x axis) and the blue-colored
measurement points (vertical, along the z axis) in front of the RF unit. b) Extrapolated
exposure to SAI and a UE logged onto the SAI small-cell for five measurement points in the
vicinity of the SAI. MP1 represents a user leaning against the SAI; MP5 represents a person
waiting at a tram station nearby in a distance of 90 m.

268
S14-2 [16:45]
STUDENT PAPER
Human exposure assessment in indoor environments using a 60 GHz personal exposure
meter
Reza Aminzadeh1, Abdou Khadir Fall2, Jerome Sol2, Arno Thielens1, 3, Philippe Besnier2, Maxim Zhadobov4,
Nele De Geeter5, Prakash Parappurath Vasudevan 1, 6, Luc Dupré5, Roel Van Holen6, Luc Martens1 & Wout
Joseph1
1Dept. of Information Technology (INTEC), Ghent University/imec, Ghent, Belgium
2Institute of Electronics and Telecommunications of Rennes, INSA of Rennes, Rennes, France
3Department of Electrical Engineering and Computer Science, University of California, Berkeley, USA
4Institute of Electronics and Telecommunications of Rennes, University of Rennes 1, Rennes, France
5Dept. of Electrical Energy, Systems and Automation (EESA), Ghent University, Ghent, Belgium
6Dept. of Electronics and Information Systems (Elis), Ghent University/imec, Ghent, Belgium
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Reza Aminzadeh
This paper presents the first mm-wave personal exposure meter (mm-PEM) to assess human
exposure to the 5th generation of mobile networks (5G) in indoor environments. The mm-PEM
consists of 9 elements of an antenna array and is calibrated on a skin-equivalent phantom in a
reverberation chamber at 60 GHz. The designed mm-PEM has a response of 1.043 (0.17 dB) at 60
GHz which is very close to the desired response of a PEM i.e. 1 (0 dB). The mm-PEM measured an
incident power density of 41 mW.m‑2 at 60 GHz for an input power of 1 mW in the empty chamber.
Introduction
The rapid progress in 60-GHz wireless technologies and the availability of the 5th generation of
mobile networks (5G) in the near future [1] has raised concerns regarding the potential adverse
health effects of mm-waves on human body. The absorption of mm-waves is limited to skin tissues
[1]. Therefore, the incident power density (IPD) is studied as a dosimetric quantity. The safety limits
of IPD are 1 mW.cm-2 and 5 mW.cm-2 averaged over 20 cm2 of the exposed area for general public
and occupational exposure, respectively [2]. Human exposure to radiofrequency (RF)
electromagnetic fields is usually measured by Personal Exposimeters (PEMs) [3, 4]. These are
portable devices worn on body allowing for continuous measurement of the electric fields strength in
several frequency bands for which protocols have been developed [5]. PEMs are calibrated in free
space while used on body. In other words, the measured values are compromised by the presence
of the human body and thus have large measurement uncertainties [6]. In order to reduce this
measurement uncertainty, personal distributed exposimeters (PDE) with multiple antennas can be
used for single [7] and multi telecommunication bands [8].
Research shows that people spend more than 80% of their times indoors [9]. This could increase
human exposure the electromagnetic fields. The total power in an indoor environment consists of
specular and diffuse multipath components. The former and the latter are due to the reflections from
large surfaces and presence of objects in a room, respectively. The DMC can contribute up to 95%
to the total power density in an indoor environment [10].
The novelties of this research are as follows: 1) For the first time, design and calibration of a mm-
wave personal exposure meter (mm-PEM) for RF exposure assessment in indoor environments. 2)
on-phantom calibration of the mm-PEM in a reverberation chamber (RC) at 60 GHz. 3) To determine
the measurement uncertainty of the proposed mm-PEM including 9 antenna elements.
Materials and methods
The goal is to determine the measurement uncertainty of the designed mm-PEM in diffuse indoor
environments. The mm-PEM consisting of 9 antennas is calibrated in an RC at 60 GHz with a
269
bandwidth of 1 GHz. The experimental setup is shown in Figure 1.The dimensions of the RC are
0.58×0.592×0.595 m3. The antenna used for this study is a microstrip four-patch antenna array (Rx)
[11]. The surface of the phantom is divided into 9 symmetric locations (i) with a distance of 4 cm.
First, the Rx is placed at every location i and the scattering (S-) parameters are measured for every
position of the Rx on the phantom and for 100 positions of the stirrer. Second, the measurements
are repeated for the Rx far from the phantom at the same locations i. This is to emulate free space
conditions for the RX when the RC is loaded with the phantom. The IPD (Si) for the Rx at every
location i is obtained as:

Where Qi is the quality factor of the RC for the Rx at location i, Pt is the transmitted power, is the
wavelength and V is the volume of the RC. The Q-factor of the RC is calculated from the
S-parameters [12].The response R of the mm-PEM is determined as the ratio of the received power
on the Rx placed on the phantom and the Rx in the loaded RC (Rx far from the phantom):

Results
Figure 2 illustrates the measured response of the mm-PEM averaged over 100 positions of the
stirrer for 9 individual antennas on the phantom and the spatial average over 9 locations (antennas)
on the phantom. Using a single antenna results in either underestimation (0.6 or -2.21 dB) or over
estimation (1.4 or 1.46 dB) of the incident fields inside the RC. Using the average over 9 antennas
the mm-PEM has an R of 1.043 (0.18 dB) at 60 GHz. This is an excellent agreement with the ideal
value of R that is 1 (0 dB) for a PEM on body. Using less than 3 antennas leads to an overestimation
of the IPD up to a factor of 1.25 to 1.4. Increasing the number of antennas up to 9 decreases the
response to the values close to 1 as discussed above. Using 9 antennas the mm-PEM has an R of
0.85 (-0.66 dB) to 1.13 (0.55 dB) in the range of 59.5 to 60.5 GHz.
Figure 3 shows the measured IPD for the antenna in the empty RC and on the skin-phantom in the
range of 59.5 to 60.5 GHz. The measured IPD at 60 GHz for an input power of 1 mW are 40.9
mW.m-2 and 12 mW.m-2 for the antenna in the empty RC and on the phantom, respectively. The
measured IPD in the empty RC is about 3.4 times higher than the measured IPD for the antenna on
the phantom.
Conclusions
A mm-PEM is designed and calibrated in a reverberation chamber in the range of 59.5 to 60.5 GHz.
The mm-PEM will be used to measure personal exposure to 5G and wireless communication
systems operating in the mm-wave band in indoor environments. We showed that using 9 antennas
on a skin-equivalent phantom provides a response of 1.043 (0.17 dB) at 60 GHz which is very close
to the desired response of a PEM i.e. 1 (0 dB). This means that the mm-PEM can measure the
incident power densities in free space but in the presence of human body.
Acknowledgment
This research was funded by the Research Foundation-Flanders (FWO-V) under grant agreement
No G003415N. A. Thielens is a post-doctoral Fellow of Flanders Innovation and Entrepreneurship
under grant No. 150752. A.T. has received funding from the European Union’s Horizon 2020
research and innovation programme under the Marie Skłodowska-Curie grant agreement No
665501 with the research Foundation Flanders (FWO). A.T. is an FWO [PEGASUS]² Marie
Skłodowska-Curie Fellow.
References

270
[1] Wu T, Rappaport T, Collins C. 2015. Safe for generations to come: Considerations of safety for
millimeter waves in wireless communications. Microwave Magazine IEEE, 16(2): 65-84.
[2] International Commission on Non-Ionizing Radiation Protection. 1998. Guidelines for limiting
exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz). Health
Physics 74: 494-522.
[3] Bolte J.F.B., Eikelboom T. 2012. Personal radiofrequency electromagnetic field measurements in
the Netherlands: Exposure level and variability for everyday activities, times of day and types of
area. Environ Int 48:133-142.
[4] Thielens A., Agneessens S., Verloock L., Tanghe E., Rogier H., Martens L., Joseph W. 2015a.
On-body calibration and processing for a combination of two radio-frequency personal
exposimeters. Radiat Prot Dosim 163:58-69.
[5] Roosli M., Frei P., Bolte J., Neubauer G., Cardis E., Feychting M., Gasjek P., Heinrich S., Joseph
W., Mann S., Martens L., Mohler E., Parslow R.C., Poulsen A.H., Radon K., Schüz J., Thuroczy G.,
Viel J., Vrijheid M. 2010. Conduct of a personal radiofrequency electromagnetic field measurement
study: Proposed study protocol. Environ Health 9:23.
[6] Bolte J.F.B. 2016. Lessons learnt on biases and uncertainties in personal exposure
measurement surveys of radiofrequency electromagnetic fields with exposimeters. Environment
International 94:724-735.
[7] Thielens A., De Clercq H., Agneessens S., Lecoutere, J., Verloock, L., Declercq, F., Vermeeren,
G., Tanghe, E., Rogier, H., Puers, R., Martens, L., and Joseph, W. 2015. "On-body calibration and
measurements using a personal, distributed exposimeter for wireless fidelity." Health Phys 108(4):
407-418.
[8] Aminzadeh R., Thielens A., Agneessens S., Van Torre P., Van den Bossche M., Dongus S.,
Eeftens M., Huss A., Vermeulen R., De Seze R., Mazet P., Cardis E., Rogier H., Röösli M., Martens
L., Joseph W. 2018. A multi-band body-worn distributed radio-frequency exposure meter: Design,
on-body calibration and study of body morphology. Sensors 18:272.
[9] C. Aliaga. How is the time of women and men distributed in Europe. statistics in focus, population
and social conditions. issn 1024-4352, ks-nk-06-004-en-n, eurostat Luxembourg (2006).
http://www.unece.org/fileadmin/DAM/stats/gender/timeuse/DataReports
/How_is_the_time_of_Women_and_Men.pdf. Accessed on (Mar. 07, 2018).
[10] Poutanen J., Salmi J., Haneda K., Kolmonen V.M., Vainikainen P. 2011. Angular and shadowing
characteristics of dense multipath components in indoor radio channels. IEEE Transactions on
Antennas and Propagation 59: 245-253.
[11] Chahat N., Zhadobov M., Le Coq L., Alekseev S., Sauleau R. 2012. Characterization of the
interactions between a 60-GHz antenna and the human body in an off-body scenario. Antennas and
Propagation, IEEE Transactions on 60 (12): 5958–5965.
[12] Besnier P., Lemoine C., Sol J. 2015. Various estimations of composite Q-factor with antennas in
a reverberation chamber. IEEE International Symposium on Electromagnetic Compatibility (EMC):
1223-1227.

Figures

271
Figure 1. The proposed experimental setup inside the reverberation chamber for on-body
calibration of the mm-PEM.

Figure 2. The response of the mm-PEM for 9 single antennas on symmetric locations on the
phantom and for the average over 9 antennas.

272
 Figure 3. The measured incident power density in the empty RC versus for the antenna on the
phantom.

S14-3 [17:00]
Microenvironmental personal and head exposure measurements of radio-frequency
electromagnetic fields in Melbourne, Australia
Arno Thielens1, 2, Matthias Van Den Bossche2, Christopher Brzozek3, Chhavi Bhatt3, Michael Abramson3,
Geza Benke3, Luc Martens2 & Wout Joseph2
1Berkeley Wireless Research Center, UC Berkeley, Berkeley, CA, USA, 94707
2Department of Information Technology, Ghent University/IMEC, Ghent, Belgium, 9052
3PRESEE, Monash University, Melbourne, Australia, 3181
Keywords: Epidemiology, RF/Microwaves, Completed (published)
Presented by: Wout Joseph
The aim of this study was to simultaneously measure personal radio frequency-electromagnetic
fields (RF-EMFs) exposure using two measurement devices. A body-worn personal exposimeter
and a head-worn personal distributed exposimeter were used for measuring body and head
exposures, respectively, in 15 microenvironments in Melbourne. The summary statistics obtained for
total RF-EMF exposure showed a high representativeness (r² > 0.87 for two paths in the same
area). The results obtained during simultaneous measurements using the two devices showed high
correlations: r² = 0.94 for the median along the measured paths.
Introduction
A considerable proportion of the human population is exposed to Radiofrequency-electromagnetic

273
fields (RF-EMFs), which are used for wireless telecommunication worldwide. This personal
exposure to RF-EMFs can be measured using so-called RF personal exposimeters (PEMs) (Bolte et
al. 2016). There exist different methods of performing methods using PEMs (Röösli et al., 2010). A
microenvironmental exposure assessment study is one of the potential approaches to measure
personal RF-EMF exposure in a systematic manner and methods for such studies have been
developed and tested in previous studies (Sagar et al., 2016; Bhatt et al. 2016).
Recently, greater research attention has focused on potential cognitive effects of RF-EMF exposure
(Roser et al., 2016), which are assumed to relate to RF-EMF exposures to the head. In order to
investigate such effects, there is a need to quantify head-specific personal exposure levels. The
current and previous generations of PEMs are unable provide head-specific exposure
measurements. In a novel approach to measure head exposure to RF-EMF, it has been proposed to
integrate and distribute a set of RF-EMF measurement nodes into headgear. Specifically, in this
study we have integrated four RF-EMF nodes that measure personal exposure in the 900 DL band
into a bicycle helmet (the personal distributed exposimeter or PDE-Helmet), so that it can measure
head-specific RF-EMF exposure in the 900 DL band.
In this study, we have simultaneously measured personal exposure to RF-EMFs of the whole body
(with body-worn PEMs) and the head (Helmet) across various microenvironments in Melbourne,
Australia.
Materials and Methods
Studied Micro-environments and Study Design
Fifteen microenvironments were defined in greater Melbourne. The studied microenvironments (see
Thielens et al., 2017) have been selected to cover different (sub-)urban activities in the area studied.
We considered the following areas (with complementary modes of transport): six residential or sub-
urban areas (three walking, two by car, and one on a bicycle), one industrial area (car), three areas
dedicated to trade, commerce, business, and tourism in Melbourne’s Central Business District
(CBD) (walking), three recreational areas (two parks on bicycle and one beach on foot), and two
college/university areas (on foot). The selected microenvironments had population densities from
250 inhabitants/km² (suburban 4-6) up to 15,000 inhabitants/km² (CBD 1-3).
The measurement study was conducted between 15th November and 22nd December 2016. In each
microenvironment, two paths were predefined and followed using one of three different modes of
transportation: walking (10), driving a car (2), and riding a bicycle (3). The paths were defined in
such a way that it took at least 15 minutes to follow them using the predefined mode of
transportation. The measurements were performed during three different timeslots during weekdays:
morning (9am – 12noon), midday (12noon – 3pm), and afternoon (3pm – 6pm). Each of the 30
studied paths was repeated five times, once on each day of the week, and was executed twice in
two different timeslots and once in the remaining third timeslot, resulting in 15 × 2 × 5 = 150
measurements.
Three microenvironments (six paths) were considered to be suitable to measure using a bicycle: two
parks and one suburban area, where every path was repeated four times (24 measurements). The
researchers’ wireless devices such as mobile phones were in flight mode during all measurements,
eliminating any contribution to RF-EMF exposure.
Measurement Devices
Two measurement devices were used in this study: one or two ExpoM-RF devices worn on one or
both hips and the PDE-Helmet (see Fig. 1), which was worn on the head during bicycle
measurements only. The ExpoM-RF is a PEM (http://www.fieldsatwork.ch), which measures electric
field strengths in 16 different frequency bands every 4 s. The PDE-Helmet is a measurement device
developed at Ghent University (Thielens et al., 2016, 2017) that uses four RF nodes distributed over
the head, in order to reduce the measurement uncertainty on the incident RF power density. The
device measures incident electric fields on the head of a subject with a frequency of 1 Hz in the 900
DL band (920-960 MHz). The detection range of the PDE-Helmet (with on-body calibration) was

274
11.3 mV/m – 113 V/m.
Data Processing
Electric field strength values registered by the devices were processed using Matlab (Mathworks,
MA, USA). The values falling below the lower limit of detection (LOD) were replaced by LOD/√2.
There were no data higher than the upper limit of detection. All electric field (E) values were
converted to power density values in each frequency band. During measurements where two
ExpoM-RF devices were used simultaneously by the same researcher, geometric means of the
power density obtained with the two ExpoM-RFs were considered. The power densities were
processed to determine the total RF-EMF exposure, which was defined as the sum of all measured
power densities in each frequency band at each time instance. The measured power densities in the
900 DL band using both devices were processed aswell. For both quantities, we calculated the 16th,
50th, 84th and 95th percentiles, and the arithmetic average along each path. We calculated these
statistics on the pooled data for the separate paths and the pooled data for all repetitions along all
paths in each microenvironment. In order to compare measurement results obtained with the two
types of measurement devices, namely the PDE-Helmet and the (pair of) ExpoM-RF(s), we
compared correlations of the summary statistics obtained with different measurement devices
obtained during simultaneous use of both devices. We compared the summary statistics obtained
from the PDE-Helmet with those obtained using (a pair of) simultaneously worn ExpoM-RF(s).
Results and Discussion
Table 1:Summary statistics of Erms using the ExpoM-RFs in 15 microenvironments in Melbourne.

Total Erms(V/m)a

μ std p16 p50 p84 p95

Microenvironment

Park 1 0.39 0.51 0.09 0.22 0.53 0.83

Park 2 0.53 0.75 0.17 0.34 0.66 1.03

CBD 1 0.89 1.07 0.36 0.64 1.17 1.66

University
0.57 0.86 0.11 0.27 0.74 1.22
Campus 1

CBD 2 0.79 0.89 0.34 0.60 1.05 1.46

CBD 3 0.72 0.88 0.31 0.53 0.93 1.36

Industrial
0.09 0.1 0.06 0.08 0.11 0.15
Area

Beach 0.21 0.25 0.09 0.14 0.26 0.42

275
 Suburban
0.36 0.62 0.04 0.09 0.39 0.83
Area 1

University
0.43 0.64 0.13 0.27 0.54 0.81
Campus 2

Suburban
0.23 0.47 0.04 0.06 0.16 0.47
Area 2

Suburban
0.25 0.41 0.05 0.12 0.28 0.52
Area 3

Suburban
0.05 0.04 0.03 0.05 0.06 0.07
Area 4

Suburban
0.06 0.06 0.04 0.06 0.08 0.10
Area 5

Suburban
0.12 0.2 0.03 0.05 0.15 0.26
Area 6

aSix quantities are listed: the mean (μ), the standard deviation on the mean (std) and four
percentiles the 16th, 50th, 84th, and 95thpercentiles, indicated by p16, p50, p84, and p95,
respectively.

Table 1 shows the summary statistics per microenvironment for ‘Total’ RF-EMF exposures
measured with ExpoM-RFs. The area with the highest ‘Total’ average exposure (0.891.07 V/m) was
CBD 1. The other areas in the CBD: CBD 2 (0.790.89V/m) and CBD 3 (0.720.88 V/m), ranked 2nd
and 3rd in highest ‘Total’ average exposure. The lowest average ‘Total’ average exposure was
measured in the less populated suburban and industrial areas: the industrial area (0.090.1 V/m),
suburban area 4 (0.050.04 V/m), suburban area 5 (0.060.06 V/m), and suburban area 6 (0.120.2
V/m). Most of the ‘Total’ RF-EMF exposure was attributed to down link RF-EMF signals. We
observed variations according to type of microenvironment. The areas in the CBD were those with
the highest total RF-EMF. The CBD has a high population, RF-EMF-user, and RF source density. A
lower total RF-EMF exposure was observed in the sub-urban areas with lower population density
than those areas in the CBD.
We found high Spearman correlations between the summary statistics of total obtained along both
paths in the same microenvironment: rs² = 0.90, 0.93, 0.95, 0.87, and 0.87, for the mean, the 16th,
50th, 84th, and 95th percentiles, respectively. Indicating that path selection within the
microenvironments was less critical. These findings are in line with those published by Sagar et al.
(2016) and Bhatt et al. (2016) who found high correlations r = 0.9 and r = 0.74 respectively, were
found for total exposure on two repetitions of the same path.
24 simultaneous measurements using both the PDE-Helmet and either one or two ExpoMs-RF were
conducted. Summary statistics were determined for the 24 measurements and correlated. These
are shown in Figure 2.The median values measured using the PDE-Helmet were in between 0.06
V/m and 0.31 V/m. We obtained very high correlations between both measurement devices, up to r²
= 0.94 for the median values. The measurements using the PDE-Helmet were approximately 5 dB

276
higher than those obtained using the ExpoM-RF(s). Bhatt et al. (2016) found an underestimation of
6-7 dB for the ExpoM-RFs. Aminzadeh et al. (2017) found 5 dB underestimation in the 900 DL band
for two commercial PEMs: the ExpoM-RF and the EME SPY, respectively, in line with our results.
Conclusions
This study showed variations of personal total RF-EMF exposure depending on the considered
microenvironment within Melbourne. The summary statistics obtained from measurements along
paths in 15 microenvironments showed high correlations with those obtained from measurements
along another path in the same environments, indicating representativeness of the selection of
paths within the chosen microenvironments for the total RF-EMF exposure. The results obtained
during simultaneous measurement using the PDE-Helmet and the ExpoM-RF show high
correlations, which serve as a validation of the measurement devices.
References
Aminzadeh, et al. 2018. Sensors 18(1):272.
Bhatt, et al. 2016. Environmental Research 151: 547–563.
Bolte, J.F.B. 2016. Environment International 94: 724-735
Röösli, et al. 2010. Environ Health. 9, 23, 1–14.
Roser K, et al. 2016. International journal of public health 2016: 9 pages.
Sagar, et al., 2016. Environ. Res 150,289–298.
Thielens, et al. 2018. Environmental Research 162: 81-96.

Figures

277
Figure 1. The PDE-Helmet. The locations of the antennas inside the helmet are shown on top.
The interior of the helmet showing the padding in black, the folded stub antennas with receiver
nodes in green-black-gold, and the battery in white, is shown on the bottom left. The device in
its final form-factor is shown on the bottom right.

278
 Figure 2. Scatterplot of the median of the Electric Field Strength in the 900 DL band.
Measured using body-worn ExpoM-RFs (horizontal axis) and the PDE-Helmet (vertical axis) in
3 preselected microenvironments in which 2 paths are repeated 4 times.

S14-4 [17:15]
Long-term spatio-temporal RF-EMF exposure assessment in sensor network
Sam Aerts1, Joe Wiart2, Luc Martens1 & Wout Joseph1
1Department of Information Technology, Ghent University / imec, Ghent, Belgium, 9052
2LTCI, Chaire C2m, Télécom ParisTech, Institut Mines-Télécom, Paris, France, 75013
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (published)
Presented by: Sam Aerts
Our exposure to environmental radiofrequency (RF) electromagnetic fields (EMF) at a given location
is inherently dynamic due to the constantly-changing nature of both our environment as well as the
telecommunications networks present in it. More than a year of measurement data was collected in
a fixed low-cost urban exposimeter network and analysed to build a spatio-temporal surrogate
model of the exposure to environmental telecommunications signals. We observed that by taking
into account the moment of the measurement in the modelling the accuracy of the resulting
surrogate model in the area under study was improved by up to 50% compared to models that
neglected the daily temporal variability of the RF signals.

279
Introduction
Human exposure to environmental radiofrequency (RF) electromagnetic fields (EMF) is not at all
constant in time, due to environmental changes and variations in the number of active users (as well
as the nature of their activity) in telecommunications networks. However, previous studies that
aimed to characterize this exposure have tended to ignore the temporal dimension. Here, the impact
of the temporal variability of telecommunications signals on outdoor RF-EMF exposure
characterization was investigated. Measurements of three downlink telecommunications signals –
i.e., from base station to user device – were collected during more than a year in a low-cost
exposimeter network within an urban setting. From this vast set of data, full spatio-temporal
surrogate models of the exposure in the area under study to the considered RF signals were built.
Furthermore, global profiles of the daily variation of the signals were composed and used to quantify
the improvement of adding the temporal dimension to established RF-EMF surrogate modelling
techniques.
Materials & Methods
Monitoring network
In the EU-FP7 LEXNET project, RF-EMF exposimeters were added to the SmartSantander Internet-
of-Things (IoT) platform (http://smartsantander.eu/) [Diez2014], covering an area of 0.4 km by 1.4
km. The exposimeters were developed for large-scale deployment, thus as cost-efficiently as
possible, and measure the environmental exposure (quantified by the electric field strength E, in
V/m) induced by the three most-used telecommunications technologies (at that time, these were
GSM at 900MHz (GSM900) and 1800 MHz (GSM1800), and UMTS at 2100 MHz). Frequency
bands specifically used by fourth generation (4G) Long Term Evolution (LTE) were not included, as
this technology was not yet in use when the exposimeters were installed in 2014 [Diez2014]. Each
of the considered frequency bands were alternately selected using an RF switch in the exposimeter,
and the nominal sample collection time (one value for each band) was either 5 or 10 minutes,
depending on the specific exposimeter. The data used in this study were collected during 14 months
– between 9th of December, 2015, and the 15th of February, 2017.
Temporal profiles
For the three considered telecom signals, the average trends of the signal strength were identified
over a day (i.e., the variation of the signal between 00:00 and 23:59, averaged over all
measurement days) or a week (i.e., the daily variation depending on the day of the week). In order
to obtain a smoother profile, and because the sampling time was not the same for all exposimeters,
hour-aggregated averages were used in this analysis – i.e., for each day, the (24) average electric-
field values captured between HH:00 and HH:59 (with HH = 00 to 23) were calculated. Furthermore,
to rule out any potential bias due to long-term variations in the signal (due to, for example, changes
in the network infrastructure), we used the relative variation of the signal compared to the daily
average (η). Finally, by aggregating and averaging all these relative values (over the entire 14-
month measuring interval) per time slot, defined as (a) hour (HH, with HH = 00 to 23), or (b) day-
hour (DDD-HH, with DDD = MON to SUN, or Monday to Sunday), temporal profiles were created.
Spatio-temporal surrogate modelling
In the final step, we investigated the possibility of creating, at any instance of time h, a spatial
surrogate model using a subset of the measurements collected over the measuring interval. This
subset, i.e., the electric-field values E(hi, Xi) (i = 1 … N) collected at instances hi and locations Xi,
would mimic a real-life measurement campaign with multiple measurement devices (of the same
type) at the same time and possibly returning to the same locations at different instances. By
rescaling the used samples E(hi, Xi), which were collected at hi, to the considered instance h, using
the identified temporal profiles, and subsequently interpolating them (using ordinary kriging) over the
study area, we created a surrogate model (at instance h) that should account for the signal’s
temporal variability while performing the measurement campaign.
To assess the validity of this spatio-temporal modelling, first, the ‘correct’ spatial model at h was

280
constructed using the exact information available at that instance (i.e., E(h, Xk), with k = 1 … the
number of exposimeters) – we call this the reference model. Then, three types of models were built
from the subset E(hi, Xi): a non-adjusted model (no rescaling), an hour-adjusted model (rescaling
according to the hour profile), and a day-hour-adjusted model (rescaling according to the day-hour
profile). All three types were compared to the reference model by calculating the relative error of the
model compared to the reference model (called temporal bias).
Results & Discussion
Monitoring network
The SmartSantander exposimeter network consisted of 36 exposimeters that were active during any
period of time between 9th of December, 2015, and the 15th of February, 2017. For all further
analysis, only the exposimeters with an adequate amount of measurements were retained to create
temporal profiles (i.e., 50 days, or 14,400 or 7,200 measurements, at respective measurement
collection times of 5 min and 10 min) and removed the (four) malfunctioning devices from the
analysis. The locations (and IDs) of the 32 considered exposimeters of the network considered in
this study are shown in Figure 1.

F 1: T (X,Y)- ID .
Temporal profiles
In order to mitigate the modelling errors induced by performing measurements at different time
instances, temporal profiles were constructed. From Figure 2, in which the hour profiles (for which
the day of the week was disregarded) are plotted, one can see that the daily variation in the UMTS
band was the highest, although the extreme variations remained quite small (0.60 – 1.40). For
GSM1800, as well as GSM900 and the total field, the electric-field strength hovered between
+/-20% of the daily average. Comparing weekdays (not shown here), Monday to Thursday appeared
to be fairly similar, although Friday, Saturday, and Sunday all boasted distinct variations.
Furthermore, the weekend nights exhibited higher variations, while they were lower for weekend
mornings and afternoons. Here as well, these trends were more outspoken for UMTS than for the
other signals.

281
 F 2: H ( ).
The fact that η remains so close to 1 (Figure 2), in contrast to the high variability observed between
single, momentaneous measurements (which, we observed, could reach 42dB, or a factor 12,000 in
power density) emphasizes the potential errors inherent to instantaneous measurements, and
justifies the use of hour-aggregated values in further analysis.
Spatio-temporal modelling
For the final step, we selected December 1st, 2016 to December 7th, 2016 as our measuring
interval. For illustration, the reference model at h = 04-12-16, 14:00 of the GSM900 exposure in the
area under study is shown in Figure 3.

F 3: S ( ) = 14:00, D 4 ( S )
GSM900 ( ). T ( F 1).
From the total set of (hour-averaged) measurements between 01-12-16, 00:00, and 07-12-16,
23:00, subsets of 100 samples, taken at random locations and time instances, were used to build
surrogate models at each time instance h in the measuring interval, using (i) the non-adjusted
values, (ii) the day-hour-adjusted values, and (iii) the hour-adjusted values. This was done for each
282
signal. Then, to quantify the usefulness of the introduced temporal profiles, for each time-instant h
within the considered measuring interval, the reference models were subsequently compared with
the (non-)adjusted surrogate models. The medians of the introduced temporal bias are listed in
Table 1.

T 1: M -, -,
- .
It is clear that using temporal profiles to rescale measurements to the correct time significantly
reduced the temporal bias introduced in the model. The day-hour-adjusted models clearly offer the
best results (on average 50% improvement; Table 1), though using the hour profiles offered a
reasonable improvement as well (on average 38% improvement; Table 1). Of the three signals, the
temporal bias, as well as the potential reduction, was the highest for UMTS, which follows the
observation that the average variation of EUMTS over a day was larger than those of the GSM
bands. This behavior indicates a larger dependency on the UMTS network, especially during peak
hours (around noon and 7pm – Figure 2).
Conclusions
Due to short-term variation over a day or over a week in environmental RF-EMF, measurements
performed at different moments during this time can introduce a temporal bias in surrogate models
of the RF-EMF exposure such as created in previous studies [e.g., Aerts2013]. It was found in this
study that, by rescaling measurements taken at different moments to the same instance of time
using temporal profiles, i.e., average trends of the electric-field strength over a certain period of
time, a reduction of the temporal bias by up to 40% can be achieved when adjusting only for the
hour, and 52% when also adjusting for the day of the week (both for UMTS). The mitigation worked
best for UMTS because of its smoother and more outspoken temporal variation during a day. Finally,
the work presented here indicates that the methodologies used in previous studies should include
time, either as a fully-fledged extra dimension (next to x and y) which is brought into the smart-
sampling methodology or through long-term temporal measurements at one or more locations within
the area to obtain an area-averaged temporal profile. This, as well as the validation of the reported
findings in other cities and environments, will be the subject of future research.
References
[Aerts2013] Aerts S., Deschrijver D., Verloock L., Dhaene T., Martens L., and Joseph W. 2013b.
Assessment of outdoor radiofrequency electromagnetic field exposure through hotspot localization
using kriging-based sequential sampling. Environ. Res. 126:184–191.
[Diez2014] Diez L. F., Anwar S. M., De Lope L. R. , et al. 2014. Design and integration of a low-
complexity dosimeter into the smart city for EMF assessment. EuCNC 2014 - European Conference
on Networks and Communications.

S14-5 [17:30]
Personal Exposimeter to monitor EMF Up-Link exposure from daily-usages of mobile phone
Serge Bories1, David Dassonville1, Sébastien Brulais1, Saifeddine Aloui1, Yenny Pinto2, Taghrid Mazloum2 &

283
Joe Wiart2
1CEA LETI Minatec Campus, Grenoble, France
2Télécom ParisTech, Paris, France
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Serge Bories
The cellular Up-Link (UL) exposure has been identified as the main contributor for the mobile phone
user. However current personal exposimeters don't estimate it properly because they don’t assess
the real transmit (Tx) power and the distance between this EMF source and the user’s body. A novel
non-invasive system monitors both of these parameters (power, usage) during daily usages. It is
integrated on a shell around the user’s mobile. Covering 5 cellular UL frequency bands, the probe is
able to log the Tx RF power from 100 Hz sampling. The usage estimation is assessed through three
small sensors worn in the user pockets. The variations of the Tx power at different time-scale are
demonstrated.
Context and challenge
Assessing the real EMF exposure of mobile phone users is a hard challenge because of the high
variability of cellular network configuration and regarding new usages.Manufacturers of
smartphones provide a SAR value where the device is used at its maximum RF output power. For
the latest generation of smartphones, the actual SAR is lower than the normative SAR, and depends
on the actual usage, and also the network in which it is operated.In project DEVIN, an add-on
module has been specified and developed, in order to measure the actual SAR of the user during
the utilization of the equipment. To this aim, two questions must be answered: what is the effective
Up Link (UL) RF power transmitted by the smartphone, and what is the position of the smartphone
relative to the user. The overall objectives of this study is to be able to assess the statistical
distribution of UL exposure, to embrace the high variability of such a metric, and its eventual
evolution regarding long term.
System description
The coined 'DEVIN' system concept is remind on figure 1 and a prototype is shown on figure 2.
The uplink RF power is measured by an RF probe that is fitted in an add-on module on the
smartphone. A plastic sheel is just covering the user terminal who can keep his habits and usage
over the whole day. No additional application is running on the mobile which can maintain its usual
power consumption. The RF probe is able to measure the evolution of the RF power in pre-defined
cellular bands, independently from the radio-access-technology (RAT). It can handle2G (GSM 900
MHz and DCS 1800 MHz), 3G (1900 MHz) and 4G (800 MHz and 2600 MHz bands) through an RF
passband filter bank that ensures proper isolation with a plus or minus 1 dB accuracy on a given
RAT band. The RF probe is just coupling -20 dB (1% power) to the antenna transmit power. This
has no impact on radio link quality, neitheir on the UL power control, nor on the SAR with the add-on
module. For all the RATs, the transmit power can be measured from +23 dBm down to -33 dBm. The
calibration of this device has been done with a given mobile phone and a Base station emulator
equipment. Some additional test have been carried on with a trace mobile in order to compare our
results with some UL signal power available from this specific mobile phone.
The data for the uplink power and the distance is instantaneously measured with 1 kHz sampling
frequency and collected in a memory after decimation (100 Hz to 1 Hz saved data). At the end of a
recording period, the data is transferred and can be analyzed offline.
The relative position of the smartphone is measured by using small beacons attached to the user's
clothes: belt, chest. By measuring the distances between the smartphone and the beacons, it is
possible to extract a proxy of the distance from the smartphone to the head or torso of the user.
Main results
The real Tx power from the mobile phone is monitored during 'played scenario' of daily usages. This
indoor scenario has been played in urban environment during late night when the cellular network

284
load is low. The base station is 2 km away with rather good propagation conditions. The same files
size are sent 1.4 Mo for a picture, 64 Mo for a first video and 247 Mo for the second video. The
mean UL power (1 point/10 ms) evolution regarding time and several usages is observed in the 4G
LTE 2600 band, figure 3a. For about half an hour, the power is significant (>-25 dBm) only when
user sends files. The +2 dBm maximum level is reached for the heavy video file when propagation
conditions are volontary perturbated (hands on the antenna). The mean level is -20 dBm over the
whole 30 min.
Figure 3b presents a zoom on the last burst of the second video transmission (12 s lenght). It shows
how fast the power control from the base station can adapt the transmitted power with more than 20
dBm variation within two first seconds. Some -10 dBm peaks appear each 0.2 s but they are too
short to contribute a lot to the -16 dBm average level on these 12 s.
Acknowledgment
The authors thank the French governemental agency ANSES for funding this study under thePNR-
ESTframe and within the project DEVIN –N°2014-RF-26.
Figures

Figure 1. Personal exposimeter concept

285
 Figure 2. Picture of the DEVIN exposimeter on the user smartphone (a), details on the
electronic boards (b)

Figure 3. Evolution of Tx power on the 4G 2600 MHz band regarding time and different
usages over half an hour (a), zoom on 12 s (b)

S14-6 [17:45]
Characterization of the exposure due to smart-home devices and other residential RF
sources
Sam Aerts1, Leen Verloock1, Matthias Van Den Bossche1, Ximena Vergara2, Luc Martens1 & Wout Joseph1
1Department of Information Technology, Ghent University / imec, Ghent, Belgium, 9052
2EMF/RF Program, Energy & Environment, Electric Power Research Institute (EPRI), Palo Alto, CA, USA
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Sam Aerts
Current Internet-of-Things (IoT) trends such as home automation and smart metering may raise
concerns about the human exposure to radiofrequency (RF) electromagnetic fields (EMF) due to

286
new wireless devices installed in residences. As signals transmitted by smart devices are usually
non-continuous, a new measurement methodology was developed to assess their exposure levels,
using the spectrogram mode of a spectrum analyzer to capture changes in frequency and/or
amplitude of the assessed signals over time. Here, the assessment of 94 residential RF sources is
presented. It was found that, in the smart home, wireless access points, smartphones, and other
personal communication devices continue to represent the bulk of our RF-EMF exposure.
Introduction
The work presented here addresses the issues concerning human exposure to radiofrequency (RF)
electromagnetic fields (EMFs) that may arise from implementing the Internet-of-Things (IoT)
approach (“everything is connected”) in residences, including current trends such as home
automation (“smart home”) and energy grid monitoring (“smart grid”). Key components of the smart
grid are so-called smart meters, which encompass a broad range of sophisticated sensors that
constantly assess the state of the grid, the availability of power flowing into the grid, and the demand
on the grid. However, as these devices are installed in the customer’s residence and often
communicate wirelessly using existing RF networks, members of the general public might feel
saddled with these new technologies due to their proximity and visibility. In this study, a novel
measurement method was designed to characterize in-situ residential sources of RF radiation (and
in particular, IoT devices and smart meters), primarily with the use of the spectrogram mode on a
spectrum analyzer. This method was then applied to a convenient sample of ten residences in
Belgium and France containing various IoT devices, smart meters, and other residential RF-emitting
sources.
Materials & Methods
As the signals transmitted by smart-home and other IoT devices are usually characterized by very
short pulses – the number of which depends on the network topology, and sometimes on the user –
for each smart apparatus the following were determined: (a) the transmission frequency, (b) the duty
cycle DC, i.e. the proportion of the time the device is actually transmitting a signal, (c) the worst-
case exposure level Emax(in V/m) – assessed at 0.2 m, 0.5 m, and 1 m – and (d) the actual (time-
averaged) exposure Eavg, calculated as Eavg= √DC*Emax. The latter can be compared to the
exposure guidelines issued by the Federal Communications Commission (FCC) in the USA (30-min
time average) [FCC1997] and/or the International Commission on Non-Ionizing Radiation Protection
(ICNIRP) elsewhere (6-min time average) [ICNIRP1998], by calculating the ratio RS:

with Eref the reference level (issued by FCC/ICNIRP) corresponding to the frequency of the signal of
the assessed device. As long as RS is lower than 1, the assessed field complies with the guidelines.
All measurements were performed with a spectrum analyzer combined with a tri-axial antenna with
frequency range 30 MHz - 3 GHz. Depending on the type of signal to be measured, appropriate
settings of the spectrum analyzer were needed. For example, as the RF signals emitted by IoT
devices are typically non-continuous, the spectrogram mode of a Rohde & Schwarz FSVA40
spectrum analyzer was used to capturing the changes in frequency and/or amplitude of the
assessed signal over time, which allowed to easily determine the device’s duty cycle.
Results & Discussion
Example measurement
Here, a specific smart-meter measurement is highlighted: the communications module (CoMo) of a
smart electricity meter. In Belgium, it is usually the electricity meter that is equipped with a wireless
communication link to the smart grid. Other smart meters possibly present in the home are
connected to the CoMo using either a wired link or a wireless Meter-Bus or Wireless Fidelity (Wi-Fi)
link.The CoMo usually communicated with the central system of the energy supplier through

287
General Packet Radio Service (GPRS) technology, similar to GSM, transmitting in the GSM900
uplink band (in this case, at 904 MHz). The signal was measured in time using the spectrogram
mode to obtain more detailed information about the rate of transmission, or in other words, about the
duty cycle. In theory, a CoMo transmits once every 15 min. However, in our measurements, we
observed durations between transmissions (denoted by Tperiod in Figure 1) as low as 43 s.
Furthermore, each transmission (or pulse) consisted of a series of bursts sent over a 3.6 s interval
(Tpulse in Figure 1). Combined with the transmissions technology’s inherent duty cycle of 1/8 (just
like GSM, GPRS uses time division multiple access or TDMA), the CoMo’s theoretical duty cycle
was 0.05% and the worst-case (withTperiod=43 s) duty cycle was 1.05%.

F 1: S '
(C M ) ( ).
Summary of the measurements
In total, the RF signals emitted by 53 in-home devices were characterized. In general, a distinction
could be made between user-controlled devices (e.g. remote controls), independent devices with
random or variable pulses (e.g. Wi-Fi routers), and independent devices with fixed pulses (e.g. most
smart meters). In addition, o put the IoT measurements in a more familiar perspective, 41
measurements of the signals emitted by a mobile (‘uplink’) using three telecommunications
technologies (Global System for Mobile Communications, Universal Mobile Telecommunications
System, and Long Term Evolution) were performed at a distance of 0.5 m from the used phone. In
Figure 2, the measurement results at 0.5 m are shown, with the assessed RF-emitting devices
grouped per category.

288
F 2: E - ( 0.5 ) RF- :
( ) ,( ) - , 6-
.
In Figure 2(a), the maximum measured electric-fields are compared. The highest field value was
measured for a CoMo of an electricity meter (Emax = 20.319 V/m). In Figure 2(b), these values are
time-averaged, using the duty cycle determined in a 6-min interval (in order to compare to the
ICNIRP guidelines). The duty cycle of non-user-controlled devices, such as smart meters, was
usually fixed. However, for user-controlled devices, the duty cycle was calculated based on a single
action (e.g. push on a button) within six minutes; for smartphones, continuous use (voice or data
transfer) during six minutes was assumed (i.e. DC = 100% for UMTS and LTE, and DC = 12.5% for
GSM, due to the time-division multiplexing of its signal); and for Wi-Fi access points, the actually
measured duty cycle of the dominant channel was used (overall,DCavg= 6.9%).
All measurements satisfied the ICNIRP (and the FCC) guidelines. The highest RS value (at 50 cm,
and assuming a realistic duty cycle) for a non-smart or IoT device was 0.025 (six-minute call with a
walkie-talkie), for a smart-home device 0.002 (smart toothbrush), and for a smart meter 0.0004
(CoMo of anelectricity meter, with theoretical DC). For uplink telecom signals, the highest RS value
found was 0.059 (LTE) – on average, UMTS uplink had the lowest RS values of the telecom signals.
Conclusions
A novel measurement method was designed to characterize the RF-EMF exposure due to in-situ
residential RF-emitting sources, which comes down to determining their transmission frequency,
their peak emitted fields at various distances, and the percentage of time they transmitted (i.e. their
duty cycle). This method was then applied to a convenient sample of ten residences in Belgium and
France containing, in total, 53 IoT devices, smart meters, and other RF-emitting devices. The
measured emissions were also compared to emissions by a mobile phone using three current
telecommunications technologies (GSM, UMTS, and LTE), as well as to the ICNIRP guidelines for
general public RF exposure.
In the smart home, wireless access points, smartphones, and other personal communication

289
devices (e.g. DECT cordless phones, walkie-talkies) continue to represent the bulk of our exposure
to RF-EMF due to their typically high emissions and use close to the body. However, smart-home
and other IoT devices with high duty cycles (e.g. motion detector, baby monitor, toothbrush) can
significantly heighten our exposure, if used close to the body. Smart meters, and in particular
communications modules wirelessly linked to the utility company’s central network, contribute only
little to our RF-EMF exposure, regardless of their sometimes high emissions, due to their relatively
rare transmissions. It should be noted that, as the specific physical environment in which RF
sources are placed has a significant impact on the RF-EMF levels in a residence, the
measurements presented here represent a sample cross-section in time of the assessed residential
RF environments. However, the described results should be sufficient to illustrate a typical RF
environment in our modern life, in which everything is connected.
Acknowledgement
This study was supported by the Electric Power Research Institute (EPRI) (USA).
References
[FCC1997] Federal Communications Commission (FCC). “Evaluating Compliance with FCC
Guidelines for Human Exposure to Radiofrequency Electromagnetic Fields”, OET Bulletin 65, edition
97-01, August 1997.
[ICNIRP1998] International Commission on Non-ionizing Radiation Protection (ICNIRP). “Guidelines
for limiting exposure to time-varying electric, magnetic, and electromagnetic fields (up 300GHz),”
Health Physics, vol. 74, pp. 494-522, 1998.

Session: S15
Dosimetry IV - Computational studies
Friday June 29, 2018 • 09:30 - 11:00
Europa C
Chairs: Bor Kos & Kenichi Yamazaki

S15-1 [09:30]
NEUROMAN: Reference computational human anatomical models for therapeutic peripheral
nerve stimulation and safety investigations
Bryn Lloyd1, Antonino Mario Cassara1, Silvia Farcito1, Esra Neufeld1, Jin Seo Park4, Beom Sun Chung3, Min
Suk Chung3 & Niels Kuster1, 2
1IT'IS Foundation, Zürich, Switzerland, 8004
2ETH Zürich, Zürich, Switzerland, 8092
3Ajou University, Suwon, Korea, 16499
4Dongguk University, Gyeongju, Korea, 38046
Keywords: Dosimetry (computational), ELF/LF, Work in Progress
Presented by: Bryn Lloyd
Therapeutic electrical stimulation of peripheral nerves is a promising technique with a wide range of
treatment applications. At the same time, the trend towards stronger magnetic fields and/or faster
gradient switching in magnetic resonance imaging poses safety risks for patients, e.g., due to tissue
heating and unwanted neurostimulation. To this end, we are developing reference human
anatomical models with unprecedented details in the peripheral nervous system, connectivity to
organs and muscles, and functionalized with compartmental nerve models to investigate interactions
with neuronal electrophysiology. We employ these phantoms to investigate current safety guidelines.
Introduction
Magnetic resonance imaging (MRI) is a powerful, non-invasive diagnostic technique routinely used

290
in clinical settings and research. While stronger magnetic fields and faster, more intense gradient
switching could improve image resolution or functional contrast, in many MRI-based technologies
(e.g., functional MRI, diffusion tensor imaging, etc.), there are accompanying potential risks of
radiofrequency (RF) tissue heating and unwanted peripheral nerve stimulation (PNS). These risks
are even greater in subjects wearing metallic implants. While computational techniques to quantify
RF exposure, power absorption, and thermal consequences are well established1,2 and regulated
by safety standards and guidelines3-7, predictive tools to study nerve electrophysiology within the
complex human anatomical environment and estimate the risk of PNS are lacking.
We previously developed the concept of neuro-functionalized models8 that includes populations of
axons geometrically represented as splines embedded in the human anatomy. Electrophysiological
models (e.g., SENN, Sweeney, etc.) assigned to the spline trajectories are used to predict the
responses of axons and nerves to electromagnetic stimuli. Embedding the nerve models in the
anatomy facilitates coupled simulations of the neuro-physiology with induced electrical fields and
non-trivial field gradients that result from the complex dielectric material distributions inside the body.
This methodology has been successfully applied in different contexts, e.g., the development of
electroceuticals9, in investigations concerning non-invasive brain stimulation10, concerning the risk
of unwanted PNS by MRI gradient switching11, and in studies aimed at critical revision of current
safety standards12. These investigations provide insight into principal mechanisms and causes of
neurostimulation and largely reproduce experimental findings. However, the lack of realistic
anatomical models with detailed nerve trajectories extracted from 3D image data prevents
quantitative prediction in the context of PNS in MRI.
To overcome these limitations and provide solutions to mitigate PNS, we initiated the NEUROMAN
project, in which we aim to create two reference high-resolution neuro-functionalized human
anatomical models. The two phantoms, a female and a male, include detailed mapping of nerve
trajectories of the peripheral nervous system, from the cranium and spinal cord to internal organs
and all major muscles of the human body.
Methods
The Visible Korean Human13 male (33y, 1.64m, 55kg) and female (26y, 1.52m, 55kg) cryosection
data are being used as the basis for the new phantoms due to the unique resolution
(0.2×0.2×0.2mm) and quality of these images. To segment important peripheral nerves, a nerve
tracing tool has been developed, which allows to semi-automatically extract smooth nerve
trajectories by specifying sparse (start/end) points along a nerve in the cryosection image stack.
Functionalization is achieved by assigning electrophysiological models of myelinated and
unmyelinated axons to axon trajectories within nerve models based on histological investigations
documented in the literature.
The computational phantoms will continue evolving, but already over 900 different tissues and
structures have been segmented, including more than 320 muscles. Following major nerves are or
will soon be completed in the male model: the vagus nerve and other cranial nerves and the lumbar,
brachial and sacral plexus. To ensure high quality standards, we follow an internal/external review
approach similar to that of the Virtual Population models2.
Results
Preliminary results on simplified implant leads and electrode geometries proximal to nerve
structures (Fig. 3) with electrophysiological axonal models exposed to electric fields in the
10–100kHz frequency range, demonstrate that neurostimulation and conduction blocking (CB) of
nerve signals are possible risks even at exposure levels in compliance with standards3,4, due to
simultaneous interference between pulsed MRI gradients and medical implants and field
enhancement at lead electrodes. Effective exposure thresholds depend on the tissue material
properties surrounding the nerve, implant trajectories and shape, pulse sequences and gradient
coils. In current work, we explore the risk of PNS and CB near implants within anatomical bodies

291
and for MRI relevant waveforms, e.g., 3D echo-planar imaging and spiral waveforms.
Conclusions
The developed phantoms, which constitute an important expansion of the Virtual Population
(www.itis.ethz.ch/vip), which includes reference models for a wide range of biomedical applications.
The NEUROMAN models are expected to significantly impact the scientific community and the field
of EM-neuron research to enable studies of multi-scale models with realistic anatomies and
electrophysiology. The results will enhance our understanding of mechanisms of neurostimulation,
e.g., in MRI, provide experimental test-beds for new therapeutic approaches and devices, and
enable study of safety aspects, providing a tool to facilitate regulatory submissions and
standardization activities.
Acknowledgements
This research has received funding and support from the Swiss Commission for Technology and
Innovation (CTI 25290.1 PFLS-LS) and the Korean Institute for Advancement of Technology.
References
1. Murbach M, Neufeld E, Kainz W et al. Whole-body and local RF absorption in human models as a
function of anatomy and position within 1.5T MR body coil. Magn Reson Med 2014;71:839–845
2. Gosselin M-C, Neufeld E, Moser H, et al. Development of a new generation of high-resolution
anatomical models for medical device evaluation: the Virtual Population 3.0. Phys Med Biol
2014;59:5287–5303
3. ICNIRP 1998 Guidelines for limiting exposure to time-varying electric, magnetic, and
electromagnetic fields (up to 300 GHz) Health Phys. 74 494–521
4. ICNIRP 2010 Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz to
100 kHz) Health Phys. 99 818–36
5. IEEE-C95.1 2006 IEEE Standard for Safety Levels with Respect to Human Exposure to Radio
Frequency Electromagnetic Fields, 3 kHz to 300 GHz (New York: IEEE)
6. IEEE-C95.6 2002 IEEE Standard for Safety Levels with Respect to Human Exposure to
Electromagnetic Fields, 0 to 3 kHz (New York: IEEE)
7. Reilly PJ, Hirata A, Low-frequency electrical dosimetry: research agenda of the IEEE International
Committee on Electromagnetic Safety, Phys Med Biol 2016, 61(12):138
8. Neufeld E, Cassará AM, Montanaro H, Kuster N, Kainz W. Functionalized anatomical models for
EM-neuron interaction modeling. Phys Med Biol 2016, 61(12):4390
9. Cassara AM, Neufeld E, Kuster N. Vagus Nerve Stimulation: Comparison of Thresholds for A-,
B-and C-Fiber Recruitment for Cuff Electrodes and Alternative Stimulation Configurations in
Functionalized Generic and Anatomical Models
10. Cassará AM, Neufeld E, Guidon M et al. Large-Scale Multi Neuronal Simulation within an
Anatomical Head Model for Transcranial Alternative Current Stimulation (tACS) Investigations, Proc
Joint Meeting of BEMS and EBEA 2016
11. Cassará AM, Neufeld E, Hagberg G et al. Peripheral Nerve Stimulation in MRI: Insights from a
three level analysis and coupled EM-electrophysiological simulations in neuro-functionalized human
models, Proc ISMRM 2017
12. Neufeld E, Oikonomidis IV, Iacono MI, et al. Investigation of assumptions underlying current
safety guidelines on EM-induced nerve stimulation. Phys Med Biol 2016, 61(12):4466
13. Park JS, Chung MS, Hwang SB et al. Visible Korean human: improved serially sectioned images
of the entire body. IEEE Trans Med Imaging 2005, 24(3):352-60
Figures

292
Figure 1. The NEUROMAN male anatomical phantom already contains over 900 tissues and
structures, including more than 320 individual muscles. On the left only the muscles are
depicted. The right shows internal organs, bones and the skin (transparent).

293
Figure 2. A slice through the NEUROMAN male right arm with the median nerve highlighted
(red circle) in the cryosection image (left) and segmentation (right).

Figure 3. Model used to quantify exposure thresholds for activation (or conduction blocking) of
axons within a nerve surrounded by electrodes, as a consequence of interaction between
incident fields in the kHz frequency range and implants, resulting in field enhancement around
the electrodes. In this example, cuff electrodes surround a model of a simplified vagus nerve
featuring fascicles and electrophysiological models of axons. The incident field is parallel to
the lead. Small red spheres indicate spiking neurons.

294
S15-2 [09:45]
Enabling LF dosimetry starting from the knowlege of the magnetic flux density
Fabio Freschi1, 2, Luca Giaccone1, Vincenzo Cirimele1 & Aldo Canova1
1Dipartimento Energia "G. Ferraris", Politecnico di Torino, Torino, Italy
2School of Information Technology and Electrical Engineering, The University of Queensland, Brisbane,
Australia
Keywords: Dosimetry (computational), ELF/LF, Completed (published)
Presented by: Luca Giaccone
LF dosimetry is commonly assessed by evaluating the electric field in the human body using the
scalar potential finite difference method. This method is effective only when the sources of the
magnetic field are completely known and the magnetic vector potential can be analytically
computed. This paper presents a rigorous method to characterize the source term when only the
magnetic flux density is available at discrete points. The proposed method is validated and also
proved to be useful and effective when applied to a real-world scenario, where the magnetic flux
density is measured in proximity of a power transformer.
Introduction
The computation of the in situ electric field, in the low-frequency range, is often done using the
quasi-static approximation (Chen et al 2014, Dawson and Stuchly 1998, Laakso et al 2012) which
leades to:

where σ is the electric conductivity, φ is the electric scalar potential and A is the magnetic vector
potential due to the sources. Under this approximation, the magnetic vector potential is considered
as unperturbed by the presence of the biological tissues and the study can be limited to the human
body only. This approximation holds true for frequencies up to100 kHz where the contribution of the
electric permittivity can also be neglected (Dawson and Stuchly 1998). This hypothesis is exploited
in Dawson and Stuchly (1996) where a numerical scheme called scalar potential finite difference
(SPFD) is proposed. This method has been extensively used in the literature (So et al 2004,
Dimbylow 2005, Canova et al 2010, Hirata et al 2010, Zoppetti et al 2011, Laakso and Hirata 2012,
Canova et al 2016a). The SPFD is simple and effective when the source can be accurately modeled
and the right-hand side of the SPFD can be analytically evaluated (Barchanski et al 2006). However,
in some real cases, its application is not straightforward. When the source is not completely known,
its effect can be only characterized with the measurement of the magnetic flux density. In this case,
the magnetic vector potential is not readily available and the application of SPFD is not possible.
Another case is a simulation scenario when the source field is numerically evaluated using suitable
softwares (Cirimele et al 2016, Zang et al 2017). In most cases, softwares do not provide the
magnetic vector potential but only the magnetic flux density.
The literature covers methods to obtain a compatible magnetic vector potential directly from the
magnetic flux density. For instance, the inverse-curl operator is discussed and applied to analytical
examples in Sahoo (2008). Recently, this approach has been implemented in the context of
numerical dosimetry in Laakso et al (2017). Authors reconstruct the magnetic vector potential in a
bounded rectangular domain where the magnetic field is known on a grid of regularly spaced points.
The method is not tested using real measurements, but results include the effect of random noise.
Another attempt to invert the curl operator is done in Cimala et al (2015) and Zang et al (2017),
where authors analyze an exposure scenario related to a wireless power transfer system. The
numerical formulation used to model the wireless system does not provide directly the magnetic

295
vector potential, therefore the unperturbed magnetic flux density is computed in the human body
voxels. The line integral of the magnetic vector potential is gauged using a tree-cotree
decomposition and calculated solving a overdetermined system.
On the basis of the simple idea used in Cimala et al (2015) and extending the preliminary work
reported in Canova et al (2016b), we developed a consistent and systematic method to identify the
line integral of the vector potential directly on the computational grid to be used with the standard
SPFD. The work is published in Canova et al (2018). Starting from the knowledge of the magnetic
flux density on a set of discrete points, the magnetic flux through the faces of the computational grid
is calculated. Using topological identities, a well-posed linear system is built and solved to
reconstruct a compatible right-hand side for the SPFD, overcoming the computational issues that
arise using the simple tree-cotree decomposition. When the magnetic flux density data are not
directly provided on the computational grid, an interpolation step is necessary. To preserve the
solenoidality of the magnetic flux density, a suitable divergence-free interpolation method based on
vector radial basis functions is used.
Dosimetry from B-field measurements: an example
The details of the proposed method can be found in Canova et al (2018) whereas in this abstract we
show the application of the method starting from measurements.The magnetic field generated by
the 15 kV/400 V, 630 kVA transformer shown in Figure 1 is analyzed. The voltage at the primary
side is regulated in order to provide a secondary current of approximately 820 A rms.Figure 1(a)
also shows the wooden structure in front of the transformer used to measure the magnetic field in
the region with dimensions 0.5 m x 1 m x 2 m shown in Figure 1(b). The grid size for measurements
is 10 cm wide, that corresponds to 1386 regularly spaced points.The measurement setup provides
the three orthogonal components of the magnetic flux density at each investigation point.
Measurements are freely available for download here https://github.com/giaccone
/field_measurements.
The B-field is measured in a region with a grid size coarser than the one of the computational
domain. Therefore, it is required the interpolation of the mesurements on the human model. It is
shown that a simple linear interpolation causes numerical artifacts like the one in Figure 2(a). To
avoid this problem one has to interpolate preserving the solenoidality of the magnetic flux density.
This is obtained by a suitable interpolator based on vector radial basis functions. With this
interpolator (fully described in Canova et al (2018)) the proposed proceure provides the results in
Figure 2(b).
Some representative tissues are selected to show the maximum exposure obtained with the two
interpolators. As it is shown in Figure 3, the simple linear interpolation causes a significant
overestimation of the exposure.
Conclusions
This paper presented a rigorous method to identify a compatible right-hand-side for the SPFD
formulation due to sources of magnetic field. The method requires the availability of the magnetic
flux density coming from measurements or numerically evaluated. The magnetic fluxes through the
faces of the discretization are obtained by integration, then the source contributions are obtained
directly on the computational mesh in terms of line integral of the magnetic vector potential. The
procedure is based on the solution of a linear system coming from the discrete magnetic curl
equation. Differently from the up-to-date literature, the linear system is reduced according to
topological information that make the system consistent with a unique solution. The procedure is
purely based on the topological information of the computational mesh, and for this reason is
completely automatic. The method may be applied to the standard scalar potential formulation. In
addition, it can be extended to any discretization type (e.g. unstructured meshes using tetrahedra)
since the topological identities are not limited to specific elements.
Results show that the complexity of the method in terms of computational time is almost linear up to
20 million of elements. The additional effort required for the calculation of the right-hand side is

296
compatible in terms of computational time with respect to a standard dosimetric simulation.The
application of the method is restricted to simply-connected domains. This limitation required the
preprocessing of the computational domain (i.e. the voxel model) in order to remove the holes and
handles such that it could be continuously deformed into a sphere.
The method were tested on a benchmark problem and on a real exposure scenario where the
magnetic field due to a power transformer was measured. The overall computational time including
the calculation of the right-hand side, the solution of the SPFD equation and the post-processing is
less than 5 min that is more than acceptable in these applications. The execution of the field
measurements is the most time-consuming step. However, it is worth noting that the method is
advantageous also when the complexity of the sources of the magnetic field requires the use of
dedicated software that does not provide specific tools for the dosimetric analysis. In these cases,
the magnetic flux density at discrete points can be calculated and converted in a compatible vector
potential used to drive the standard SPFD.
When the input is a set of magnetic flux density measurements, attention must be paid on the
interpolation procedure. A simple tri-linear interpolation can lead to non-physical results due to the
non-solenoidality of the interpolated field. This issue were addressed using div-free interpolator
based on radial basis functions.
References
Barchanski A et al 2006 Efficient calculation of current densities in the human body induced by
arbitrarily shaped, low-frequency magneticfield sources J. Comput. Phys. 214 81–95
Canova A et al 2010 Exposure of working population to pulsed magnetic fields IEEE Trans.
Magn. 46 2819–22
Canova A et al 2016a A simplified procedure for the exposure to the magnetic eld produced by
resistance spot welding guns IEEE Trans. Magn. 52 5000404
Canova A et al 2016b From the magnetic eld measurement to the numerical evaluation of the
human exposure Proc. of 17th Biennial IEEE Conf. on Electromagnetic Field Computation
(Miami, FL, 13–6 November 2016) (https://doi.org/10.1109/CEFC.2016.7816215)
Canova A et al 2018 Numerical assessment of low-frequency dosimetry from sampled magnetic
fields, Phys. Med. Biol. 63 (2018) 015029 (14pp)
Chen X L et al 2014 Human exposure to close-range resonant wireless power transfer systems
as a function of design parameters IEEE Trans. Electromagn. Compat. 56 1027–34
Cimala C et al 2015 Numerical schemes for high-resolution dosimetry simulations of automotive
low frequency inductive power transfer systems Proc. of the 2015 Int. Conf. on Electromagnetics
in Advanced Applications (Torino, Italy, 7–11 September 2015) (IEEE) pp 1044–7
Cirimele V et al 2016 Finite formulation of surface impedance boundary conditions IEEE Trans.
Magn. 52 7206504
Dawson T and Stuchly M 1996 Analytic validation of a three-dimensional scalar-potential finite-
difference code for low-frequency magnetic induction Appl. Comput. Electromagn. Soc. J. 11
72–81
Dawson T and Stuchly M 1998 High-resolution organ dosimetry for human exposure to low-
frequency magnetic fields IEEE Trans. Magn. 34 708–18
Dimbylow P 2005 Development of the female voxel phantom, NAOMI, and its application to
calculations of induced current densities and electric fields from applied low frequency magnetic
and electric fields Phys. Med. Biol. 50 1047–70
Gandhi O et al 1984 Impedence method for calculation of power deposition patterns in
magnetically induced hyperthermia IEEE Trans. Biomed. Eng. 31 644–51
Laakso I and Hirata A 2012 Reducing the staircasing error in computational dosimetry of low-
frequency electromagnetic elds Phys. Med. Biol. 57 N25–34
Laakso I et al 2012 Evaluation of SAR in a human body model due to wireless power
transmission in the 10 MHz band Phys. Med. Biol. 57 4991–5002

297
 Laakso I et al 2017 Modelling of induced electric fields based on incompletely known magnetic
elds Phys. Med. Biol. 62 6567
Sahoo S 2008 Inverse vector operators (arXiv: 0804.2239)
Zoppetti N et al 2011 Evaluation and characterization of fetal exposures to low frequency
magnetic fields generated by laptop computers Prog. Biophys. Mol. Biol. 107 456–63
Figures

Figure 1. (a) 630 kVA transformer in front of the wooden structure used as a guide to perform
the measurements. (b) Representation of the region in front of the transformer with 1386
investigation points and the human model.

298
Figure 2. In situ electric field evaluated in the Duke model: (a) linear interpolation and (b) div-
free interpolation of the magnetic flux density measurements. (a) Emax = 0.151V m−1. (b)
Emax = 0.060 V m−1.

Figure 3. Exposure index evaluated at some representative tissues of the peripheral and
central nervous system using the divergence-free and linear interpolations.

299
S15-3 [10:00]
Estimation of electric fields and eddy currents induced by an AC electromagnetic field in
human subjects
Hideyuki Okano1, Ilkka Laakso2, Tsukasa Kondo3, Hiromi Ishiwatari4 & Keiichi Watanuki1, 3, 5
1Advanced Institute of Innovative Technology, Saitama University, Saitama, Japan, 338-8570
2Department of Electrical Engineering and Automation, Aalto University, Espoo, Finland, 02150
3Graduate School of Science and Engineering, Saitama University, Saitama, Japan, 338-8570
4Soken Medical Co., Ltd., Tokyo, Japan, 171-0033
5Brain and Body System Science Institute, Saitama University, Saitama, Japan, 338-8570
Keywords: Dosimetry (computational), ELF/LF, Work in Progress
Presented by: Hideyuki Okano
This study focuses on computational estimation of the induced electric fields (EFs) and eddy
currents induced by an AC electromagnetic field (EMF; 50 Hz, Bmax 180 mT) in human subjects.
The induced EFs were numerically calculated using the finite-element method. The left forearm was
exposed to heterogeneous EMF generated by double coils. For the EF intensity at the skin surface
as well as in muscle and tendons, the simulation results indicated that the highest EF was located in
the regions in and around the first coil. The EF intensity at the artery was higher near the second
coil than the first coil. The present numerical analysis may contribute to the understanding of
physiological mechanisms of AC EMFs for hemodynamic responses.
Introduction
AC electromagnetic field (EMF) therapy has been used to relieve chronic pain, muscle stiffness,
muscle fatigue and so on, for more than 30 years in Japan, since approved by Japanese Ministry of
Health, Labour and Welfare for improvement of muscle stiffness and blood circulation in the EMF-
exposed area in which the peak magnetic flux density Bmax of the surface area should be ranging
between 35 and 180 mT. However, physiological mechanisms are not clear enough to support the
effectiveness of this therapeutic approach, especially due to EMF-enhanced blood circulation and
recovery of muscle fatigue and pain [1-6]. This study focuses on the computational estimation of the
induced electric fields and eddy currents induced by an AC EMF exposure (50 Hz, Bmax180 mT) in
healthy human subjects.
Materials and Methods
AC EMF Exposure Device
An AC EMF exposure device (Soken MS), which was manufactured by Soken (Toride, Ibaraki,
Japan), was utilized for research purpose. Two separate electromagnetic coils (the coil radius, 5 cm;
the distance between the center of the coils, 16 cm) are set horizontally inside the EMF exposure
device and the value of the Bmax is 180 mT on the surface of the EMF exposure device above the
center of the coils. The spatial distribution of the Bmax from the surface of the EMF exposure device
is shown in Fig. 1. The magnetic flux density values of AC EMF decrease exponentially with
distance. The estimated Bmax value in ulnar artery is approximately 13 mT in which the distance
from the surface Bmax 180 mT of the EMF exposure device is approximately 3 cm.
Measurement of AC EMF
The measurement of root mean square (rms) magnetic flux density values (a) was made by means
of a Hall probe magnetometer in AC mode (AC/DC Magnetometer, AlphaLab, Salt Lake City, UT,
USA). Here, the Bmax values (b) of AC EMF were calculated using the following the equation:
b = √2 a (1)
The room temperature and the temperature in the surface of the EMF exposure device during the
EMF exposure period (for up to 15 min) was maintained at 25±0.5°C. The relative humidity was
controlled from 50±10%.

300
Electric Field (EF) Modelling and Simulation
The left forearm was exposed to EMF generated by double coils, in which the first coil was placed
near the elbow and the second coil was placed near the wrist. Computational methods for
determining the EMF-induced electric fields (EFs) in heterogeneous anatomical models are based
on the quasi-static approximation of Maxwell’s equations. Under the quasi-static approximation, the
induced EFs can be represented in terms of the electric scalar potential which satisfies an elliptic
partial-differential equation. The EFs were numerically calculated using the finite-element method
(FEM) [7-8] with first-order cubical elements [9]. Based on the measurement of AC EMF, the EMF
exposure coil was modeled using a collection of thin circular current loops. For modeling the
anatomy of the hand and forearm, a voxel model with a resolution of 0.5 mm was constructed based
on the BodyParts3D database (http://lifesciencedb.jp/bp3d/). The voxel model included detailed
models of vasculature and consisted of ten different tissues, which are listed in Table 1. To reduce
numerical artefacts, the voxels with the highest 0.01% induced EF/current density were excluded
from the analysis.
Results and Discussions
The values of eddy currents (current densities) in various tissues were obtained from the
multiplication of the induced EFs by the electrical conductivities. Here the values of electrical
conductivities in various tissues have been described and reviewed in detail elsewhere [10-12].
Their conductivity values are listed in Table 1.
Table 1. Electric conductivities used in the simulations

Tissue Conductivity (S/m)

Skin 0.10

Fat 0.04

Muscle 0.35

Tendon 0.30

Ligament 0.30

Artery 0.70

Vein 0.70

Blood 0.70

Cancellous bone 0.07

Cortical bone 0.02

Fig. 2 shows the distribution of the induced EFs in-situ. With regard to the intensity of the electric
field at the skin surface as well as in muscle and tendons, the simulation results indicated that the
highest EF was located in the regions in and around the first coil (Fig. 2 A, B). Because the arm is
thicker near the first coil, larger currents are induced. As for the EF intensity at the artery, the EF
intensity was higher near the second coil than the first coil (Fig. 2 C). For this reason, the artery near
the second coil is closer to the edge of the coil, where the largest currents are induced. The
streamlines of the induced eddy currents are drawn at randomly selected locations (Fig. 2 D).
According to the Faraday law, the induced current is proportional both to the time derivative of the
magnetic field and the surface area.
Table 2 indicates the maximum values of induced EFs and current densities in various tissues. The

301
current density depends on the location. It appears that one of veins (109 mA/m2) is in a region with
a higher current density than the arteries (69 mA/m2). It has been reviewed that for extremely low
frequency (ELF) frequency range, few systematic studies have been carried out to define the
threshold field characteristics for producing significant perturbations of biological functions [13].
Nevertheless, available evidence suggests that ELF-EMFs must induce current densities in tissues
and extracellular fluids that exceed 10 mA/m2, in order to produce significant alterations [13]. In
particular, between 10 and 100 mA/m2 (above 5-50 mT at 50/60 Hz) there are well established
effects, including visual and nervous system effects [13]. In addition, facilitation of bone fracture
reunion has been reported [13]. In in vitro studies, various phenomena have been reported in the
1-10 mA/m2 range, but their health significance has not been determined [13]. It has been proposed
that therapeutic applications of magnetic fields make use of this range [13].
Table 2. Electric fields in tissues induced by an AC electromagnetic field

Tissue Max electric field (V/m) Max current density (A/m2)

Skin 0.4042 0.0404

Fat 0.4357 0.0174

Muscle and tendon 0.2178 0.0762

Artery 0.0991 0.0693

Vein 0.1552 0.1087

Bone 0.3371 0.0092

Conclusions
We estimated the electric fields and eddy currents induced by an AC EMF (50 Hz, Bmax 180 mT) in
human subjects. The left forearm was exposed to heterogenerous EMF generated by double coils,
in which the first coil was placed near the elbow and the second coil was placed near the wrist. With
regard to the EF intensity at the skin surface as well as in muscle and tendons, the simulation
results indicated that the highest EF was located in the regions in and around the first coil.
Considering the EF intensity at the artery, the EF intensity was higher near the second coil than the
first coil. The present numerical analysis may contribute to the understanding of physiological
mechanisms of AC EMFs for hemodynamic responses.
References
[1] McKay JC, Prato FS, Thomas AW. A literature review: the effects of magnetic field exposure on
blood flow and blood vessels in the microvasculature. Bioelectromagnetics, 28:81-98. 2007.
[2] Ohkubo C, Okano H, Masuda H, Ushiyama A. EMF effects on microcirculatory system. The
Environmentalist, 27:395-402. 2007.
[3] McNamee DA, Corbacio M, Weller JK, Brown S, Prato FS, Thomas AW, Legros AG. The
cardiovascular response to an acute 1800-microT, 60-Hz magnetic field exposure in humans. Int
Arch Occup Environ Health, 83:441-454. 2010.
[4] McNamee DA, Corbacio M, Weller JK, Brown S, Stodilka RZ, Prato FS, Bureau Y, Thomas AW,
Legros AG. The response of the human circulatory system to an acute 200-μT, 60-Hz magnetic field
exposure. Int Arch Occup Environ Health, 84:267-277. 2011.
[5] Ueno S, Okano H. Static, low frequency and pulsed magnetic fields in biological systems. In:
"Electromagnetic Fields in Biological Systems." James C. Lin (ed). Boca Raton: CRC Press.
115-196. 2011.

302
[6] Ohkubo C, Okano H. Magnetic field influences on the microcirculation. In: "Electromagnetic
Fields in Biology and Medicine." Marko S. Markov (ed). Boca Raton: CRC Press. 103-128. 2015.
[7] Wang W, Eisenberg S. A three-dimensional finite element method for computing magnetically
induced currents in tissues. IEEE Trans Magn, 30:5015-5023. 1994.
[8] Wagner T, Fregni F, Eden U, Ramos-Estebanez C, Grodzinsky A, Zahn M, Pascual-Leone A.
Transcranial magnetic stimulation and stroke: a computer-based human model study. Neuroimage,
30:857-870. 2006.
[9] Laakso I, Hirata A. Fast multigrid-based computation of the induced electric field for transcranial
magnetic stimulation. Phys Med Biol, 57:7753-7765. 2012.
[10] Gustrau F, Bahr A, Rittwenger M, Goltz S, Eggert S. Simulation of induced current densities in
the human body at industrial induction heating frequencies. IEEE Trans Electromagn Compat, 41:
480-486. 1999.
[11] Li Y, Hand JW, Wills T, Hajnal JV. Numerically-simulated induced electric field and current
density within a human model located close to a z-gradient coil. J Magn Reson Imaging,
26:1286-1295. 2007.
[12] Laakso I, Tanaka S, Mikkonen M, Koyama S, Sadato N, Hirata A. Electric fields of motor and
frontal tDCS in a standard brain space: A computer simulation study. Neuroimage, 137:140-151.
2016.
[13] World Health Organization (WHO). In: "Environmental Health Criteria 69: Magnetic Fields."
WHO, Geneva, Switzerland, ISBN 92-4-154269-1, 1987.
[14] Okano H, Fujimura A, Ishiwatari H, Watanuki K. The physiological influence of alternating
current electromagnetic field exposure on human subjects. In: "IEEE International Conference on
Systems, Man, and Cybernetics (SMC)." ISBN 9781538616451 2442-2447. 2017.
Figures

Figure 1. Spatial distribution of the Bmax values.

303
 Figure 2. Induced electric fields (V/m). View of the left arm from the posterior and medial
direction. Circles indicate the centers of the magnetic coils. A. Electric fields at a depth of 2
mm below the skin surface. B. Electric fields in muscles and tendons. C. Electric fields in the
trunk of the ulnar artery. D. Streamlines of the induced eddy currents. The ulnar artery is
highlighten in red.

S15-4 [10:15]
Extended numerical modelling of cortical electrostimulation
Jose Gomez-Tames1 & Akimasa Hirata1
1Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology, Nagoya, Japan,
466-8555
Keywords: Dosimetry (computational), ELF/LF, Review, Commentary, Recommendation, Evaluation
Presented by: Jose Gomez-Tames
Computational models have been used to understand local brain stimulation. However, no a
computational model revealing the stimulation process from the cortex to MNs has not yet been
proposed. The aim of the current study was to develop a corticomotoneuronal (CMN) model to
investigate intraoperative stimulation during surgery
Introduction
Direct cortical stimulation is applied to identify and monitor structures of the primary motor cortex
and pyramidal tract during intraoperative mapping/monitoring. This prevents neurological
deterioration during brain tumour surgery, while maximizing resection to increase the survival rate
[1]. During cortical stimulation, descending action potentials travel through the corticospinal axons,
which synapse with the anterior horn, where cell bodies of alpha MNs (motor neurons) are located.
An excitatory postsynaptic potential is generated in the postsynaptic membrane, which can elicit an
action potential that travels to the endplate, producing a motor evoked potentials. Thus, in addition
to acting on the number of activated pyramidal neurons at the cortical level, stimulation

304
conditions/parameters also have an effect at the synaptic level, where MN responses are enhanced
by temporal/spatial summation and neural facilitation. From a physics viewpoint, several
computational studies have estimated the effect of stimulation parameters on the electric field
distribution in nervous tissue [2]–[4]. These models are pivotal to understanding the location of
stimulation hotspots in the brain. However, the effects of the interaction between frequency and
pulses delivered to the brain cannot be addressed by those models without considering the
integration process at cortical and synaptic levels. The current study newly developed a hybrid
computational model of corticomotoneuronal (CMN) response to describe the activation process
from cortical stimulation to motor evoked potential.
Method
A three-step computational model of direct CMN stimulation was designed to explain the effect of
externally applied electric fields on the temporal-spatial activation of target neurons, leading to motor
evoked potentials (Fig. 1A). In the first step, the electric potential generated by direct cortical
stimulation was computed using a volume conductor model. In the second step, the induced electric
potential obtained by the volume conductor model was coupled with a compartment model of a
myelinated pyramidal neuron axon. Finally, the descending action potential through the corticospinal
tract depolarizes the postsynaptic MN, which was implemented using a synaptic conductance
model.
Results
Figure 1B shows that the proposed CMN model agrees with experimental data [5], in which a
reduction of the stimulation threshold is due to the effect of the stimulating frequency and number of
pulses delivered to the motor area. Taniguchi et al. [6] demonstrated that an anodal rectangular train
of 3–5 pulses, with frequencies between 200–700 Hz, applied to the primary motor cortex,
generated motor evoked potentials. The temporal summation of frequencies lower than 100 Hz was
very small. For frequencies larger than 600 Hz, the refractory period limited repetitive generation of
action potentials.
Conclusion
For the first time, the proposed model can investigate the interaction between frequency and
number of pulses by integrating cortical stimulation and corticospinal neuron-, synaptic-, and motor
neuron-activation into one system.
References
[1] T. Saito, Y. Muragaki, T. Maruyama, M. Tamura, M. Nitta, and Y. Okada, “Intraoperative functional
mapping and monitoring during glioma surgery.,” Neurol. Med. Chir. (Tokyo)., vol. 55, no. 1, pp.
1–13, 2015.
[2] I. Laakso, S. Tanaka, S. Koyama, V. De Santis, and A. Hirata, “Inter-subject Variability in Electric
Fields of Motor Cortical tDCS,” Brain Stimul., vol. 8, no. 5, pp. 906–913, 2015.
[3] A. Nummenmaa, J. A. McNab, P. Savadjiev, Y. Okada, M. S. Hämäläinen, R. Wang, L. L. Wald,
A. Pascual-Leone, V. J. Wedeen, and T. Raij, “Targeting of White Matter Tracts with Transcranial
Magnetic Stimulation,” Brain Stimul., vol. 7, no. 1, pp. 80–84, Jan. 2014.
[4] M. Parazzini, S. Fiocchi, I. Liorni, A. Priori, and P. Ravazzani, “Computational modeling of
transcranial direct current stimulation in the child brain: implications for the treatment of refractory
childhood focal epilepsy,” Int. J. Neural Syst., vol. 24, no. 2, p. 1430006, Mar. 2014.
[5] J. C. Lilly, G. M. Austin, and W. W. Chambers, “Threshold movements produced by excitation of
cerebral cortex and efferent fibers with some parametric regions of rectangular current pulses (cats
and monkeys).,” J. Neurophysiol., vol. 15, no. 4, pp. 319–41, Jul. 1952.
[6] M. Taniguchi, C. Cedzich, and J. Schramm, “Modification of cortical stimulation for motor evoked
potentials under general anesthesia: technical description.,” Neurosurgery, vol. 32, no. 2, pp.
219–26, Feb. 1993.

305
Figures

Figure 1. (A) Proposed corticomotoneuronal (CMN) model for cortical stimulation. (B) Cortical
maps of excitation thresholds using monopolar stimulation for a pulse duration of 200 μs. (C)
Excitation thresholds using variable frequencies and numbers of pulses (frequency refers to
repetition rate between pulses). PN: pyramidal neuron; MN: motor neuron.

S15-5 [10:30]
Patient semi-specific computational modeling of electromagnetic stimulation applied to
regenerative treatments in acute ischemic stroke
Micol Colella1, Francesca Camera1, Fioravante Capone2, Stefania Setti3, Ruggero Cadossi3, Francesca
Apollonio1, Vincenzo Di Lazzaro2 & Micaela Liberti1
1Department of Information Engineering, Electronics and Telecommunications, University of Rome "La
Sapienza", Rome, Italy
2Unit of Neurology Neurophysiology, Neurobiology, Department of Medicine, Università Campus Bio-Medico
di Roma, Rome, Italy
3IGEA Biophysics Laboratory, Carpi, Italy
Keywords: Dosimetry (computational), Pulsed, Completed (unpublished)
Presented by: Micaela Liberti
Specific pulsed electromagnetic field (PEMF) have been shown to be effective in producing changes
in the excitability of human cerebral cortex and to have neuroprotective potential on animal models

306
of brain ischemia and on in-vitro studies. For this reason, the interest of validating the PEMF
stimulation as non-invasive, safe and effective tool to promote recovery in acute ischemic stroke
patients is spreading. In such a context, an accurate dosimetry that can be patient semi-specific
and, so, describe the electrical and magnetic quantities induced in each treatment, is clearly a
powerful tool to understand brain magnetic stimulation and the related interaction mechanisms with
biological tissues.
Introduction
Ischemic stroke, is a leading cause of death and disability worldwide, bearing a deep impact on the
socioeconomic burden for healthcare. To date, thrombolysis is the only approved treatment
however, the time window is limited to the first hours after stroke, and thus there is an urgent need
of different therapies capable to reduce the catastrophic consequences of brain ischemia beyond
this short period of intervention.
Extremely low frequency magnetic fields (ELF-MF) could be an alternative neuroprotective approach
for ischemic stroke because preclinical studies have demonstrated their effects on the mechanisms
underlying ischemic damage. In-vitro studies have shown that ELF-MFs modify genes expression,
promote neurite outgrowth [1] and, in particular, ELF-MF delivered in pulsed mode can selectively
modulate glutamate [2] and adenosine [3] receptors. Moreover, pulsed ELF-MFs applied on animal
models of brain ischemia have shown to have neuroprotective potential [4-6].
Besides being effective in attenuating ischemic edema, the signal used in [5] is also currently used
in orthopedic practice and is able to produce changes in the excitability of human cerebral cortex [7].
Recently, I-NIC an open-label, one arm, dose-escalation, exploratory study evaluated the safety and
tolerability of ELF-MF in patients with acute ischemic stroke [8new]. Within 48 hours from the stroke
onset, patients started ELF-MF treatment, daily for 5 consecutive days. Clinical follow-up lasted 12
months. Brain MRI was performed before and 1 month after the treatment. Six patients were
stimulated, three for 45 min/day and three for 120 min/day. None of them reported adverse events.
Clinical conditions improved in all the patients. Lesion size was reduced in one patient stimulated for
45 minutes and in all the patients stimulated for 120 minutes. Magnetic field intensity within the
ischemic lesion was above 1 mT, the minimum value able to trigger a biological effect in preclinical
studies. Such a pilot study demonstrates that ELF-MF are safe and tolerable in acute stroke
patients. A prospective, randomized, placebo-controlled, double-blind study is ongoing with 120
minutes treatments in order to clarify whether ELF-MFs could represent a potential therapeutic
approach
Other studies have shown that, however, not all ELF-MF pulsed signals and intensities are effective
in elicit such effects, suggesting that an accurate dosimetry, i.e., the estimation of the magnetic field
B and the other electrical quantities (E and J) induced in the brain for considered pulsed
electromagnetic fields (PEMF) is a key issue to understand brain magnetic stimulation postulate
thresholds and study the related interaction mechanisms.
The aim of this work is to perform an accurate dosimetric analysis of the three patients subjected to
a 120 minutes treatment by taking advantage of a patient semi-specific dosimetric model, that
combines the generalized human computational models (Virtual Population 1.0 [9], ZMT AG) and, at
the same time, using the geometric information of the ischemic area obtained from the magnetic
resonance imaging (MRI) data of the patients undergone to the treatment. In this way, we aim to
correlate the amount of the computed magnetic and electrical quantities induced by the stimulation
and the effectiveness of the treatment.
Models and Methods
Simulative software and human head model
The software used for the dosimetric analysis is Sim4Life v.3.4 (ZMT, Zurich MedTech AG), that is a
simulation platform that can deal with medical image data obtained from MRI; in particular, with this
software it is possible to convert the typical voxeled data obtained from MRI in volumes and
smoothed surfaces and assign electric properties to them.

307
The human head model was obtained from Duke of the ViP (v.1.0 [5], ZMT AG); we selected a
subdomain of 164x238x178 mm3, in order to include all the head structures.
The ischemic model
The geometric model of the ischemic tissue was obtained directly from the images of the magnetic
resonance of the patients undergone to the treatment. In these images, for each voxel, a number
between 0 and 255 (one byte) encodes for the gray intensity, i.e. for the kind of tissue. The ischemic
area is hyperintense, i.e. brighter that the healthy tissues; in particular, for our patient images, all the
voxels that overcome an established threshold of 140 have been considered ischemic tissue.
In order to extract the volume corresponding to the ischemia, the MRI images were all imported in
MATLAB, where, for each slice, a boolean equivalent was generated, assigning "0" to all the voxels
represented by a number equal or minor than 140, and "1" to all the ones represented by a number
major than 140; in this way, we obtain an equivalent set of data, but neglecting all the information
related to all the tissues different from the ischemia.
All these new slices were imported in Sim4Life v.1.2 and processed in order to generate the
surfaces of the ischemia. Since the resolution of these data, especially in the longitudinal axis, is
higher that the head model one, we chose to apply a smoothing function with an iterative algorithm
that makes an implicit interpolation in order to adapt the two different grids.
Once that the ischemic volume has been obtained, it has been placed, in Sim4Life environment, in
Duke's head model; the correct positioning has been achieved looking at the MRI images and
estimating the distance of the ischemia from the skull and other brain structures.
We assign to the ischemic tissue the electric properties of the edema [10], i.e. s=1.7 S/m and the
same er of the gray matter.
The same procedure has been followed based on the MRI images of the ischemic lesion after 30
days, in order to follow the evolution of the model towards the field exposure.
The coil model and simulations parameters
In the I-NIC clinical study, the medical device of IGEA company is used. This device is constituted
by a generator of low frequency pulsed electromagnetic field which generates a current pulse signal
and delivers this signal to an external coil, generating a pulsed electromagnetic field with specific
characteristics [7].
The pulsed signal has the following temporal characteristics:
duration of the active phase of the signal: 1.3 ± 0.1 ms;
repetition frequency: 75 Hz (which is equivalent to a repetition time period between two
successive pulses of 13.3 ms);
amplitude of the peak value: 1.05 A.
amplitude of the peak of magnetic field intensity: about 2 mT.
Moreover, a "helmet" is placed on the head of the patient and allows to hold in position the coil (by
means of a Velcro strap) during the treatment period.
The coil used in the simulation environment is a single turn rectangular coil of 14x10.6 cm that is
able to reproduce the same magnetic field of the experimental set up (data not shown).
We chose a stimulation current amplitude of 240 A because it is able to generate in the center of the
coil a magnetic field intensity of about 2 mT.
The single turn has been then warped in order to simulate the coil properties of being adaptable to
the head shape; this warping has been obtained computing the mean head curvature in the coronal
plane. The coil has been placed then as close as possible to the head and in order to have the
ischemic volume on the coil axis. Fig 1 shows the final geometry of the model and all the steps.
Results
Since the presence of the head doesn't alter the distribution of B, the tissues experience a magnetic

308
field that depends only on the distance from the coil.
In figure 2 it is possible to observe the B field induced on the three patients streamlines in the whole
volume are shown and B field in the ischemic volumes, in the inset the post treatment volume is
superposed, these data suggest a stronger volume reduction for B field intensities greater than 1.5
mT.
For a deeper analysis the volumes are analysed on slices in the coordinate plane, Fig. 3 shows
results on a xy plane, a clear reduction of pre-treatment volume with respect to post treatment one is
evident for B fields greater than 1.5 mT and J field greater than 10 mA/m2.
These results could be further deepen in order to confirm such thresholds for the three patients.
Conclusions
This work has been made possible by the collaboration with IGEA company, that developed the
exposure system and patented the pulsed signal. This system has been set up with the purpose of
reaching a peak value of the magnetic field of about 2 mT; this intensity, in fact, has shown to have
effects on the biological systems even at molecular level, inducing changing on the binding of
membrane receptors.
Simulations showed that tissues in the brain and in particular the ischemic volume experience a
magnetic field comparable to the one for which biological effects have been attested. Thus, it will be
possible to study a B-direct effect on the unhealthy tissue responsible for the regenerative effect
experimentally observed.
It has been a patient semi-specific dosimetric model, using the geometric information of the
ischemic area obtained from the magnetic resonance imaging (MRI) data of the patient that have
been then placed inside a standard computational model. This methodology can be used for any
patient undergone to the treatment and, so, can be considered a powerful tool to correlate the
amount of the computed magnetic and electrical quantities induced by the stimulation and the
effectiveness of the treatment in order to confirm the thresholds on 1.5 mT for B field and 10 mA/m2
suggested by these preliminary results.
References
[1] McFarlane et al., Bioelectrochemistry, 2000, 52(1): 23-8
[2] Wieraszko, Bioelectromagnetics, 2004, 25(7): 537-44
[3] Varani et al., Bioelectromagnetics, 2012, 33(4): 279-87
[4] Capone et al., Austin J Cerebrovasc Dis & Stroke, 2012, 1(1): id1002
[5] Grant et al., Bioelectromagnetics, 1994, 15(3): 205-16
[6] Pena-Philippides et al., Translational Stroke Research, 2014, 5(4): 491-500
[7] Capone et al., Journal of Neural Transmission, 2009, 116(3): 257-265
[8] Capone et al. Scientific ReportsOpen AccessVolume 7, Issue 1, 1 December 2017, Article
number 12145
[9] Christ et al., Physics in medicine and biology, 2009, 55(2): N23
[10] Kent et al., Journal of Neural Engineering, 2014, 11(4): 046010
Figures

309
Figure 1. Sketch of workflow, (a) step 1: solenoid model (experimentally validated), (b) step 2,
semi-specific model based on reconstructed ischemic volume for each patient , (c) step 3,
dosimetric evaluations

Figure 2. B field distribution at the time peak in the pre treatment ischemic volumes and
streamline view of B along “coil frontal plane”. In the inset the superposition of pre and post
treatment lesions.

310
Figure 3. B field (b) and J field (c) at the peak for patient A in a plane xy (z= 10 cm) in the
ischemic volume, post treatment volume superposed

311
S15-6 [10:45]
Development of the 0.16THz in vivo exposure system for evaluating acute ocular damages
with the high power gyrotron electromagnetic wave source
Yukihisa Suzuki1, Masami Kojima2, Takafumi Tasaki2, Yoshinori Tatematsu3, Masafumi Fukunari3, Maya
Mizuno4, Kensuke Sasaki4, Soichi Watanabe4, Masao Taki1, Masahiko Tani3 & Hiroshi Sasaki2
1Tokyo Metropolitan University, Tokyo, Japan, 192-0397
2Kanazawa Medical University, Ishikawa, Japan, 920-0265
3University of Fukui, Fukui, Japan, 910-8507
4National Institute of Information and Communications Technology, Tokyo, Japan, 184-8795
Keywords: In vivo, THz, Work in Progress
Presented by: Yukihisa Suzuki
In this study, a new in vivo exposure system for 0.16 THz exposure has developed to investigate the
threshold level of power density inducing thermal ocular damage. To realize this system, the
gyrotron FU CW GV was used for the 0.16THz high power EM wave source. When the irradiated
power density was 500mW/cm2, the maximum temperature elevation at 6min from the onset of
exposure became 26.4ºC and 19.8ºC for an agar phantom and a cornea of rabbit eye, respectively.
Consequently, we have achieved the sufficient temperature elevation to observe acute cornea
damages due to the thermal effect at 0.16THz
Introduction
Electromagnetic (EM) waves with millimeter wave and THz frequency region are expected to be
used for advanced technologies, for example high-speed and broadband communications, wireless
power transmission, remote sensing, and so on. In near future, these technologies will used in our
daily life. Therefore, it is required to obtain knowledge about the health effect caused by EM wave
exposure with these frequencies. There are several in vivo studies relating to non-thermal effects, in
these frequency region (for example [1,2]). However, there are no in vivo studies to investigate
acute thermal effect, because it is difficult to obtain THz electromagnetic wave sources which
generate sufficiently large power to observe thermal acute effects. Therefore, the purpose of this
study is to establish a new in vivo exposure system to investigate the threshold level of power
density inducing thermal ocular damage and to evaluate composed system.

Materials and Methods

In this study, we tried to apply a gyrotron as a high power electromagnetic wave source. Figure 1
shows a block diagram of developed 0.16THz in vivo exposure system. THz EM waves with
0.16THz continuous wave was provided by the gyrotron FU CW GV [3], and guided along the
φ28mm over-sized circular waveguide. A taper transition attached to the end of the over-sized
waveguide to connect the WR-6 wave guide system. EM wave power level introduced into WR-6
waveguide system was regulated by P-I-N attenuator which controlled attenuation ratio by applying
DC current. Power level was measured by power sensor (PM5, Virginia Diodes, Inc., USA) by
connecting to the directional coupler. The lens antenna with 200 mm focal point and 6mm FWHM
beam waist was used to irradiate THz EM wave to a cornea. Values of power density irradiated to
the surface of cornea is evaluated by averaging the distribution of power densities inside φ13mm
area centered at the apex of cornea. By obeying that principle, the power density of 0.743mW/cm2
was obtained, when the input power into the lens antenna was 1.00mW. For the evaluation of the
exposure system, temperature elevation during 6min exposure is measured at the surface of an
agar phantom and a cornea of rabbit eye. The surface of temperature is measured by the infrared
thermography camera (CPA-T620, FLIR, USA).

Results

312
We achieved the exposure power density up to 600mW/cm2 (antenna input power: 808mW) with the
developed exposure system. Figure 2 indicates the dependence of temperature elevation on the
incident power density for an agar phantom and a cornea. When the irradiated power density was
500mW/cm2, the maximum temperature elevation at 6min from the onset of exposure became
26.4ºC and 19.8ºC for an agar phantom and a cornea of rabbit eye, respectively. Here the
conditions of the room temperature and relative humidity were 24.0±1.0ºC and 60±10%,
respectively. It was found that we have achieved the sufficient temperature elevation to observe
acute cornea damages due to the thermal effect by using developed exposure system with gyrotron
EM wave source at 0.16THz.
Acknowledgement
This work was supported by the Committee to Promote Research on the Potential Biological Effects
of Electromagnetic Fields, Ministry of Internal Affairs and Communications, Japan.
References
[1] V. F. Kirichuk, et al., “Effect of High Power Terahertz Irradiation on Platelet Aggregation and
Behavioral Reactions of Albino Rats” Bull. Exp. Biol. Med., Vol. 48, pp. 746-749, 2009.
[2] N. P. Bondar, et al., “Behavioral effect of terahertz waves in male mice” Bull. Exp. Biol. Med., Vol.
145, pp. 401–405, 2008.
[3] Y. Tatematsu, et al., “Development of the Multifrequency Gyrotron FU CW GV with Gaussian
Beam Output”, Int. J. Infrared Milli. Waves Vol. 38, pp. 697-708, 2015.
Figures

Figure 1. A block diagram of the developed 0.16THz exposure system

313
Figure 2. The dependence of temperature elevation on the incident power density for an agar
phantom and a cornea

314
 Session: S16
Public Policy II
Friday June 29, 2018 • 09:30 - 11:00
Europa D
Chairs: Azadeh Peyman & Theodoros Samaras

S16-1 [09:30]
Understanding and overcoming ‘Risk Communication Traps’
Frederik Freudenstein1, 2, 3, 4, Peter Wiedemann1, 2, 4, Rodney Croft1, 2, 3, 4 & Sarah Loughran1, 2, 3, 4
1Australian Centre for Electromagnetic Bioeffects Research, Wollongong, Australia, 2500
2University of Wollongong, School of Psychology, Wollongong, Australia, 2500
3Illawarra Health and Medical Research Institute, Wollongong, Australia, 2500
4Centre for Population Health Research on Electromagnetic Energy, Monash Uni, Melbourne, Australia, 3800
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Frederik Freudenstein
In the presented research we analyse the effects of a special class of problems, called ‘risk
communication traps’. These traps hinder the proper understanding of risk assessments, leading to
misunderstandings of both the very existence of and magnitude of a risk. We focus on a set of five
traps related to RF EMF (radio frequency electromagnetic fields) risk assessments. As an example,
research on the ‘framing trap’ is presented. This trap becomes relevant to risk communication when
hazard identifications (being just the fist step in risk assessment) are misunderstood as risk
assessments. An online survey investigates wether this effects RF EMF risk perception and higher
confidence that cancer risks will occur from cell phone usage.
Introduction
When communicating risk issues the presented information should be correct, easy to access,
understandable and help people to make evidence-based decisions. Although this is the theory,
whenever risk-related information is communicated it is prone to create exaggerated fears or cause
misunderstandings among patients, public, politicians or any other receivers of the message. The
way in which scientific findings are understood and that risks are perceived is highly dependent on
the selection and structure of the information presented (Covey, 2007). Hence risk communication
faces the problem of ensuring that information communicated results in ‘appropriate’ risk perception.
Whenever this fails – be it unintentional or otherwise – we talk of the mechanism leading to such
failure as a ‘risk trap’.
Our RF-EMF (radio frequency electromagnetic fields) risk research focuses on a set of five traps
that might affect the understanding of risk information: (1) The selectivity trap, (2) the
operationalization trap, (3) the generalisation trap, (4) the comparison trap, and (5) the framing trap.
Without commenting on IARC’s (International Agency for Research on Cancer) work in general,
their press release from 2011 (IARC, 2011) can be used to describe these traps. In 2011 IARC
classified RF EMF as ‘possibly carcinogenic’. A decisive scientific basis for this categorization was
the Interphone Study from 2010, (Interphone Study Group, 2010) which reported a 40% increased
risk of glioma (brain cancer) for heavy cell phone users with a cumulative call time of more than
1640 hours in the last 10 years. The 40% risk increase found in this study was only for this particular
group of heavy users, and this fact was emphasized in the press release. This is an example of the
‘selectivity trap’, where highlighting some information while omitting other information may result in
biased views and interpretation of a risk assessment. Furthermore the IARC text provided
information regarding a relative cancer risk increase of 40 %. As lay people have difficulty
interpreting numeric and probabilistic risk expressions, using relative risks may lead to an
overestimation of risk (Wiedemann et al., 2014). This is an example of the‘operationalization trap’,
which deals with the importance of choosing the right parameters when communicating risk related

315
information. The title of the press release, ”IARC classifies RF EMFs as possibly carcinogenic to
humans”, poses further problems for risk perception. Even if clarified in the text, simplifications like
this may result in an overgeneralisation (the ‘generalization trap’), whereby all RF EMF emitting
devices are viewed as equally risky (e.g. cell phones, Wi-Fi, base stations). Also related to this is the
‘comparison trap’. The use of comparisons is a well-known strategy in risk communication (Johnson,
2002). However, in terms of IARC’s classification scheme, which can be used to provide
comparisons with other agents with the same classification (in this case ‘possibly carcinogenic’),
problems can again arise. This is because not all agents with same ‘hazard’ classification will be
associated with the same risk. However this is not communicated when such comparisons are made
which ultimately may influence RF EMF risk perception in different ways. Finally, there is the
‘framing trap’, which becomes relevant when hazard identifications are inappropriately viewed as
risk assessments. For example, readers may misunderstand IARC´s conclusions given that,
although IARC conducted a hazard assessment only, the press release used the word ‘risk’ several
times, and referred to the already-mentioned 40% increase of ‘risk’ for heavy users of cell phones.
However, IARC conducted a hazard identification, and not a risk assessment, i.e., not providing
actual information about the magnitude of risk as dose response relations and exposure analysis
are excluded.
Our program aims to conduct experimental research to determine how ‘risk communication traps’
may affect lay people’s intuitive risk perception, and through this to develop recommendations for
risk communication so as to avoid these risk traps. This presentation will focus on the first of these,
the ‘framing trap’, presenting data aimed at determining whether the framing trap leads to
misunderstanding and inappropriate risk perception.
Methods
An online survey was conducted in 2017/2018 with a total of 1026 participants from Australia; 830
respondents’ data remaining after quality control. A Randomized 1-factorial design with 3 levels was
used (for experimental setup see Table 1). After providing an introduction and background
information on RF EMF, an excerpt from the IARC press release from 2011 was presented to all
participants. Group 1 (G1) received additional information specifying that IARC is doing cancer risk
assessments. Group 2 (G2) received additional explicit information clarifying that IARC is doing
hazard identifications, and stressing the importance of the distinction between hazard and risk
assessment. Group 3 (G3) was used as a control group, where participants received IARC press
release only. The additional text modules provided to the different groups are shown in Table 2. A
manipulation check was also used to see whether the respondents interpreted the IARC press
release as a hazard identification or risk assessment.

To investigate differences between the treatment groups, three dependent variables were tested: (1)
Respondents risk perception of heavy mobile phone usage was measured on a 4-point Likert scale
using the following question, which including a pictorial description of the scenario: “How dangerous
do you consider using a cell phone (like the person pictured above) for 30 min per day over a 10-
year period?”, from 1 = Not dangerous, to 4 = very dangerous. The picture showed a person using
the mobile phone for a voice call at the side of the head; (2) An item created to measure general
EMF risk perception (“How concerned are you about the potential health risks of electromagnetic
fields in general”, from 1 = Not at all concerned, to 4 = very concerned); (3) a question about the
certainty with which they believed that cancer risk was increased when using mobile phones: “How
confident are you that RF EMF exposure from cell phones increases cancer risks?”, on a scale from
0 -100, with 100 = I am 100% confident that RF EMF exposure from cell phones does increase
cancer risks. For G1, who received false information about the purpose of the IARC evaluation, and
G3, who received no additional information about the press release, a short debriefing was then
provided to clarify that the IARC assessment was a hazard identification and not a risk assessment.
Results
An ANOVA showed no differences between the groups in terms of risk perception of heavy mobile

316
phone usage (F(2, 827) = 0.085, p= 0.919), general risk perception (F(2, 827) = 0.858, p = 0.424),
and confidence that mobile phone usage causes cancer (F(2, 827) = 0.255, p= 0.777) (group sizes:
nG1 = 281; nG2 = 280; nG3 = 269).
When looking at the distribution of respondents’ understanding of IARC’s role in risk assessments,
the majority in each group was convinced that this institution provides details about the magnitude of
a cancer risk (i.e. that it was a risk assessment): G1 = 87.5% (246 out of 281), G2 = 73.2% (205 out
of 280) and G3 = 84.4% (227 out of 269), and across groups a total of 81.7% (678 from 830
respondents) Based on this, participants were then re-classified according to whether they believed
that the IARC cancer assessment was a hazard identification or risk assessment (regardless of the
additional information they received during the experimental). This lead to two subgroups: Group 4
(G4, nG4 = 678; risk understanding) who believed IARC was doing risk assessment, and Group 5
(G5, nG5 = 152; hazard understanding) who believed IARC was doing hazard identification.
A t-test with the two new subgroups showed that people framing the IARC text as risk assessment
had a significant higher risk perception of heavy mobile phone usage, t(828) = 4.63, p < 0.001 and
significantly higher general EMF risk perception t(828) = 5.14, p < 0.001, as well as higher
confidence that EMF causes cancer t(828) = 3.64, p < 0.001, when compared to those who
understood the text as hazard identification (see Table 3).
Discussion
The current experimental study shows no effects of our experimental conditions. A possible
explanation for this could be that the additional texts modules used in the experiment might not have
been effective enough to reach the intended framing effects. However, an exploratory analysis
indicates that whether the IARC text is understood as a risk assessment or as hazard identification
has an impact on RF EMF risk perception and on the confidence that mobile phone usage causes
cancer.
Conclusion
Further research should test an effective presentation format and investigate ways to effectively
avoid the ‘framing trap’ in communication.
References
Covey J. (2007). A meta-analysis of the effects of presenting treatment benefits in different formats.
Medical Decision Making, 27:638–654.
IARC (2011). IARC Classifies Radiofrequency Electromagnetic Fields As Possibility Carcinogenic to
Humans, 2011. Available online: http://www.iarc.fr/en/media-centre/pr/2011/pdfs/pr208_E.pdf
Interphone Study Group (2010). Brain tumour risk in relation to mobile telephone use: results of the
INTERPHONE international case‐control study. Int J Epidemiol, 39(3):675‐694.
Johnson B. (2002). Stability and inoculation of risk comparisons‘ effects under conflict: Replicating
and extending the „asbestos jury“ study by Slovic et al. Risk Analysis, 22(4):777-788.
Wiedemann PM, Boerner F, Repacholi M. (2014). Do people understand IARC’s 2B cate- gorization
of RF fields from cell phones? Bioelectromagnetics, 35(5):373–378.
Figures

317
Figure 1. Overview experimental setup. RPgeneral = general risk perception, RPMP = risk
perception mobile phones, Confidence cancer risk = Confidence that mobile phone usage
causes cancer.

Figure 2. Additional text modules provided to respondents after press release in the different
groups.

318
 Figure 3. t-test for the depended variables ’risk perception of mobile phones’, ‘general risk
perception’, and ’confidence that cell phone use increases cancer risks’, between the
respondents who believed that the IARC press release was a risk assessment (G4) and those
who believe that it was a hazard identification (G5). M = mean; SD = standard deviation; RP =
risk perception; Confidence cancer risk = Confidence that mobile phone usage causes cancer.

S16-2 [09:45]
Post-Normal Science and the management of uncertainty in bioelectromagnetic
controversies
Andrew Wood1, 2
1School of Health Sciences, Swinburne University of Technology, Melbourne, Australia, Vic 3122
2Biomedical Engineering, RMIT University, Melbourne, Australia, Vic 3083
Keywords: Public Health Policy, All Frequencies, Review, Commentary, Recommendation, Evaluation
Presented by: Andrew Wood
The incorporation of scientific findings into public policy continues to present significant challenges
in many areas of environmental protection, including bioelectromagnetics. At the same time public
trust in ‘experts’ has diminished and science faces what some regard as a crisis. Uncertainty is often
acknowledged but rarely dissected. This paper will suggest ways in which uncertainties can be
properly managed, with particular reference to the work of Funtowicz and others on ‘Post-Normal
Science’ and more recently ‘What is science’s crisis really about?’
Introduction
Despite over half a century of scientific research into the safety of Electric and Magnetic Fields
(EMFs: from static through to 300 GHz) safety levels and advice to policy makers has not changed
substantially. However, public concern over lack of adequate recognition of ‘low level’ effects
persists with ‘expert groups’ apparently divided (Sage 2012) (SCENIHR 2015) and the methods of
collation, analysis and comparison of evidence criticised. However, EMF is not alone in receiving
this type of criticism. Many issues, including climate change, genetically-modified crops, air-borne
pollutants etc. also face controversy. Having been involved in reviews of EMF-related scientific
studies both in Australia and as part of international panels, the author is aware of the difficulties in
producing balanced and at the same time insightful reviews. In relation to this, the work of Funtowicz
and Ravetz (1993) on the concept of ‘Post-Normal Science’ is useful and has been applied to a
range of environmental issues, including, as examples, mobile technology (Kastenhofer 2011)
(Spruijt et al. 2014). However, there is more to be said, particularly in the light of more recent

319
discussions on the notion of science in crisis (Saltelli and Funtowicz 2017).
Post-Normal Science
Post-Normal Science (PNS) deals with situations where uncertainty and/or decision stakes are high
(Figure 1). The intensity of these quantities is higher than would be found in conventional applied
science or professional consultancy. The decision stakes could involve conflicting purposes (e.g.
access to fast telecommunications versus the need to protect vulnerable individuals from possible
harm) and the uncertainties involve the distinctions between justified belief and opinion and on the
ethics of pursuing particular policies. More so than in professional consulting, uncertainty and
complexity needs to be explicitly addressed, to allow critical reflection and further, uncertainty is to
be managed. Part of this is the recognition that the core values of the expert leads to coalitions of
experts with similar core values and that stakeholders are people with strong personal (or corporate)
motivation to be part of interaction. The more recent work of Funtowicz focusses on reasons for a
perceived crisis in science: specifically, problems with reproducibility; its good governance and its
use or misuse for policy formation (Saltelli and Funtowicz 2017). These problems are identified as
being associated with the exponential growth in papers, a community with high ideals becoming
enslaved to objective measures and the myth of value-neutral nature of science, particularly where
political pressures exist, but are inadequately acknowledged. This paper also reflects on instances
of imprudent inferences from flawed work and the use of maths to intimidate adversaries. The rise of
predatory journals and the patchy nature of the peer-review system are also listed as contributing to
the perceived crisis. Symptoms include: scientists resorting to ‘questionable research practices’
including falsification of data (Fanelli 2009); desperate measures on the part of younger scientists
because of lack of funding and stable positions and poor supervision/results checking by senior
scientists because of time pressures. Remedies suggested by Saltelli and Funtowicz (2017) include:
moving to better quality metrics for scientific papers; avoiding speaking in the name of ‘science’;
adopting models of ‘extended participation’; working deliberatively within imperfections and having a
careful appreciation of uncertainty and ignorance.
Application to EMF
A number of studies have examined issues of uncertainty regarding scientific input to the EMF and
other debates, with some reference to PNS. A helpful analysis by Spruijt et al. (2014) of 267
publications (including some of EMF, but in passing) identifies 4 different styles of ‘experts’ namely:
regulatory advocate; engaged expert; humble scientist and noninterventionist. This suggests
improved ways in which experts should advise on complex issues including: democratizing science
(i.e. public participation and stakeholder dialogs); transparency in methods and assumptions;
professional attitude of humility and making different points of view within the expert community
explicit. Further Borraz (2011) suggests that in relation to EMF, the risk framework itself has tended
to amplify uncertainty with the reputation of decision-makers becoming a source of additional
uncertainty.
Discussion
The so-called ‘pluralist approach’, which broadens the types of knowledge, the providers of
knowledge and the procedures used to produce evidence, has a ready appeal, but it tends to treat
all inputs as equivalent. The recent book ‘The Death of Expertise’ (Nichols 2017) contends that ‘the
information age has helped fuel a surge of narcissistic and misguided intellectual egalitarianism that
has crippled debates on any number of issues’. Information freely available has ‘created an army of
ill-informed and angry citizens who denounce intellectual achievement and distrust experts’. Further,
Konig, Borsen, and Emmeche (2017) has noted that ‘Stringent Conflict of Interest (CoI) policies may
bar the most experienced & knowledgeable’. There is a need to recognise specialist input while at
the same time making the assessments accessible to stakeholders to allow measured comment.
Suggestions for moving forward
The weight of evidence approach is regarded as the standard method for assessing safety (EFSA
Scientific Committee 2017). However, if this consists merely of a ‘paper count’ of effect versus no-

320
effect, or the weighing of evidence is not well specified or made available for scrutiny its value is
limited. In addition, if there is pressure to meet deadlines across complex organisational structures
this can further compromise the process. Therefore, the setting of realistic deadlines, the method of
identification and management of large volumes of published studies and the weighing procedure
needs to be carefully addressed. There are additional analyses which can be undertaken to help
elucidate basic questions and these should be considered: examples are (Vijayalaxmi and Prihoda
2012), (Wood 2017), Figure 2. Greater availability of source data would be of assistance: there
seems to be a great reluctance of authors to do this, perhaps arising from a culture of over-
protection of reputation, in which exposure of trivial errors of analysis can lead to severe
consequences. The transparency of data from the recent NTP study is to be applauded in this
regard. Journal editorial policies should encourage having more supplementary files available, with,
in addition to data, the intermediate steps in complex calculations, etc. Due regard needs to be
given to recognising appropriate expertise, which will involve the careful declaration and
management of CoIs, which would normally involve active stakeholder participation. Regarding
uncertainty, this should be carefully catalogued and be available for scrutiny by stakeholders.
Sensitivity analyses should be carried out to determine the extent to which policy advice should
change allowing for the range of uncertainties identified. Uncertainties in the setting of inclusion
criteria for meta-analysis and in management of perceived bias also require appropriate recognition.
There are several examples of some of these measures having been implemented, which will be
discussed in the presentation.
References
AGNIR, 2012. (Advisory Group on Non-Ionising Radiation). Health Effects from Radiofrequency
Electromagntic Fields. Chilton, Didcot: Health Protection Agency.
Borraz, O. 2011. "From risk to the government of uncertainty: the case of mobile telephony." Journal
of Risk Research 14 (8):969-982.
EFSA Scientific Committee. 2017. "Guidance on the use of the weight of evidence approach in
scientific assessments." EFSA Journal 15 (8):e04971-n/a.
Fanelli, D. 2009. "How many scientists fabricate and falsify research? A systematic review and
meta-analysis of survey data." PLoS One 4 (5):e5738.
Funtowicz, S. O., and J. R. Ravetz. 1993. "Science for the Post-Normal Age." Futures 25
(7):739-755.
Kastenhofer, K. 2011. "Risk assessment of emerging technologies and Post-Normal Science."
Science, Technology and Human Values 36 (3):307-333.
Konig, N., T. Borsen, and C. Emmeche. 2017. "The ethos of post-normal science." Futures
91:12-24.
Nichols, T. 2017. The death of expertise: the campaign against established knowledge and why it
matters. New York, NY: Oxford University Press.
Sage, C. 2012. "Bioinitiative 2012: Summary for the Public." In, ed C Sage.
http://www.bioinitiative.org/report/wp-content/uploads/pdfs/BioInitiativeReport2012.pdf.
Saltelli, A., and S. Funtowicz. 2017. "What is science's crisis really about?" Futures 91:5-11.
SCENIHR. 2015. (Scientific Committee on Emerging and Newly Identified Health Risks). Opinion on
Potential health effects of exposure to electromagnetic fields (EMF). European Commission.
Spruijt, P., A. B. Knol, E. Vasileiadou, J. Devilee, E. Lebret, and A. C. Petersen. 2014. "Roles of
scientists as policy advisers on complex issues: A literature review." Environmental Science & Policy
40:16-25.
Vijayalaxmi, and T. J. Prihoda. 2012. "Genetic damage in human cells exposed to non-ionizing
radiofrequency fields: a meta-analysis of the data from 88 publications (1990-2011)." Mutat Res 749
(1-2):1-16.
Wood, A.W. 2017. "Possible low-level radiofrequency effects." In Non-ionizing Radiation Protection,
321
edited by A.W. Wood and K. Karipidis, 239-256. Hoboken, NJ: Wiley.
Figures

Figure 1. Domain of 'Post-normal Science' in relation to Applied Science and Professional

Consultancy: the decision stakes and systems uncertainties tend to be high. Diagram based
on Funtowicz and Ravez (1993).

322
 Figure 2. SAR values for studies reported in the AGNIR (2012) and SCENIHR (2015) reports:
From top: those reporting EMF 'no-effects' are on odd rows and 'effects' on even rows. Bars
indicate average (± SD). For a threshold behaviour, the average SAR for 'effect' would be
expected to be greater than 'no-effect'. Diagram based on Wood (2017).

S16-3 [10:00]
Quarterly newsletters synthesizing research for policy and society: an innovative approach
piloted by the interdisciplinary Swiss expert group on electromagnetic fields and non-
ionising radiation (BERENIS)
Stefan Dongus1, Peter Achermann2, Jürg Fröhlich3, Jürg Kesselring4, Meike Mevissen5, David
Schuermann6, Edith Steiner7 & Martin Röösli1
1Environmental Exposures and Health Unit, Swiss Tropical and Public Health Institute, Basel, Switzerland,
4002
2Institute of Pharmacology and Toxicology, University of Zurich, Zurich, Switzerland, 8057
3Fields at Work GmbH, Zurich, Switzerland, 8006
4Rehabilitation Centre, Clinics of Valens, Valens, Switzerland, 7317
5Vet-Pharmacology & Toxicology, Vetsuisse Faculty, University of Bern, Bern, Switzerland, 3012
6Department of Biomedicine, University of Basel, Basel, Switzerland, 4058
7Ärztinnen und Ärzte für Umweltschutz, Basel, Switzerland, 4019
Keywords: Public Health Policy, All Frequencies, Work in Progress
Presented by: Stefan Dongus
In 2014, the interdisciplinary Swiss expert group on electromagnetic fields and non-ionising radiation
(BERENIS) started to continuously assess newly published human experimental and

323
epidemiological studies, mammalian in vivo and in vitro studies, as well as exposure and dosimetry
studies. Based on defined criteria, BERENIS identifies publications most relevant for health risk.
During the 3-years pilot phase of BERENIS, ten quarterly newsletters have been published,
providing short summaries and evaluations of 55 publications. Topics of particular relevance
identified by BERENIS include 1) oxidative stress, 2) cryptochrome as EMF sensor, 3) cancer
promotion, and 4) neurodegeneration.
Background
In Switzerland, the Federal Office for the Environment (FOEN) is the government body responsible
for the protection of the public against non-ionising radiation (NIR) emitted by fixed installations. This
includes monitoring and assessing research on health effects, informing and updating the public but
also other stakeholders about the current state of research, which provides the basis for the ambient
limit values stated in the Swiss “Ordinance relating to Protection from Non-Ionising Radiation”.
The FOEN has nominated a consultative group of Swiss experts, which has commenced its work in
July 2014. The group is called “BERENIS”, based on the acronym of the respective German term.
The pilot phase of BERENIS lasted three years, until June 2017.
BERENIS organisational setup, mandate, and course of action
BERENIS consists of seven Swiss experts from various disciplines with scientific competence
regarding electromagnetic fields and NIR. This expert team is assessing human experimental and
epidemiological studies, in vivo and in vitro studies, as well as exposure and dosimetry studies.
While the scientists in the team are primarily responsible for the assessment of publications related
to their respective areas of expertise, the main role of the two physicians is to critically reflect and
complement the scientists’ assessments, and to bring in the view of affected patients. The experts
are complemented by two representatives of the FOEN, supporting the literature screening process,
transmitting specific requests of the government to the group, and acting as discussion partners.
BERENIS also invites and is open to observers from other ministries.
The BERENIS experts regularly screen the scientific literature to identify publications they consider
relevant for the protection of the public and the environment from potentially adverse effects, as well
as for the understanding of the interaction modes between NIR and cells or organisms, and provide
a detailed and critical assessment of these publications. The experts also give their opinion on
questions raised by the FOEN, or topics attracting interest of the general public, and advise the
FOEN on research topics of particular priority.
Quarterly, the scientific secretariat compiles a list of newly published literature, noncommittally
assigned to individual BERENIS members according to their expertise. Before a new list is shared
with the whole BERENIS group, the two physicians and a FOEN representative mark publications
they consider relevant for human health. The experts screen the complete list, and select the
publications relevant for health and scientific knowledge and provide a detailed assessment of them.
Defined criteria are used in the selection as well as the assessment process. The assessments are
presented and discussed during plenary meetings held on a quarterly basis. The results of the
concerted evaluation are published in quarterly newsletters, which are publicly accessible at the
FOEN’s official website:
https://www.bafu.admin.ch/bafu/en/home/topics/electrosmog/newsletter-of-the-swiss-expert-group-
on-electromagnetic-fields-a/beratende-expertengruppe-nis-berenis.html
The same website provides the defined criteria guiding the process of literature selection and
assessment, which have been presented at the BioEM conference 2016 (Dongus et al. 2016).
BERENIS also occasionally discusses relevant review articles, and provides short summaries
without evaluations in the newsletter. The target group for the newsletters are interested lay people
and authorities. The newsletters are available in English, German and French. Subscription to an
automatic email distribution list for newly published newsletters is also available.
Results and conclusions

324
During the 3-years pilot phase, 12 quarterly BERENIS meetings have been held. During this period,
BERENIS has identified 1267 new publications. Out of these, 132 have been considered to be of
particular relevance according to the selection criteria, and have thus been selected for assessment
and discussion. The respective experts evaluated the selected publications according to the
assessment criteria. The evaluations of all publications selected for a meeting were presented and
jointly discussed in that meeting. A mutual decision was then taken in regard to the inclusion and
evaluation in the newsletter. The attendance rate of experts at the meetings was more than 90%,
enabling an efficient work flow. Until June 2017, 10 newsletters have been assembled and
published, comprising short summaries and assessments of a total of 55 publications (Table 1).
Table 1: Publications presented and discussed in newsletters until June 2017
Extremely low Radiofrequency
Assessed frequency electromagnetic
Both (ELF-MF Intermediate
publications Total magnetic fields (RF-
and RF-EMF) frequency
(n=55) fields (ELF- EMF)
MF)

Experimental 10 4
16 2
cell studies

Epidemiological 5 7
13 1
studies

Experimental 7 7
animal 15 1
studies

Human 1
experimental 5 4
studies

Review 2 2
4
articles

Dosimetric 2
and
2
mechanistic
studies
A considerable number of high quality peer-reviewed publications combining mechanistic data with
descriptive effects have been assessed. Several experimental animal and cell studies
demonstrating effects were considered to be of particular interest, as they suggested possible
interaction mechanisms between non-ionising radiation and cells or organisms. These studies
focussed on developmental processes and reproduction, memory function and neurodegeneration,
and cancer promotion. An increase in oxidative stress stood out as particularly relevant because of
its involvement in the pathogenesis of various diseases including neurodegenerative diseases and
cancer. Cryptochrome or the actin in the cytoskeleton were identified as putative molecular sensors.
However, their potential implications for health remain ambiguous and need to be understood better.
New epidemiological and human experimental studies did not strongly indicate health effects,
although a few associations have been reported, for example potential effects on figural memory
performance, and brain activity during sleep. Monitoring of non-ionising radiation research thus
remains important to continuously inform stakeholders about the state of new research to eventually
tighten measures to protect the health of the population.
The monitoring approach of the BERENIS team over its 3-years pilot phase has proven to be

325
feasible, timely and relatively efficient, with a very low likelihood for missing studies with relevant
findings. The interdisciplinary group setup enables the synthesis of results from different study types
of various disciplines, thus leading to a better overall understanding of EMF health issues. The
BERENIS approach is providing an "early warning system", while efficiently transferring new
research insights into policy and society. BERENIS will continue its work for another period of four
years, which started in July 2017.
References
Dongus S, Achermann P, Fröhlich J, Kesselring J, Mevissen M, Schuermann D, Steiner E, Röösli M
(2016): The interdisciplinary Swiss expert group on electromagnetic fields and non-ionising radiation
(BERENIS) – results and conclusions after 18 months of work. Annual Meeting of the
Bioelectromagnetics Society (BEMS) and the European Bioelectromagnetics Association (EBEA),
BioEM2016, Ghent, Belgium, Jun 5 - 10, 2016.

S16-4 [10:15]
Revision of IEEE Standards C95.1-2005 and C95.6-2002
C. K. Chou1, Ron Petersen2, Kenneth Foster3, Akimasa Hirata4, Marvin Ziskin5, J. Patrick Reilly6, Richard
Tell7, Antonio Faraone8, B. Jon Klauenberg9, Robert Kavet10, Kevin Graf11, Robert Cleveland12, Artnarong
Thansandote13, Jerrold Bushberg14, William Bailey15, John Osepchuk16, Alexandre Legros17, Kenichi
Yamazaki18 & Ralf Bodemann19
1C-K. Chou Consulting, Dublin, CA, USA, 94568
2R C Petersen Associates, LLC, Bedminster, NJ, USA, 07921
3University of Pennsylvania, Philadelphia, PA, USA, 19104
4Nagoya Institute of Technology, Nagoya, Japan
5Temple University Medical School, Philadelphia, PA, USA, 19140
6Metatec Associates, Silver Spring, MD, USA, 20904
7Richard Tell Associates, Inc, Madison, AL, USA, 35756
8Motorola Solutions, Inc., Fort Lauderdale, FL, USA, 33322
9Airman Systems Directorate, Fort Sam Houston, TX, USA, 78234
10Kavet Consulting LLC, Oakland, CA, USA, 94619
11Exponent Inc., Menlo Park, CA, USA, 94025
12EMF Consulting , Portland, OR, USA, 97219
13Health Canada (retired), Ottawa, Ontario, Canada, K1A 1C1
14University of California, Davis Health, Sacramento, CA, USA, 95817
15Exponent Inc., Bowie, MD, USA, 20715
16Full Spectrum Consulting, Concord, MA, USA, 01742
17Lawson Health Research Institute, London, Ontario, Canada
18Central Research Institute of Electric Power Industry, Yokosuka, Japan
19Siemens AG, Munich, Germany, D-81739
Keywords: Standards, All Frequencies, Completed (unpublished)
Presented by: C-K. Chou
C95.1 was first published as a radiofrequency (RF) protection standard in 1966 by the United States
of America Standards Institute. Major revisions of C95.1 were published in 1974, 1982, 1991, 1999,
2005, and 2014. This latest revision, a merging and updating of IEEE Std C95.1-2005 (3 kHz – 300
GHz) and C95.6-2002 (0 – 3 kHz), also relies upon C95.1-2345-2014 (military workplaces and
personnel protection), which also combined and updated C95.1 and C95.6. In this presentation,
similarities and differences of this draft revision with previous standards are discussed. New limits
for exposures between 6 GHz and 300 GHz are proposed based on thermal modeling studies. The

326
current draft revision is presently undergoing IEEE voting procedures.
Introduction
Recent literature reviews have not revealed sufficient evidence of any adverse health effects caused
by exposure to electromagnetic fields between 0-300 GHz that would change the current scientific
basis for determining safe levels of exposure in humans. For low frequencies, the established
adverse effect to be protected against is painful or aversive electrostimulation; at high frequencies it
is behavior-altering tissue heating. Improved numerical and measurement methods in low frequency
and RF dosimetry have increased knowledge concerning induced fields and related sensory effects
(electrostimulation), as well as power density and SAR-temperature relationships following RF
energy deposition in human tissue, which are essential to assessing potential biological and health
effects of EMF exposures. This draft revision incorporates a reasonably large margin of safety and
prescribes the implementation of a RF safety program to prevent exposures above the upper tier
restricted exposure limit. This draft revision is considered especially conservative, since the safety
factors are applied against perception phenomena (electrostimulation and behavioral disruption),
which are far less serious effects than any permanent or reversible tissue damage that might
conceivably occur above the perception threshold. This revision of IEEE Stds C95.1TM-2005 [1] and
C95.6-2002 [2] maintains many of the characteristics of these standards but also contains new
provisions that address recent dosimetry findings with the overarching objective of simplifying the
use and application of the standard.
Similarities
a) Research on the biological effects of electromagnetic interactions with tissues has not changed
the scientific basis of the adverse effect levels, i.e., electrostimulation for low frequencies and
heating for high frequencies.
b) Exposure limits for electrostimulation effects are kept the same as in IEEE Stds C95.6TM-2002
and C95.1TM-2005.
c) Exposure limits, termed dosimetric reference limits (DRLs), previously basic restrictions, on whole
body average and peak spatial-average specific absorption rates (SARs) remain the same to
prevent heating effects from exposure over much of the RF spectrum.
d) The exposure reference levels (ERLs), previously maximum permissible exposure (MPE) levels,
for the lower tier remain the same as in IEEE C95.1TM-2005.
e) The proposed draft standard continues to support the position of the earlier editions, i.e., upper
tier ERLs are protective of public health and safety and that the risk of harm to such fields below the
ERLs is unmeasurable.
Differences
a) IEEE Std C95.1TM-2005 contains two tiers; an upper tier for “people in controlled environments”
and a lower tier “action level” for implementing an RF safety program or MPE for the general public
when an RF safety program is not available. In this revision, operational definitions are introduced.
The terms lower and upper tiers are considered equivalent to the operational definitions “persons in
unrestricted exposure environments” and “persons permitted in restricted exposure environments”,
also termed “unrestricted tier” and “restricted tier” respectively.
b) A dosimetry reference limit (DRL) replaces basic restriction, and exposure reference limit (ERL)
replaces MPE, the terms used in the existing standards. The intent is to make the terms more
explicit and understandable.
c) The safety program initiation level (previously “action level”) is clarified as the ERL marking the
transition point between the lower (unrestricted) tier and the upper (restricted) tier.
d) The upper frequency boundary for whole body average (WBA) SAR has been changed from 3
GHz to 6 GHz because of improved measurement capabilities and to harmonize with the upcoming
revised ICNIRP guidelines. (Table 5 in the draft revision)
327
e) The averaging time is 30 minutes for whole body RF exposure and 6 minutes for local exposure.
It is scientifically more appropriate to select the averaging time according to the absorbing mass,
and not according to the exposure tier. (Tables 5-8 in the draft revision)
f) The term “extremities” as used in IEEE Std C95.1TM-2005 is changed to “limbs” involving the
whole arms and legs, instead of portions distal to the elbows and knees. This change is to
harmonize with IEEE Std C95.6TM-2002 and the ICNIRP guidelines [3].
g) Following the publication of novel dosimetry findings, the local exposure ERL is now frequency
dependent, instead of being a fixed factor of 20 times the whole-body ERL regardless of frequency.
(Table 9 in the draft revision)
h) The upper tier whole-body exposure ERLs above 300 MHz are different from those in IEEE Std
C95.1TM-2005 to maintain a consistent 5x factor between tiers and to harmonize with ICNIRP
guidelines.
i) The local exposure DRL and ERL for frequencies between 6 GHz and 300 GHz now are different.
The DRL is the epithelial power density inside the body surface, and ERL is the incident power
density outside the body. DRLs and ERLs for the limbs and pinnae are relaxed by a factor of two as
compared to the head and torso, consistent with the SAR limits below 6 GHz. Averaging power
density area is defined as a 4 cm2 square. (Table 10 in the draft revision)
j) Peak DRL and ERL limits for local exposures to pulsed RF fields are defined, and new fluence
limits for single RF-modulated pulses above 30 GHz are introduced. The averaging area for single
pulse fluence is 1 cm2 square.
k) The former induced current limit for both feet is considered an unrealistic condition and is
removed. The induced current limits for a single foot are retained. (Table 11 in the draft revision).
l) RMS induced and contact current limits for continuous sinusoidal waveforms (100 kHz to 110
MHz) are changed from those in Table 7 of C95.1-2005 to frequency dependent values. The new
limits are now identical to those in C95.1-2345-2014 [4] which was developed as a standard for
NATO military operation. (Table 12 in the draft revision)
Exposure Limits
In recognition of the differing biological effects of exposure to particular frequencies, the standard
addresses three frequency bands. The limits in the first band minimize adverse effects associated
with electrostimulation. This overlaps the second band where the limits also protect against effects
associated with heating. The limits in the third band protect against effects associated with body
surface heating.
0 Hz to 100 kHz (up to 5 MHz for pulsed fields):
The revision defines DRL in terms of the in situ (within biological tissue) electric fields for different
regions of the body. Magnetic field ERLs are specified for the arms and legs and for the head and
torso. Compliance with the standard can be demonstrated for uniform sinusoidal magnetic fields by
showing that either the in situ electric field DRL or the magnetic field ERL is satisfied. If the magnetic
field is not constant over the head and torso, compliance with the DRLs is satisfied if the spatial
peak of the magnetic field ERL is not exceeded. Averaging time for an RMS measurement is 0.2 s.
Formulas have been included for determining maximum permitted peak electric fields for both in situ
and environmental considerations. For non-sinusoidal, or pulsed sinusoidal fields, DRLs and ERLs
are specified based either on instantaneous temporal measurements, or on Fourier components of
the exposure waveform. The choice between these two compliance tests is left to the discretion of
the user. These provisions remain unchanged relative to the existing standard.
100 kHz to 6 GHz:
In this frequency range where SAR is the controlling criterion, the revised standard confirms the
presumed threshold WBA SAR of 4 W/kg for potentially adverse effects. Local DRLs expressed as
the peak spatial average SAR (psSAR) in 10 g of tissue in the shape of a cube for any body tissue

328
including the arms, legs and pinnae, are specified. This revision retains the limits for psSAR of 10
W/kg for the upper/restricted tier and 2 W/kg for the lower/unrestricted tier, and those for the limbs
and pinnae 20 W/kg and 4 W/kg for the two tiers respectively. Table 5 of the draft revision shows the
DRL for this frequency range. The contact current limits for the frequency range of 100 kHz to 110
MHz have been frequency dependently subdivided into touch and grasping conditions, with the
grasping condition constrained to the restricted environment. The permissible touch contact current
has been reduced for both the restricted environment and the unrestricted environment. The lower
portion of this frequency range, i.e., 100 kHz to 5 MHz, is a transition region where both the limits
protecting against electrostimulation and the limits protecting against effects associated with heating
must be met. The draft revision identifies factors associated with research on contact currents that
need additional study to ensure that the limits are applicable to practical conditions found in the
workplace.
6 GHz to 300 GHz:
The principal changes within this band have been in the local exposure DRLs and ERLs. Since the
interactions are mostly at or just within the body surface and are quasi-optical, the DRLs are
expressed in terms of a newly introduced metric, the “epithelial power density” and associated
exposure averaging interval. The corresponding ERLs are defined as function of frequency to
account for the increasing transmitted power across the body/air interface. The local exposure DRLs
and ERLs are both averaged over 4 cm2. There are limited experimental data upon which to set
limits on exposures above 6 GHz. Pending availability of more data, the limits above 6 GHz were
revised to provide a similar level of protection against thermal hazards as that provided in the
current limits below 6 GHz, using thermal modeling and dosimetric studies. ICES recognizes the
need for additional data in this frequency range. The principal changes within this band have been in
the DRL and ERL. At 300 GHz, the ERL for the restricted environment is equal to the reference level
(MPE) in the laser standards, the coverage of which begins at 300 GHz. Attached Tables 6 and 10
from the draft revision show the new local exposure DRLs and ERLs for this frequency band. For
whole body exposures, the lower tier ERL limits remain the same but the higher tier ERL was
adjusted to keep a constant 5x between the two tiers.
Summary
This standard revises and combines IEEE C95.6-2002 and C95.1-2005 as well as components of
the C95.1-2345-2014 into a single standard. The main changes are the limits in the frequency range
of 6 GHz to 300 GHz. Local exposure DRLs and ERLs are derived from recent thermal modeling
studies. The “epithelial power density” at the body surface is a newly introduced DRL for frequencies
above 6 GHz. Limits for frequencies below 6 GHz are mostly kept the same as before, with some
modifications. Due to thermophysiological considerations, time averaging now depends on whether
it is a whole-body or local exposure rather than restricted or unrestricted exposure tier. Voting is in
progress. It is expected to finish the IEEE Standards Association approval process at the end of
2018.
References
1. IEEE C95.1-2005, “IEEE Standard for Safety Levels with Respect to Human Exposure to Radio
Frequency Electromagnetic Fields, 3 kHz to 300 GHz.” Piscataway, NJ, USA.
2. IEEE C95.6-2002, “IEEE Standard for Safety Levels with Respect to Human Exposure to
Electromagnetic Fields, 0–3 kHz.” Piscataway, NJ, USA.
3. ICNIRP, “Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic
fields (up to 300 GHz)," Health Physics, Vol. 74, No. 40, 494-522, 1998.
4. IEEE C95.1-2345-2014, “IEEE Standard for Military Workplaces—Force Health Protection
Regarding Personnel Exposure to Electric, Magnetic, and Electromagnetic Fields, 0 Hz to 300 GHz”
Piscataway, NJ, USA.
Figures

329
Figure 1. DRLs (100 kHz to 6 GHz) [Table 5 in draft revision]

Figure 2. Local exposure DRLs (6 GHz to 300 GHz) [Table 6 in draft revision]

330
 Figure 3. Local exposure ERLs (6 GHz to 300 GHz) [Table 10 in draft revision]

S16-5 [10:30]
Safety in electromagnetic fields - Simplifying risk assessments using exposure zones
Peter Jeschke1, Hannah Heinrich2, Claudine Neumann3, Ingo Bömmels3, Stephan Joosten4, Mathias
Hoffmann5, Hannelore Neuschulz1 & Erik Romanus1
1Physical Agents, Federal Institute for Occupational Safety and Health, Dortmund, Germany, 44149
22h-enginnering & -research, Hausen, Germany, 91353
3NIR - EMF, DGUV Institute for Occupational Safety and Health, St. Augustin, Germany, 53757
4EMF, German Social Accident Insurance, Collogne, Germany, 50960
5AB 35: Radiation Protection and NIR, NLWKN, Hildesheim, Germany, 31135
Keywords: Occupational, All Frequencies, Work in Progress
Presented by: Peter Jeschke
The European Directive 2013/35/EU has been implemented into German legislation by the EMF
Ordinance. To improve the comprehensibility and usability of the EMF Ordinance, Technical Rules
are currently developed to assist employers, particularly in small and medium sized enterprises, to
implement state of the art occupational safety and health measures. Using the information and
measures given in the Technical Rules ensure compliance with the EMF Ordinance. “Exposure
Zones”, which link exposure conditions at workplaces with Action Levels and associated protective
measures, provide a simple and guided approach for employers to fulfil the requirements of the EMF
Ordinance.
The presentation describes how the approach of “Exposure Zones” simplifies the risk assessments
required by both the German Ordinance on the Protection of Workers against the Risks arising from
Electromagnetic Fields (EMF Ordinance) and the European Directive on Electromagnetic Fields
(2013/35/EU) and provides insight into the current work on Technical Rules for the EMF Ordinance
(TREMF).

331
Implementation of European Directive on EMF (2013/35/EU) into German occupational safety
and health legislation
Similar to all other European member states, the European Directive 2013/35/EU (Directive on
Electromagnetic Fields) had to be implemented into German occupational safety and health
legislation. It was open to the member states whether to implement the European Directive on
Electromagnetic Fields into national legislation directly or to amend existing regulations as in
Germany. For a detailed comparison of international policies on EMF please refer to Stam (2018).
The EMF Ordinance came into force in November 2016 and is a binding regulation for any risks
arising from the exposure to electromagnetic fields at workplaces. To support particularly small and
medium sized enterprises (SMEs) with an elaborated but applicable safety approach, Technical
Rules will improve the comprehensibility and usability of the German EMF Ordinance. With
Technical Rules Germany’s occupational safety and health legislation supports employers in fulfilling
their obligations arising from the German EMF Ordinance and the European Directive on
Electromagnetic Fields (2013/35/EU) respectively. Comparable with the role of harmonized
standards, Technical Rules give presumption of conformity to the German EMF Ordinance.
Structure and contents of the Technical Rules on EMF
The German Federal Ministry of Labour and Social Affairs has assigned the Advisory Committee on
Operational Safety to develop Technical Rules for the German EMF Ordinance. Representatives of
employers’ associations, labour unions, the German Social Accident Insurance, and governmental
occupational safety and health authorities along with scientists are appointed members to this
Advisory Committee. At the end of the development process, all appointed members are obliged to
reach consensus on the matter. Such process guarantees a well-defined balance between
effectiveness and efficiency of safety measures. It differs significantly from the development process
of international standards, which could be prone to be driven by other interests than for instance
occupational safety and health. To summarize, the consensus driven development process of
Technical Rules along with presumption of conformity to the German EMF Ordinance mark
substantial differences compared to the European non-binding guide for implementing the Directive
on Electromagnetic Fields (2013/35/EU).
Based on the EMF Ordinance, three Technical Rules are currently developed by working groups of
the Advisory Committee on Operational Safety covering the frequency range beginning with static
electric and magnetic fields leading up to 300 GHz EMF. Refer to figure 1 for the effects of EMF and
the frequency ranges covered by the scope of the three Technical Rules.
All three Technical Rules are comprised of chapters specific to static and low frequency, high
frequency, and frequency ranges of Magnetic Resonance applications elaborating on “risk
assessment”, “measurement and calculation”, and “provisions aimed at avoiding or reducing risks”
(refer to figure 2). This separation proves appropriate to account for different direct and indirect
effects specific to each frequency range. Common chapters to all three Technical Rules include
definitions of legal terms, physical quantities, and an overview of the safety concept along with
Exposure Zones. Appendices provide background information and descriptions of direct and indirect
effects of electromagnetic fields, exposure scenarios, or elaborated sample risk assessments.
Development of Exposure Zones
The safety concept of the European Directive on Electromagnetic Fields and consequently its
German counterpart is somewhat difficult to apply for nonEMF-expert employers, especially at
SMEs. Action Levels as “operational levels established for the purpose of simplifying the process of
demonstrating the compliance with relevant ELVs [exposure limit values]“ (Art. 2 (g), 2013/35/EU)
are a first step towards an easy to use safety approach. However, applying Action Levels still require
a sound knowledge to draw correct conclusions and to derive protection provisions. In order to
enhance comprehensibility and usability, Exposure Zones promise to ease the burden of performing
risk assessments and hence to better protect employees at risk. As part of safety approaches for
example in current international standards (e.g. prEN50499, FprEN 50496), Exposure Zones are
well established in Germany too. Exposure Zones as defined in DGUV-R 103-013 (DGUV 2006)

332
have proven practicability in numerous risk assessments of a wide variety of workplaces. Based on
the German EMF Ordinance, an updated version of Exposure Zones is central to the approach of
the Technical Rules on EMF currently under development.
Based on Action Levels, the updated Exposure Zones slightly differ from existing concepts as
Exposure Zones considers workers at particular risk based on the research report FB 451 (BMAS
2015). Workers at particular risks are for example those who wear active or passive implantable
medical devices, e.g. cardiac pacemakers or stents. Along with implementing Exposure Zones as an
easy to use approach for Technical Rules on EMF selected research questions have to be solved,
for example the connection of Action Levels for non-thermal and thermal effects at intermediate
frequencies (please refer to figure 3). This missing link is caused partly by revisions of ICNIRP
guidelines on low frequencies resulting in a raise of exposure limit values and Action Levels
respectively (ICNIRP 2010). The ICNIRP guidelines on high frequencies are under revision since
2014 (ICNIRP 2017).
In summary, Exposure Zones relate exposure conditions with Actions Levels as well as required
protective provisions. Hence Exposure Zones provide employers, particularly of SMEs, with a
comprehensible and transparent easy to use safety approach to assess risks for workers arising
from electromagnetic fields.
References
[2013/35/EU]: “Directive on the minimum health and safety requirements regarding the exposure of
workers to the risks arising from physical agents (electromagnetic fields)”; entered into force on
29.06.2013.
[EMFV]: “German Ordinance on the Protection of Workers against the Risks arising from
Electromagnetic Fields (EMF Ordinance)”, entered into force on 15.11.2016.
[Stam, 2018]: “Comparison of international policies on electromagnetic fields (power frequency and
radiofrequency fields)”, 20 pages, Bilthoven: RIVM.
[EC, 2016]: “Non-binding guide to good practice for implementing Directive 2013/35/EU
Electromagnetic Fields”; Volume 1, 2, and Guide for SMEs.
[prEN 50499:2017]: „Procedure for the assessment of the exposure of workers to electromagnetic
fields“.
[FprEN 50496]: „Determination of workers' exposure to electromagnetic fields and assessment of
risk at a broadcast site“.
[DGUV, 2006]: „Electromagnetic Fields“, DGUV-R 103-013 (former BGR B11), 53 pages .
[BMAS, 2015]: “Electromagnetic fields at workplaces - Safety of workers at particular risk” FB 451,
46 pages, ISSN 0174-4992.
[ICNIRP, 2010]: “ICNIRP Guidelines for limiting exposure to time-varying electric and magnetic fields
(1Hz - 100kHz)”, published in Health Physics, 99(6):818 ‐ 836; 2010.
[ICNIRP, 2017]: “Revision of the HF Guidelines”, accessed on: 23.02.18, http://www.icnirp.org/en/ac-
tivities/news/news-article/revision-of-hf-guidelines-2017.html.
Figures

333
Figure 1. Frequency range of German EMF Ordinance and the scope of the corresponding
Technical Rules on ElectroMagnetic Fields: 1 - TREMF static and low frequency EMF, 0 ≤ f ≤
10MHz | 2 - TREMF high frequency EMF 100 kHz ≤ f ≤ 300 GHz | 3 - TREMF Magnetic
Resonance Applications (imaging and spectroscopy) with static, low, and high frequency EMF

Figure 2. Structure and contents of TREMF

334
 Figure 3. Sample frequency dependency of Exposure Zones for magnetic fields as promoted
in the TREMF with exposure level at work place in accordance to EZ 0 - AL AIMD (AL for
active implantable medical devices), AL PIMD (AL for passive implantable medical devices) |
EZ 1 - low AL (for f ≤ 400Hz) and high AL (f>400Hz) | EZ 2 - high AL (f > 400Hz) | EZ 3 - AL of
limbs

S16-6 [10:45]
An analytical model to derive safety limits for pulsed/modulated 5G radiation exposure
Esra Neufeld1 & Niels Kuster1, 2
1IT'IS Foundation for Research on Information Technologies in Society, Zürich, Switzerland, 8004
2Swiss Federal Institute of Technology (ETHZ), Zürich, Switzerland, 8092
Keywords: Standards, RF/Microwaves, Completed (published)
Presented by: Niels Kuster
5G technology foresees data transmission in short bursts. Signals with high peak-to-average ratios
can lead to critical tissue heating even when time- and area-averaged power density values respect
current safety guidelines. An analytical model of heating due to plane-wave and localized pulsed
exposure was developed and has been used in combination with thermal dose limits to derive
propositions for the adaptation of current standards.
Background
5G telecommunication technology foresees data transmission in short bursts. Even though the time-

335
and area-averaged power density values might respect current safety guidelines (e.g., [1]), the high
peak-to-average ratio (PAR), in combination with rapid temperature increases associated with
surface energy deposition, may result in tissue damage. To address this issue, limits on fluence or
radiant exposure have been proposed previously. We have created an analytical model of pulsed
tissue heating and resulting thermal dose to investigate the impact of averaging-time and PAR, and
to derive guidance for standardization. Both, plane-wave and localized exposure are considered.
Methods
A PAR- and averaging-time dependent analytical formulation for the quasi-steady-state temperature
oscillation resulting from pulsed exposure can be derived from the step-response function. It
reduces to a simple approximation formula for the relative temperature oscillation magnitude for high
PAR, but can also be used to compute thermal dose according to the CEM43 model (cumulative
equivalent minutes of heating at 43°C). The general analytical approach was evaluated by using the
step-response function of the Pennes Bioheat equation applied to a surface heated, semi-infinite,
and perfused tissue. To consider localized exposure, the thermal time constant was adapted
according to published values for effective time-constants.
Results
The results demonstrate that currently foreseen averaging-times, intensity levels, and PAR
thresholds of 5G technology can easily result in tissue damage. By introducing CEM43 thresholds
for the general public and occupational exposure (using literature data on damage thresholds and
safety factors previously employed in safety standards), as well as assumptions about the maximal
exposure duration, exposure intensity and PAR-limit dependent limits on averaging time can be
derived. Based on these evaluations, it is suggested to adapt standards to:
limit the continuous exposure temperature increase to 1K (the currently considered 4K is clearly
too high);
employ two different averaging times – i.e., a newly introduced short averaging time that allows
for rapid bursts, short enough to not significantly affect the temperature (e.g., for bursts longer
than 30us, an averaging time of 10ms reduces relative oscillation magnitudes below 20%), and
a longer averaging time similar in function to the one foreseen by the current standards;
derive limits on averaging time and maximal PAR from thermal dose considerations (reasonably
long averaging times (30-240s) are obtained in combination with reasonable PAR restrictions).
Limits on fluence alone are found to be insufficient to guarantee safety.
Conclusions
An analytical model for tissue heating and damage by transient exposure to 5G radiation was
developed, which demonstrates that current safety thresholds, as well as considered adaptations of
the standard can result in exposures that may not be consistent with the safety concept of the
guidelines. The proposed model allows to derive meaningful relationships between PAR thresholds,
averaging time, and exposure intensities, and has been used to derive propositions for the
adaptation of current standards.
References
[1] International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines for Limiting
Exposure to Time-Varying Electric, Magnetic and Electromagnetic Fields (up to 300 GHz),” Health
Physics, vol. 74, no. 4, pp. 1–32, Apr. 1998.

336
 Session: PW1
Pre Conference Workshop: EMF exposure from 5G equipment: the state of art of research
and standardization
Sunday June 24, 2018 • 13:00 - 18:00
Europa C
Chair: Ken Joyner

PW1-1 [13:00]
Workshop content
Ken Joyner1
1Australia

Although networks utilizing frequency bands below 6 GHz will still constitute the backbone of mobile
communications, new wireless access technologies, such as 5G, will operate also in higher bands.
This workshop will present the increasing amount of studies addressing millimeter wave EMF
exposure. It will also discuss novel compliance assessment methodologies and procedures that
have been developed and are being standardized for both base stations and portable devices.

PW1-2 [13:05]
Overview of 5G and the EMF Compliance Standards Development program from the
InternationalElectrotechnical Commission
Michael Wood1
1Chairman IEC TC106 and Telstra Corporation, Melbourne, Australia

PW1-3 [13:30]
EMF Exposure limits above 6 GHz
Akimasa Hirata1
1Nagoya Institute of Technology, Nagoya, Japan

PW1-4 [14:00]
EMF exposure of the skin at the mmW
C. K. Chou1
1IEEE/ICES TC95 Chairman, USA

PW1-5 [14:30]
Ocular studies of EMF exposure at the mmW
Masami Kojima1 & Yaeko Suzuki2
1Kanazawa Medical University, Uchinada, Japan
2Tokyo Metropolitan University, Hachiōji, Japan

PW1-6 [15:00]
On the averaging time of human exposure at frequencies above 6 GHz
Kenneth Foster1
1University of Pennsylvania, Pennsylvania, PA, USA

337
PW1-7 [15:30]
In vivo experimental measurements of thermal change due to mmW exposure
Leeor Alon1
1New York University School of Medicine, New York, NY, USA

PW1-8 [16:00]
EMF energy absorption mechanisms at the mmW
Andreas Christ1
1Research Consultant, Cabo Frio, Brazil

PW1-9 [16:30]
Challenges in standardization related to EMF compliance above 6 GHz
Davide Colombi1
1Ericsson Research, Stockholm, Sweden

PW1-10 [17:00]
EMF compliance assessments of 5G devices
Niels Kuster1
1IT'IS Foundation, Zurich, Switzerland

PW1-11 [17:40]
EMF compliance assessment based on a mixed approach of measurements and numerical
simulations
Wassim El Hajj1
1Intel Corporation, Nice, France

338
 Session: PW2
Pre Conference Workshop: MURI/AFOSR Workshop on Nanoelectropulse-induced
electromechanical signaling and control of biological systems
Sunday June 24, 2018 • 13:00 - 16:30
Europa D
Chairs: Andrei Pakhomov & Bennett Ibey

PW2-1 [13:00]
Workshop content
Andrei Pakhomov1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

Multidisciplinary University Research Initiative (MURI) grants are awarded by the US Air Force
Office of Scientific Research (AFOSR) to support fundamental, cutting-edge research that crosses
traditional science and engineering boundaries. This Workshop will focus on cellular and molecular
mechanisms underlying bioeffects of nanosecond pulsed electric fields (nsPEF). Talks will highlight
the impact of bipolar cancellation, a phenomenon which makes bipolar stimuli and emitted EMF less
effective. We will summarize experimental and modeling efforts in support of a concept of remote
stimulation and electroporation by “cancellation of cancellation” (CANCAN). We will conclude with a
first-time introduction and discussion of a synthetic theory which links nanoelectroporation and
bipolar cancellation mechanisms.
The workshop welcomes all BioEM participants and does not require registration.

PW2-2 [13:00]
Universality of bipolar cancellation of nanoelectropulse bioeffects
Andrei Pakhomov1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

PW2-3 [13:25]
Direct observation of membrane charging during uni- and bipolar nanosecond pulse
exposure
Bennett Ibey1
1Air Force Research Laboratory, 711th Human Performance Wing, Airman Systems Directorate, Bioeffects
Division, Radio Frequency Bioeffects Branch, Fort Sam Houston, TX, USA

PW2-4 [13:50]
Bipolar, bidirectional cancellation of electrostimulated molecular transport — Nanoscale and
microscale models and measurements
P. Thomas Vernier1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

PW2-5 [14:15]
Break

339
PW2-6 [14:25]
Nanopore behavior during and after pulsing: General and bipolar cancellation implications
James Weaver1
1Institute for Medical Engineering & Science, Massachusetts Institute of Technology, Boston, MA, USA

PW2-7 [14:45]
Activation of voltage-gated calcium channels by uni- and bipolar nsPEF
Olga Pakhomova1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

PW2-8 [15:10]
Cancellation of Ca2+ signaling by bipolar nsPEF in neural crest-derived bovine adrenal
chromaffin cells – Pulse duration is critical!
Normand Leblanc1
1Center for Cardiovascular Research, University of Nevada, Reno School of Medicine, Reno, NV, USA

PW2-9 [15:35]
Break

PW2-10 [15:45]
Open discussion: Synthetic theory of nanoporation and bipolar cancellation
Andrei Pakhomov1 & Bennett Ibey2
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA
2Air Force Research Laboratory, 711th Human Performance Wing, Airman Systems Directorate, Bioeffects
Division, Radio Frequency Bioeffects Branch, Fort Sam Houston, TX, USA

340
 Session: T1
Tutorial 1: Systematic Review Approaches in Environmental Health Sciences
Tuesday June 26, 2018 • 08:30 - 09:30
Emerald
Chairs: James Lin & Gunnhild Oftedal

T1-1 [08:30]
Systematic review approaches in environmental health sciences
Kristina Thayer1
1Integrated Risk Information System (IRIS) Division, National Center for Environmental Assessment (NCEA),
US Environmental Prote, Durham, NC, USA, 27711
Biographical sketch
Kristina Thayer is the Integrated Risk Information System (IRIS)
Division Director, located within the US Environmental Protection
Agency (EPA) National Center for Environmental Assessment
(NCEA). IRIS assessments identify the potential for a chemical to
cause cancer or non-cancer health effects in people and are
considered the top tier source of toxicity information used by EPA
and other agencies to inform national standards, clean-up levels at
local sites, and set advisory levels. IRIS uses systematic review
methods to conduct assessments. Prior to joining EPA, Dr. Thayer
was Deputy Division Director of Analysis at the National Toxicology
Program (NTP) and Director of the NTP Office of Health Assessment
and Translation (OHAT) located on the campus of the National
Institute for Environmental Health Sciences (NIEHS). She is
considered an expert on the application of systematic review
methods to environmental health topics and the use of specialized software and automation
approaches to facilitate conducting reviews.
Five relevant publications
1. NAS February 1-2, 2018 Workshop: “Review of Advances Made to the IRIS Process” (agenda
and meeting materials)
2. Howard BE, Phillips J, Miller K, Tandon A, Mav D, Shah MR, Holmgren S, Pelch KE, Walker V,
Rooney AA, Macleod M, Shah RR, Thayer KA. 2016. SWIFT-Review: a text-mining workbench
for systematic review. Systematic reviews 5(1):87. https://www.ncbi.nlm.nih.gov/pmc/articles
/PMC4877757/
3. Rooney AA, Cooper GS, Jahnke GD, Lam J, Morgan RL, Boyles AL, Ratcliffe JM, Kraft AD,
Schunemann HJ, Schwingl P, Walker TD, Thayer KA, Lunn RM. 2016. How credible are the
study results? Evaluating and applying internal validity tools to literature-based assessments of
environmental health hazards. Environment international 92-93:617-629.
https://www.sciencedirect.com/science/article/pii/S0160412016300058?via%3Dihub
4. National Toxicology Program (NTP). 2015. Handbook for Conducting a Literature-Based Health
Assessment Using NTP Office of Health Assessment and Translation (OHAT) Approach for
Systematic Review and Evidence Integration. http://ntp.niehs.nih.gov/go/38673.
5. Rooney AA, Boyles AL, Wolfe MS, Bucher JR, Thayer KA. 2014. Systematic review and
evidence integration for literature-based environmental health science assessments. Environ
Health Perspect. 122(7):711-718. https://ehp.niehs.nih.gov/1307972/
Abstract
Adapting a systematic approach on scientific revision of EMF/health related evidence in order to
bring more transparency to evaluations has become an important priority as the WHO RF EHC

341
document and ICNIRP new guidelines are being prepared. The tutorial will provide an overview of
implementation of systematic review methods to environmental health topics, to include the basic
tenants of identifying relevant literature, approaches used to assess study quality, use of structured
frameworks to convey confidence conclusions for a body of evidence, and approaches to integrate
epidemiological, toxicological, and mechanistic evidence.

Session: W1
Workshop 1: Transferring Evidence from Research and Case Reports into Policy and
Society
Tuesday June 26, 2018 • 16:30 - 18:00
Europa C
Chair: Jürg Kesselring

W1-1 [16:30]
Workshop content
Stefan Dongus1 & Jürg Fröhlich2
1Swiss TPH, Switzerland
2Fields at Work, Switzerland

Potential health effects of electromagnetic fields are not only debated by scientists, but also by
interested lay people, affected individuals, medical practitioners and policy makers. While hundreds
of scientific studies are published worldwide each year investigating a wide range of related aspects
and topics, the challenge is not only to continuously assess the potential implications for existing
and future regulatory limits, but also to translate the results for a wider public, in order to meet the
need for reliable, balanced and comprehensive information. This workshop will provide examples for
existing approaches of synthesizing research for policy and society, experiences from policy makers
providing guidelines for regulatory limits, and insights from medical practitioners, closing with a
roundtable discussion.

W1-2 [16:35]
EU-level risk assessment by SCHEER – the Scientific Committee on Health, Environmental
and Emerging Risks, European Commission
Theodoros Samaras1
1Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece

W1-3 [16:45]
Country-level risk assessment in France by ANSES – the French National Agency for Food
Safety, Environment and Labor
Johanna Fite1
1Department of Risk Assessment, ANSES (French National Agency for Food Safety, Environment and
Labor), Maisons-Alfort, France

W1-4 [16:55]
The perspective of the European Academy for Environmental Medicine (EUROPAEM)
Michael Kundi1
1Center for Public Health, Department of Environmental Health, Medical University of Vienna, Viena, Austria

342
W1-5 [17:05]
Values and Evidence in Risk Assessment: Results from a Comparative Analysis
Gregor Durrenberger1
1Swiss Research Foundation on Mobile Communication - FSM, ETH Zurich, Zurich, Switzerland

W1-6 [17:15]
Insights from 30 years of medical practice in a neurology clinic
Jürg Kesselring1
1Rehabilitation Centre, Clinics of Valens, Valens, Switzerland, 7317

W1-7 [17:25]
Round Table Discussion
Jürg Kesselring1
1Rehabilitation Centre, Clinics of Valens, Valens, Switzerland, 7317

Session: T2
Tutorial 2: Magnetic Nanoparticles as Therapeutic Agents: Focusing on the Role of a
Scientist and Engineer
Wednesday June 27, 2018 • 09:00 - 10:00
Emerald
Chairs: Antonio Sarolic & Lluis M. Mir

T2-1 [09:00]
Magnetic nanoparticles as therapeutic agents: Focusing on the role of a scientist and
engineer
Theodoros Samaras1
1Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece, GR-54124

Biographical sketch
Theodoros Samaras is Professor of Bioelectromagnetics at the
Department of Physics, Aristotle University of Thessaloniki, Greece.
He is a trained Medical Physicist specializing in the non-ionizing
spectrum. He is working with computational multiphysics/multiscale
methods for applied electromagnetics. His research interests include
exposure assessment (both numerical and experimental) to
electromagnetic fields and radiation; applications of electromagnetic
fields in cancer therapy (hyperthermia, thermal ablation techniques);
non-invasive brain stimulation with transcranial techniques; patient
monitoring with electrical impedance tomography. In recent years, he
has been active in the area of biomedical uses of magnetic
nanoparticle, where he is attempting to transfer knowledge in
dosimetry from other areas of bioelectromagnetics He has been
serving as a reviewer for several journals and funding organizations
and as the national representative to European research co-ordination actions (BM0704, BM1309,
TD1104, EMF-NET) and standardization committees (IEC TC106). He is currently member of the
European Commission's Scientific Committee on Health, Environmental and Emerging Risks
(SCHEER) and has been elected twice in the Council of the European Bioelectromagnetics
Association (EBEA) and in the Board of the European Society for Hyperthermic Oncology.

343
Five relevant publications
1. Maniotis, N., Makridis, A., Myrovali, E., Theopoulos, A., Samaras, T., Angelakeris, M. “Magneto-
mechanical action of multimodal field configurations on magnetic nanoparticle environments”
(2017) Journal of Magnetism and Magnetic Materials. Article in Press.
2. Simeonidis, K., Martinez-Boubeta, C., Rivera-Gil, P., Ashraf, S., Samaras, T., Angelakeris, M.,
Tresintsi, S., Mitrakas, M., Parak, W.J., Monty, C., Balcells, L. “Regeneration of arsenic spent
adsorbents by Fe/MgO nanoparticles” (2017) Journal of Chemical Technology and
Biotechnology, 92 (8), pp. 1876-1883.
3. Myrovali, E., Maniotis, N., Makridis, A., Terzopoulou, A., Ntomprougkidis, V., Simeonidis, K.,
Sakellari, D., Kalogirou, O., Samaras, T., Salikhov, R., Spasova, M., Farle, M., Wiedwald, U.,
Angelakeris, M. “Arrangement at the nanoscale: Effect on magnetic particle hyperthermia”
(2016) Scientific Reports, 6, art. no. 37934.
4. Makridis, A., Tziomaki, M., Topouridou, K., Yavropoulou, M.P., Yovos, J.G., Kalogirou, O.,
Samaras, T., Angelakeris, M. “A novel strategy combining magnetic particle hyperthermia pulses
with enhanced performance binary ferrite carriers for effective in vitro manipulation of primary
human osteogenic sarcoma cells” (2016) International Journal of Hyperthermia, 32 (7), pp.
778-785.
5. Chalkidou, A., Simeonidis, K., Angelakeris, M., Samaras, T., Martinez-Boubeta, C., Balcells, L.,
Papazisis, K., Dendrinou-Samara, C., Kalogirou, O. “In vitro application of Fe/MgO nanoparticles
as magnetically mediated hyperthermia agents for cancer treatment” (2011) Journal of
Magnetism and Magnetic Materials, 323 (6), pp. 775-780.
Abstract
The tutorial addresses the use of magnetic nanoparticles (MNP) in medical therapy with a brief
review of existing and envisaged techniques. Since nanotechnology is a multidisciplinary field, the
tutorial focuses on the role of scientists and engineers in this emerging field of biomedicine. This
role is not restricted to the manufacturing and physical characterization of MNP, using techniques
like electron microscopy, magnetization measurements, etc. It extends to their in vitro and in vivo
dosimetry, where dosimetric quantities, like specific absorption rate (SAR) or specific loss power
(SLP), familiar to the bioelectromagnetics audience are used. However, new dosimetric concepts,
like force per cell, may need to be introduced for the medical applications of MNP. Scientists and
engineers working in this field are also challenged by uncertainty quantification and the reporting of
MNP properties, which can vary a lot. Another area of interest is the design of appropriate magnetic
fields and devices, reaching to other disciplines of computational multiphysics, like computational
fluid dynamics. In this tutorial I will discuss, with the appropriate examples, the involvement of
scientists and engineers in turning MNP to therapeutic agents.

344
 Session: T3
Tutorial 3: Evaluation of Animal Cancer Studies (overview)
Thursday June 28, 2018 • 08:30 - 09:30
Emerald
Chairs: Isabelle Lagroye & Christopher Portier

T3-1 [08:30]
Evaluation of animal cancer studies
Chad Blystone1
1National Toxicology Program (NTP), NIEHS, National Institute of Health, Durham, NC, USA, 27709

Biographical sketch
Dr. Chad Blystone is a Toxicologist with the National Toxicology Program (NTP)
located within the National Institute of Environmental Health Sciences (NIEHS). His
expertise includes designing and interpreting animal cancer studies. He leads the
NTP’s Chronic Toxicity and Carcinogenicity Technical Reports, a program that has
evaluated chemicals and other agents in the two-year bioassay for over 30 years.
These studies have contributed a significant amount of data for identifying potential
human health hazards and NTP’s nearly 600 studies have provided an extensive
database for researchers and international risk assessment organizations.
Abstract
In order to protect human health, rodent cancer studies have been used for decades to identify
potential carcinogens. These studies are specifically designed to identify carcinogenic activity of
tested agents, identify target organs, and characterize the exposure-response relationship after a
long term exposure. Various health regulatory agencies may require rodent cancer studies during
the development of new pharmaceuticals, pesticides, or other agents of potential concern. In the risk
assessment of the tested agent, no observed adverse effect levels (NOAELs) or benchmark dose
(BMD) will often be used to extrapolate the results to typically low human exposure levels. This
tutorial will go through the various elements in designing, conducting, and interpreting rodent cancer
studies to provide a background on these prominent evaluations of possible human health hazards.

Session: W2
Workshop 2: EMF and Cancer: Interaction Mechanisms Leading to Hazardous or
Beneficial effects
Thursday June 28, 2018 • 13:30 - 15:00
Europa D
Chairs: Antonio Sarolic & Olga Zeni

W2-1 [13:30]
Workshop content
Mirjana Moser1
1Bern, Switzerland
While EMFs in the extremely-low frequency and radiofrequency ranges are classified as possible
carcinogens for humans, there are also an increasing number of clinical applications of these EMFs,
including for cancer treatment. The aim of the workshop is to present and discuss available
knowledge on possible EMF interaction mechanisms that are leading to risk of cancer, vs.
interaction mechanisms that are leading to “beneficial” effects that can be used for therapeutical
purposes. A change of view from the risk perspective only, to a scenario with possible benficial

345
effects as well may be the beginning of a pradigm change. Is it also so that these novel insights
provide a better understanding of the interaction mechanisms?

W2-2 [13:40]
Research on environmental EMF as possible carcinogens in relation with interaction
mechanisms also leading to beneficial effects
Isabelle Lagroye1
1Ecole Pratique des Hautes Etudes, Paris, France

The topics of EMF and cancer is one of the most debated in the “environment and health” field since
about 40 years. The main frequency ranges leading to public exposures are extremely-low
frequency magnetic fields (ELF-MF, 50-60 Hz) and radiofrequency fields (RF, around 1 GHz).
Extremely-low frequency fields are classified as possible carcinogens for humans since 2002 for
exposure of more than 0.3-0.4 µT in relation with childhood leukemia. Radiofrequency are classified
as possible carcinogens for humans since 2011 for RF emitted by mobile phones and glioma and
acoustic neuroma. The so-called 2B classification means that the causal relationship couldn’t be
demonstrated, and notably here, interaction mechanism underlying such effects couldn’t be clearly
identified.
In the recent decades however, the use of EMF in therapeutic applications has emerged. A variety
of electric (nanopulses), magnetic fields (pulsed magnetic fields), and electromagnetic fields (RF)
are investigated for cancer therapy. The former, however, are not relevant for public exposures.
Medical application of EMF in the field of oncology are often seen as a complementary treatment
with no side-effects.
Tissue heating is one of the best example of interaction mechanisms leading to both beneficial and
deleterious effects. Indeed, tissue heating is identified in the radiofrequency range as the interaction
mechanism responsible for cancer treatment with recognized clinical input. While deleterious effects
are understood as driven by the same heating mechanism, the evidence that cancer would be one
of these effects seems weak.
A summary of the current knowledge on the mechanisms investigated for the effects of
environmental and therapeutic EMF will be presented. For both deleterious and beneficial effects,
main biological mechanisms investigated are genotoxicity, apoptosis, and reactive oxygen species
production.
Should we only consider that the dose makes the poison? If not, what else?
In this regards, a key element to be considered is whether EMF have differential effects in normal
versus cancer cells. Finally, will autophagy and the so-called adaptive response, recently
investigated in bioelectromagnetics, shed some light on in the field?

W2-3 [14:05]
Research on use of EMF in cancer treatment from the point of view of the COST Action EMF-
MED
Mats-Olof Mattsson1, 2
1AIT Austrian Institute of Technology, Tulln, Austria, 3430
2SciProof International AB, Östersund, Sweden, 83132

The COST Action BM1309 (European network for innovative uses of EMFs in biomedical
applications, EMF-MED) provides a cooperative framework to support research on beneficial
biological effects of non-ionizing electromagnetic fields and their use in biomedical applications. One
of the Working Groups of the Action focuses on “Cancer-EMF interactions and applications”, where
a platform for research dealing with diagnosis and treatment of cancer using EMF and/or EMF-

346
based technologies has been established. This presentation has the focus on interaction
mechanisms and will summarize past and present Action activities related to these objectives, and
indicate promising future research activities that have been identified by the Action.
The COST Action BM1309 (European network for innovative uses of EMFs in biomedical
applications, EMF-MED) provides a cooperative framework to support the research on beneficial
biological effects of non-ionizing electromagnetic fields (EMFs) and their use in biomedical
applications. One of the Working Groups (WG) of the Action focuses on “Cancer-EMF interactions
and applications”, where a platform for research dealing with diagnosis and treatment of cancer
using EMF and/or EMF-based technologies has been established. The objectives of the WG are to:
– establish the scientific rationale of cancer treatments based on low level and high level EMFs;
– optimize the administration and control of EMF-based cancer treatment;
– develop and/or improve EMF-based cancer diagnostic modalities;
– develop the associated technology for clinical use.
This presentation has the focus on interaction mechanisms and will summarize past and present
Action activities related to these objectives, and indicate promising future research activities that
have been identified by the Action.

W2-4 [14:30]
Protective effects of non-ionizing radio frequency fields in mammalian cells damaged by
mutagens
Maria Scarfi1
1CNR-Institute for Electromagnetic Sensing of the Environment (IREA), Naples, Italy

Session: W3
Workshop 3: Advanced Approaches for Analysis of Biological Effects of Pulsed Electric
Fields
Thursday June 28, 2018 • 15:30 - 18:00
Europa D
Chairs: Michal Cifra & Lluis M. Mir

W3-1 [15:30]
Workshop content
Michal Cifra1 & Lluis M. Mir2
1Institute of Photonics and Electronics, The Czech Academy of Sciences, Prague, Czech Republic
2Laboratory of Vectorology and Anticancer Therapies, UMR 8203, CNRS, Univ.Paris-Sud, Université Paris-
Saclay, Gustave Roussy, Villejuif, France, 94805
The proposed workshop is dedicated to presentation and discussion of advanced technologies and
methods to study the effects of pulsed electric fields on living cells mainly at the cellular and
molecular levels, for improving the knowledge and applicability of future medical applications. The
focus will be on technologies based on advanced chemiluminescence techniques, advanced linear
and nonlinear optical and imaging techniques (Raman, CARS and THz microspectroscopy),
numerical microdosimetry and molecular dynamics modeling.

W3-2 [15:40]
Endogenous biological chemiluminescence for probing oxidative effects of electric pulses
Michal Cifra1, Djamel E. Chafai1, Daniel Havelka1 & Lluis M. Mir2

347
1Institute
of Photonics and Electronics, The Czech Academy of Sciences, Prague, Czech Republic, 18200
2Laboratory of Vectorology and Anticancer Therapies, UMR 8203, CNRS, Univ.Paris-Sud, Université Paris-
Saclay, Gustave Roussy, Villejuif, France, 94805
Oxidative reactions in biological systems generate endogenous chemiluminescence. We report how
this endogenous light is affected by intense electric pulses. It was found that a train of several short
pulses tends to induce a stronger response of endogenous chemiluminescence from cells than the
single long pulse while total duration and other conditions are the same. We suggest that that the
endogenous biological chemiluminescence can be used for label-free real time monitoring of
biological oxidation initiated by intense electric pulses.
1. Introduction
Living cells are sources of electromagnetic field. From a spectral perspective, low frequency (< few
kHz) part of the cellular electromagnetic spectrum is already extensively explored and known to
arise from membrane channels and pumps activity and consequent transmembrane potential [1].
There is a gap in our knowledge in the radiofrequency and THz region where little is known about
cellular electromagnetic field [1]. In the optical band, it is proved that cells emit electromagnetic field
due to biochemical activity – a continuous flux of light of ultraweak intensity can be detected. Here
we don’t mean a bioluminescence from specialized organisms (such as fireflies) which is visible by
naked eye, but a very general phenomenon of ultraweak chemiluminescence which is present in all
organisms [2].
Ultraweak chemiluminescence (also termed as ultraweak photon emission – UPE) reflects a rate of
oxidation taking place in the biological sample. Therefore, UPE can be used to monitor oxidation.
The main advantage of this method is that it provides real-time information on oxidation and it does
not require the addition of any external labels to biosamples because the signal comes from the
oxidation of the biomolecules endogenously present in the sample – hence the term ”endogenous”
chemiluminescence. The aspect of “endogenousness” also brings in an inherent potential drawback
because the scaling of the signal depends on the biosample composition under study and needs to
be benchmarked a priori or a posteriori with other analytic methods. Furthermore, a typical yield of
chemiluminescence per oxidation of one biomolecule is rather low for an average ensemble of
biomolecules. Therefore, the amount of biomolecules (or cells, proportionally) in the sample has to
be large enough to detect any signal at all.
Nevertheless, as the method of detecting endogenous chemiluminescence from biosamples is real-
time and simple to use and it offers a possibility to monitor oxidation, we decided to test whether it
can be used to monitor oxidation caused by pulsed electric field. Several works showed that the
electric pulses cause oxidation in biological and biomolecular systems [3-7]. Oxidation of
biomolecules is very well known to be caused by the electrochemical processes at the interface with
the electrodes when the voltage is applied, but it has been also found after contactless delivery of
intense electric pulses through e.g. an antenna [8]. Furthermore, electric pulses have been
demonstrated to induce cellular stress and biological production of reactive oxygen species [5-7].
The possibility that electric field can cause or modulate the biological oxidation processes is
potentially of tremendous impact on our understanding of electromagnetic field bioeffects. Oxidative
effects of electric pulses, therefore, have to be carefully monitored both to avoid them when
undesired or to utilize and fine-tune them for beneficial effects such as the stimulation of protective
antioxidative response of the biological system.
2. Methods
The work was focused on the effect of electric pulses applied to DC-3F (Chinese hamster lung
fibroblast line) cells on the endogenous chemiluminescence from the cell suspension. We tested the
luminescence measurement system compatibility with μs-ms pulses delivered by the Cliniporator
device and optimized the set-up. Electrodes and the sample chamber/holder used for sample
treatment had a gap of 5 mm. Hence, for example 1 kV voltage corresponded to 0.2 MV/m (2
kV/cm) voltage to electrode distance ratio. The sample volume was typically 0.5 mL. Cliniporator

348
generator was used to generate and deliver max. 1 kV, 5 ms, and 15 A pulses. Shortest pulse was
100 μs, in a high voltage regime and max. 200 V, 50 ms, and 15 A in a low voltage regime. DC-3F
cells were cultivated in a full medium (MEM + 10% FBS + antibiotics (penicillin + streptomycin)) in
5% CO2, 37 °C and harvested before full confluence by trypsinization and suspended in
electroporation buffer (SHM – sucrose/HEPES/Mg+ ions) for the experiment. The final cell density
for experiment is described lower.
Ultra-sensitive luminescence detection system based on a selected Hamamatsu H7360-01
photomultiplier and light-tight chamber was built in the Institute of Photonics and Electronics, The
Czech Academy of Sciences. Combined experimental setup was comprised of an electric pulse
generator (Cliniporator) connected to the dark light-tight chamber with luminescence detection.
3. Results
The typical chemiluminescence signal from cell suspension after application of electric pulses is in
Fig. 1. At first, we quantified the dependence of the chemiluminescence intensity on the cell number
present in the sample. The exposure parameters were fixed as 20x pulses 0.2 MV/m, 250 μs, and 1
Hz. At least 20-30 million cells are needed to obtain reasonable signal above the baseline.
At second, we analyzed if there is any difference of response due to single long pulse vs. several
short pulses with 1 s gap (both cases have the same total duration). This question is based on the
recently published hypothesis [9] which suggests that time gaps between the pulses will cause
larger surface of the cells to be electropermeabilized because the lipids in the membrane will have
enough time to diffuse along the membrane between the pulses. Here it is important to note that
only those areas of the cell membranes which are oriented towards the electrodes are
permeabilized during the electropermeabilization process. Specifically, we compared single 0.2
MV/m and 5 ms pulse vs. 20x 0.2 MV/m and 250 μs pulses. From five experiments, we found that
the train of 20 pulses seems to elicit a stronger response than a single pulse while both have the
same current dose.
At third, we tested if there is any “memory” effect of previous pulsing. The hypothesis leading to this
question is following: we expect that the first pulsing will irreversibly permeate the membrane of
some fraction of cells leading to release of biomolecules to the extracellular medium thus making
these biomolecules available for oxidation at the electrodes. Then, the second pulsing will oxidize
these free biomolecules leading to chemiluminescence of higher intensity than in the first pulsing
when the molecules were not present in the medium. We have demonstrated in few experiments
that the second pulsing indeed brings about stronger response. However, further tests are needed
clarify if biomolecules are indeed released to a significant extent under our experimental conditions
and if it is the presence of these biomolecules which causes higher chemiluminescence during
second pulsing.
4. Discussion and future challenges
For future work, we aim to achieve a higher sensitivity of the chemiluminescence detection by
decreasing the light detector dark count using a cooled (-30 °C) photomultiplier system available.
Furthermore, endogenous chemiluminescence seems to have significant spectral content in the red
band (>600 nm) [10] where the currently used photomultiplier has a rather low quantum efficiency.
We plan to employ the photomultiplier with a higher sensitivity in the red band and compatibility with
our cooling housing. Higher sensitivity will enable detection of chemiluminescence signals from
smaller number of cells and/or under influence of weaker/shorter electric pulses.
There are also further open questions based on our research. One question is whether the
chemiluminescence signal originates from the oxidation at the electrodes or in the larger volume of
the sample. This question could be verified by imaging the chemiluminescence with ultra-sensitive
cooled EM-CCD camera. Tounequivocally provethe oxidation, we also aim to employ a method for
carbonyl detection using DNPH derivatization, since one of the common products of biomolecule
oxidation is the formation of carbonyl groups
In conclusion, we were able to find the experimental settings under which we can detect the effect of

349
intense pulsed electric field on chemiluminescence. We believe that the results and upcoming
follow-up will help to better understand the oxidative component of electropermeabilization
mechanism to optimize cancer and gene therapy therapeutic protocols in the future.
Acknowledgements
We acknowledge STSM 160117-081320 within COST Action BM1309. We also acknowledge
support from the Czech Science Foundation, project no. P102/15-17102S. Authors also participate
in COST Action CA15211 and bilateral exchange project between Czech and Slovak Academies of
Sciences, no.SAV-18-11. Adeline MUSCAT is acknowledged for culturing the cells, Marie BRETON
for assistance with chemicals and Tomás GARCÍA-SÁNCHEZ for electrode and cuvette fabrication.
References
[1] O. Kucera, K. Cervinkova, M. Nerudova, and M. Cifra. “Spectral Perspective on the
Electromagnetic Activity of Cells.” Current Topics in Medicinal Chemistry 15, no. 6, pp. 513–522.
2015
[2] M. Cifra and P. Pospíšil, “Ultra-weak photon emission from biological samples: Definition,
mechanisms, properties, detection and applications,” Journal of Photochemistry and Photobiology
B: Biology, vol. 139, pp. 2–10, Oct. 2014.
[3] M. Maccarrone, M. R. Bladergroen, N. Rosato, and A. F. Agro, “Role of lipid peroxidation in
electroporation-induced cell permeability,” Biochemical and biophysical research communications,
vol. 209, no. 2, pp. 417–425, 1995.
[4] M. Maccarrone, C. Fantini, A. F. Agrò, and N. Rosato, “Kinetics of ultraweak light emission from
human erythroleukemia K562 cells upon electroporation,” Biochimica et Biophysica Acta (BBA)-
Biomembranes, vol. 1414, no. 1, pp. 43–50, 1998.
[5] M. Maccarrone, N. Rosato, and A. F. Agro, “Electroporation enhances cell membrane
peroxidation and luminescence,” Biochemical and biophysical research communications, vol. 206,
no. 1, pp. 238–245, 1995.
[6] B. Gabriel and J. Teissie, “Generation of reactive-oxygen species induced by
electropermeabilization of Chinese hamster ovary cells and their consequence on cell viability,”
European Journal of Biochemistry, vol. 223, no. 1, pp. 25–33, 1994.
[7] O. N. Pakhomova et al., “Oxidative effects of nanosecond pulsed electric field exposure in cells
and cell-free media,” Archives of Biochemistry and Biophysics, vol. 527, no. 1, pp. 55–64, Nov.
2012.
[8] M. Breton, A. Catrain, R. Vezinet, and Ll. M. Mir, “Chemical study of the effects of UltraWide
Band and Narrow Band signals on membranes,” in Proceedings of BIOEM2016, pp. 17–19.
[9] M. Leguèbe, A. Silve, L. M. Mir, and C. Poignard, “Conducting and permeable states of cell
membrane submitted to high voltage pulses: Mathematical and numerical studies validated by the
experiments,” Journal of Theoretical Biology, vol. 360, pp. 83–94, Nov. 2014.
[10] M. Nerudová, K. Červinková, J. Hašek, and M. Cifra, “Optical spectral analysis of ultra-weak
photon emission from tissue culture and yeast cells,” in SPIE Proceedings, p. 94500O, 2015.
Figures

350
 Figure 1. Typical chemiluminescence signal from cell suspension undergoing pulsed electric
field treatment. Here, 20 pulses (0.2 MV/m, 250 μs) fired at 1 Hz rate, starting from 122s.

W3-3 [16:00]
Molecular models used in the interaction between nanosecond pulses electric fields and
biomolecules
Paolo Marracino1, Elena della Valle1, Maura Casciola2, Micaela Liberti1 & Francesca Apollonio1
1Department of Information Engineering, Electronics and Telecommunications, University Sapienza of Rome,
Rome, Italy
2Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA

When dealing with the interaction of electric fields and bio-systems, the very first mechanism is
observed at molecular level; therefore, the need for accurate and reproducible models able to
describe the time evolution of the target in presence of the external fields are needed.
The aim of computer simulations of molecular systems is to compute macroscopic behavior from
microscopic interactions. A model of the real world is constructed, both measurable and not
measurable properties are computed and the former are compared with experimentally determined
properties. If the model used is validated by the experimental comparison, it could even be used to
predict unknown or not measurable quantities [1-2]. Theoretical methods and computational
approaches applied to reproduce and predict the behavior of molecules at the atomic level are all
part of the general category of molecular modeling. In particular, Molecular Dynamics (MD) is a
simulation methodology often used for studying the conformational rearrangements of molecules
and their interactions with other molecular species in a range of environments. The method provides
a dynamic description of the temporal behavior of atoms and molecules by using finite difference
methods to numerically solve Newton's equations of motion. MD methods produce a so-called
trajectory, i.e. coordinates and velocities, of the simulated set of interacting particles (e.g. atoms) [1].
The behavior and energetics of molecules are fundamentally quantum mechanical. However,

351
quantum mechanical models are computationally demanding with a sever limitation of the system
size and time scale of the simulation. Approximating atomic and molecular interactions with classical
mechanics, based on the classical motion of nuclei allows to increase simulation size and time scale
limits, while preserving accuracy for the description of a number of properties that do not depend on
the electronic distribution in a molecule. In classical MD, the potential energy of and forces acting on
a many body system are calculated from an empirical force field [3].
MD simulations are becoming an increasingly powerful tool for the prediction of molecular
properties: based on the MD trajectories, one can access information related to structural,
dynamical and thermodynamics properties of the modeled system complementary and of support to
experimental observations. The main strength of molecular dynamics simulations is the amount of
detail: all or nearly all atoms are incorporated in the model, and few assumptions are necessary. Its
main weaknesses are the computational cost and the associated limited time and limited length
scale and the approximations inherent in the potential function. The maximum time-step allowing the
stability of the integration of the equations of motion, which limits, up to now, the lengths of current
simulations realistically to about hundreds of nanoseconds, imposes that even the largest systems
currently being handled are of the order of tens of thousands of particles [4-5]. An overall schematic
of how molecular dynamics works is reported in Fig. 1.
In this context, the interaction with external electric fields is modelled as an electrostatic force acting
on each single particle represented in the simulation. In particular, nanosecond pulses electric fields
used in laboratory experiments related to electropermeabilization, due to their time length ranging
from some to hundreds of nanoseconds, are suitable to be modelled in molecular dynamics
simulations.
Therefore, in this work several examples of the interaction between nanosecond pulses electric
fields and biosystems will be presented [6-19]: i) intracellular protein; ii) transmembrane protein; iii)
patch of lipid bilayer; iv) analysis of the pore formation mechanism. The final aim is to attempt to
classify such interaction at molecular level and investigate whether there could be a specificity going
beyond the well-known mechanism for which at its very first stage the electric field couples with
charges and dipoles.
References
[1] M.P. Allen, D.J. Tildesley, Computer Simulation of Liquids, Clarendon Press, 1989
[2] A.R. Leach, Molecular Modelling: Principles and Applications, Pearson Education, 2001.
[3] English, N.J., Waldron, C.J. Perspectives on external electric fields in molecular simulation:
Progress, prospects and challenges (2015) Physical Chemistry Chemical Physics, 17 (19), pp.
12407-12440.
[4] Tamar Schlick, Eric Barth, Rev. Biophys. Biomol. Struct. 26 (1997) 181-222.
[5] Berkowitz, M.L. Biochim. Biophys. Acta 1788 (2009) 86-96
[6] Apollonio et al., 2013, IEEE MTT, 61, 2031-2045.
[7] Apollonio, F., Liberti, M., Amadei, A., Aschi, M., Pellegrino, M., D'Alessandro, M., D'Abramo, M.,
Di Nola, A., D'Inzeo, G. (2008) IEEE Transactions on Microwave Theory and Techniques, 56 (11),
art. no. 4657388, pp. 2511-2519.
[8] Marracino, P., Castellani, F., Vernier, P.T., Liberti, M., Apollonio, F. (2017) Journal of Membrane
Biology, 250 (1), pp. 11-19.
[9] Marracino, P., Liberti, M., Vernier, P.T., Apollonio, F., (2017) RSC Advances, 7 (51), pp.
31997-32007.
[10] Marracino, P., Liberti, M., Trapani, E., Burnham, C.J., Avena, M., Garate, J.-A., Apollonio, F.,
English, N.J. (2016) International Journal of Molecular Sciences, 17 (7), art. no. 1133.
[11] Marracino, P., Liberti, M., d'Inzeo, G., Apollonio, F. (2015) Bioelectromagnetics, 36 (5), pp.
377-385.

352
[12] Casciola, M., Bonhenry, D., Liberti, M., Apollonio, F., Tarek, M. (2014) Bioelectrochemistry, 100,
pp. 11-17.
[13] Reale, R., English, N.J., Marracino, P., Liberti, M., Apollonio, F. (2014) Molecular Physics, 112
(14), pp. 1870-1878.
[14] Marracino, P., Paffi, A., Liberti, M., D'Inzeo, G., Apollonio, F. (2014) European Microwave Week
2014: 44th European Microwave Conference, art. no. 6986555, pp. 798-801.
[15] Marracino, P., Paffi, A., Banno, A., Apollonio, F., Liberti, M., D'Inzeo, G. (2014) 2014 31th URSI
General Assembly and Scientific Symposium, URSI GASS 2014, art. no. 6930070
[16] Reale, R., English, N.J., Garate, J.-A., Marracino, P., Liberti, M., Apollonio, F. (2013) Journal of
Chemical Physics, 139 (20), art. no. 205101
[17] Zanetti-Polzi, L., Marracino, P., Aschi, M., Daidone, I., Fontana, A., Apollonio, F., Liberti, M.,
D'Inzeo, G., Amadei, A. (2013) Theoretical Chemistry Accounts, 132 (11), pp. 1-10.
[18] Reale, R., English, N.J., Marracino, P., Liberti, M., Apollonio, F. (2013) Chemical Physics
Letters, 582, pp. 60-65.
[19] Marracino, P., Apollonio, F., Liberti, M., D'Inzeo, G., Amadei, A. (2013) Journal of Physical
Chemistry B, 117 (8), pp. 2273-2279.
Figures

Figure 1. Scheme of inputs and outputs of MD simulations

W3-4 [16:20]
Speaker 3
Mounir Tarek1
1Centre National de La Recherche Scientifique (CNRS), Université de Lorraine, Nancy, France

353
W3-5 [16:40]
Shared knowledge, gaps and challenges of microdosimetry: Importance of realistic models
at the cellular and subcellular level
Agnese Denzi1, Caterina Merla2, Francesca Apollonio1, Lluis M. Mir3 & Micaela Liberti1
1Department of Information Engineering, Electronics and Telecommunications, University of Rome "La
Sapienza", Rome, Italy, 00184
2SSPT - Division of Health Protection Technologies, ENEA, Rome, Italy
3Gustave Roussy, CNRS, Univ. Paris-Sud, Université Paris-Saclay, Villejuif, France

In the study of the biophysical mechanisms at the basis of the EM field interaction leading to
biomedical applications, one needs also to know the field strength at the microscopic scale to
establish a quantitative relation between the field and the observed effect. Aim of this work is to
establish the shared scientific bases of microdosimetry and main results, and to identify the needs
for future research. In particular, a recent important improvement is represented by taking into
account the irregularly shaped cells and internal organelles.
Introduction
Electromagnetic microdosimetry is the quantitative study of the spatial, temporal, and spectral
distributions of EM fields imparted to cellular and subcellular biological structures [1].
In the study of the biophysical mechanisms at the basis of the EM field interaction leading to
biomedical applications, one needs also to know the field strength at the microscopic scale to
establish a quantitative relation between the field and the observed effect [1]. Microdosimetric
techniques are needed for calculating the real distribution of the field on sub-cellular structures, such
as cell membranes and organelles [2]-[5]. Many papers have been published in these last years on
this topic, and a great impulse came also from the key role that microdosimetry plays in some
diagnostic techniques based on dielectric spectroscopy and in the estimation of the transmembrane
potential in electroporation of cells and tissues [6]-[9].
Literature Overview
To approach the microdosimetric problem, two fundamental steps are needed. The first one involves
the setup of a proper dielectric and geometrical cell model. Typically, the biological cell is considered
as a multilayered structure. The most common choice is the three-layered cell [2]–[11]: cytoplasm,
plasma membrane, and extracellular medium. Some papers consider also simplified intracellular
compartments as endoplasmic reticulum, nucleus or bound water layers [12]-[14].
As for the dielectric model, main attempts are in order to dielectrically characterize the plasmatic
membrane, which is generally considered one of the main EM interaction targets. The initial models
adopted considered a frequency-independent behavior of the cell membrane, signifying a pure
capacitive behavior [2], [10], [11], [15]. Recently, dispersive models have been considered
unavoidable to study CW EM fields with frequency higher than hundreds of MHz and nanosecond
pulsed electric fields [3]-[5], [8], [9], [12]-[14]. Some authors focused also on cytoplasm
characterization [16].
The second step concerns the setup of appropriate EM solutions of the cell environment. Almost all
authors choose a quasi-static approach. Indeed, as the theoretical problem is addressed at the
single-cell level, in the RF range, the involved wavelengths are always greater than some
millimeters and the cell radius is of the order of tens of µm. The frequency dependence of the
induced field, when necessary, is usually taken into account through the frequency dependence of
the complex permittivity of the cellular compartments. Three main kinds of solutions exist: circuit
models, analytical solutions applied to simplified cell shapes, such as spherical and spheroidal multi-
shell models, and numerical methods applied to irregular-shaped cells [2]–[19].
Some general consensus exists that up to few kHz the cell membrane insulates quite well the cell
interior with respect to the external applied electric field; at higher frequencies such a capacitive
effect decreases and the electric field penetrates inside the cell and reaches internal organelles.

354
One can roughly say that around 100 V/m at low frequencies and 10 kV/m at high frequencies are
necessary to induce values of transmembrane potential (TMP) comparable to the ones of normal
cell functions (mV). For what concerns the high frequency components it is possible to say that
membrane dielectric values influence inversely the induced TMP: the lower the membrane
permittivity, the higher the TMP. At high frequencies, dielectric modeling is predominant in
determining E field intensities also when dealing with realistic cell shapes [4], [9], [10], [15].
In particular, the use of ultrashort pulsed electric fields has become a promising application in
cancer therapy thanks to their effect of membrane electroporation with a formation of pores and a
consequent increase in membrane permeability [20]. Usually, pulses with duration in the range of μs
and amplitude of kV/m (causing “classical” electroporation) are used to permeabilize the plasma
membrane in order to permit the uptake by the cells of drugs and non-permeant molecules [21]. The
shorter pulses, with duration of nanoseconds and amplitude of MV/m, due to their higher frequency
content, are used also to permeabilize the internal structures (such as mitochondria and
endoplasmic reticulum) [22], [23]. Hence, microdosimetric models that include internal organelles
like the endoplasmic reticulum, have assumed increasing importance.
To obtain microdosimety model representative of the real experiments, the starting point is to
consider images obtained with an optical microscope using different staining techniques able to
highlight different cell compartments [23].
Successively, commercial codes or custom scripts can be developed to recognize cell and interior
borders at the end the images must be transformed in vector ones and imported in an
electromagnetic simulator.
At this point a numerical model can be realized based on the dielectric properties of a cell like in
[24]. In Fig. 1 is reported the scheme on how we build microdosimetry.
Observables are the classical of electromagnetic dosimetry (i.e. Electric Field, Electric Current
Density) plus the Transmembrane Potential on the plasma membrane and the ones of the internal
compartments of interest.
It is worth to note that we must also stress that the induced TMP can be calculated in linear regime
as long as one can assume that the membrane is intact and its conductance is low. When the
conductance increases due to electroporation, the membrane partially discharges through the
conductive pathways, i.e., pores. This consequently reduces the TMP where electroporation
occurred, increasing the field that penetrates the cell that consequently can porate the internal
membranes. This can be taken into account with theoretical models which considered dependence
of the membrane conductivity on the electric field.
Conclusions
Aim of this work is to establish the shared scientific bases of microdosimetry and main results, and
to identify the needs for future research. In particular, a recent important improvement is
represented by taking into account the irregularly shaped cells and organelles. In spite of the huge
work done in literature, some open issues are still alive and represent key challenges in the
development of microdosimetry. The first one is represented by the microthermal issue. Second
important point is represented by the computational/numerical difficulty to relate the microscopic
field to the field at the mesoscopic scale where clusters of cells are to be considered. Finally, the
importance of molecular details like the structure of the membrane bilayer, the anisotropy, the
presence of solvated shells, the water at interface should be better investigated.
Major challenge is represented by the capability to solidly validate microdosimetric results with
reliable experimental works.
References
[1] Apollonio et al., 2013, IEEE MTT, 61, 2031-2045.
[2] Liu and Cleary, Bioelectromagn., vol. 16, pp. 160–171, 1995.
[3] Kotnik and D. Miklavcic, Bioelectromagn., vol. 21, pp. 385–394, 2000.

355
[4] Liberti, et al., IEEE Antennas Wireless Propag. Lett., vol. 8, pp. 865–868, 2009.
[5] Merla et al. IEEE Trans. Microw. Theory Techn., vol. 58, no. 3, pp. 691–698, Mar. 2010.
[6] Smith et al. IEEE Trans. Plasma Sci., vol. 34, no. 4, pt. 2, pp. 1394–1404, Aug. 2006.
[7] Joshi et al. IEEE Trans. Plasma Sci., vol. 32, no. 4, pp. 1677–1686, Aug. 2004.
[8] Merla et al., IEEE Trans. Biomed. Eng, vol. 58, no. 5, pp. 1294–1302, 2011.
[9] Merla et al., IEEE Trans. Biomed. Eng., vol. 49, no. 8, 2302 - 2311, 2012.
[10] Martin et al. Bioelectromagn., vol. 25, pp. 631–633, 2004.
[11] Foster, Schwan, Handbook of Biological Effects of Electromagnetic Fields, 1986, pp. 27–9
[12] Denzi et al., J Membrane Biol., 246(10):761-767, 2013.
[13] Simenova and Gimsa, Bioelectromagnetics, vol. 27, pp. 652–666, 2006.
[14] Kotnik and D. Miklavcic, IEEE Trans. Biomed. Eng., vol. 47, no. 8, 2000.
[15] Munoz et al. Phys Med Biol 48: 1649–1659, 2003.
[16] Simenova et al., Bioelectrochemistry, vol. 56, pp. 215–218, 2002.
[17] Kotnik and D. Miklavcic, Biophys. J., vol. 90, pp. 480–491, 2006.
[18] Maswiwat et al., Bioelectrochemistry, vol. 74, pp. 130-141, 2008.
[19] Merla et al., European Conference on Antennas and Propagation, Eucap 2012.
[20] Breton, M., & Mir, L. M. (2012). Bioelectromagnetics, 33(2), 106-123.
[21] Cadossi, R. et al., Future Oncology, 10(5), 877-890.
[22] de Menorval etal., Scientific Reports. 6, 32331, (2016).
[23] Azan A et al. Advs Anatomy, Vol. 227, Transport Across Natural and Modified Biological
Membranes and its Implications in Physiology and Therapy, 2017.
[24] Hanna, H., Denzi, A., Liberti, M., Andre, F.M., Mir, L.M. (2017) Scientific Reports, 7 (1), art. no.
13079.
[25] Denzi, A., Hanna, H., Andre, F.M., Mir, L.M., Apollonio, F., Liberti, M. (2017) SMACD 2017 - art.
no. 7981606.
Figures

Figure 1. Scheme on how to build Microdosimetry

356
W3-6 [17:00]
Characterization of the interaction between pulsed electric fields and live cells by confocal
Raman microspectroscopy
Lluis M. Mir1, Antoine Azan1, Valérie Untereiner2, 3, Caterina Merla1, 4, Cyril Gobinet2, Marie Breton1, Ali
Tfayli5 & Olivier Piot2, 3
1Vectorology and Anticancer Therapies, UMR 8203 CNRS, Univ. Paris-Sud, Gustave Roussy, Université
Paris-Saclay, Villejuif, France, 94805
2MeDIAN, Biophotonics and Technologies for Health, MEDyC, UMR 7369 CNRS, University of Reims
Champagne-Ardenne, Reims, France, 51096
3Cellular and Tissular Imaging Platform PICT, Faculty of Pharmacy, University of Reims Champagne-
Ardenne, Reims, France, 51096
4Laboratorio Tecnologie Biomediche, ENEA, C.R. Casaccia, Roma, Italy, 00123
5Laboratoire de Chimie Analytique, Univ. Paris Sud, Châtenay-Malabry , Italy, 92296

Part of the workshop on "Advanced approaches for analysis of biological effects of pulsed electric
fields", this presentation deals with the characterization of the interaction between pulsed electric
fields and live cells by confocal Raman microspectroscopy.
The main consequence of the interaction between pulsed electric fields and biological cells is the
destabilisation of the plasma membrane. This leads to the cell membrane permeabilization (allowing
the membrane crossing by non-permeant molecules). This bio-electromagnetic interaction, termed
“electropermeabilization”, or “electroporation”, has led to many applications in food industry and
medicine. Although electropermeabilization has been known for decades, in our team its underlying
mechanisms are under investigation because the phenomenon is still not fully understood.
Recent studies are promoting chemical modifications as the initiating mechanism of the membrane
destabilization. Thus, confocal Raman microspectroscopy can be used to study, in a non-invasive
and label-free manner, the cell electropermeabilization process from a molecular point of view.
Raman signatures of control and pulsed live human adipose-derived mesenchymal stem cells were
acquired. The effect of 100 µs electric pulses on the biochemical composition of live cells could then
be investigated. Spectra were acquired on attached cells: due to the large size of these cells,
spectra could be acquired at different places of the cell, to investigate if the vectoriality of the
porating effects of the electric pulses can be detected using confocal Raman spectroscopy.
Vibrational modes of proteins (phenylalanine and amide I) and lipids were found to be modified by
the electropermeabilization process with a statistically significant difference. In particular, the relative
magnitude of four phenylalanine peaks was modified showing that cell electropermeabilization
affects not only the lipid bilayer but also the proteins [1]. Different regions of interest around the
nucleus of the cells were investigated as well as the dose-effect relationship related to different
electric pulse parameters. Membrane resealing was monitored by pulsing the cells in reversible or
irreversible electropermeabilization conditions at different temperatures. The chemical changes
were similar in different locations around the cell nucleus. By sweeping the field magnitude, the
number of electric pulses or their repetition rate, the Raman signatures of live cells appeared to be
related to the electropermeabilization state, verified by Yo-Pro-1 uptake [2]. More details will be
presented.
References
[1] A. Azan, V. Untereiner, C. Gobinet, G. D. Sockalingum, M. Breton, O. Piot and L. M. Mir.
Demonstration of Protein Involvement in Living Cell Electropulsation using Confocal Raman
Microspectroscopy. Scientific Reports 7. 297–306, 2017. doi:10.1038/srep40448.
[2] A. Azan, V. Untereiner, L. Descamps, C. Merla, C. Gobinet, M. Breton, O. Piot and L. M. Mir.
Comprehensive Characterization of the Interaction between Pulsed Electric Fields and Live Cells by
Confocal Raman Microspectroscopy. Analytical Chemistry 89, 10790-10797, 2017.

357
W3-7 [17:20]
A wide-field CARS setup for spectroscopy under electropulsation of biological media
Caterina Merla1, 2, Brigitte Attal-Trétout3, Michaël Scherman3, Martina Nardoni4, Amina Ghorbel1, Micaela
Liberti5, Francesca Apollonio5, Stefania Petralito4 & Lluis M. Mir1
1CNRS UMR 8203, Gustave Roussy, Univ. Paris-Sud, Université Paris Saclay, Villejuif, France, 94805
2ENEA SSPT Division of Health Protection Technologies, Rome, Italy, 00123
3Onera BP 80100, Palaiseau Cedex, France, 91123
4Department of Pharmaceutical Technologies, "Sapienza" University of Rome, Rome, Italy, 00195
5Department of Information Engineering Electronic and Telecommunications, "Sapienza" University of Rome,
Rome, Italy, 00184
To understand basic phenomena of electropulsation of biological media, a new experimental setup
is used. It combines a wide field CARS microscope with a wide band coplanar waveguide. This
setup allows to acquire CARS hyper-spectra of specific Raman bands of small unilateral vesicles
(i.e. liposomes) and cells contemporary exposed to high amplitude ultra-short (microsecond or
nanosecond duration) electric pulses. Being the physico-chemical bases of electropermeabilization
still debated, our setup allows the experimental assessment of the role of water molecules and
phospholipids during and after the delivery of the electric pulses.
Introduction
To deep more insight into basic phenomena occurring during and after electropulsation of biological
media, a new experimental setup is used. It combines a wide field CARS microscope [1] with a wide
band coplanar waveguide (CCPW) [2]. This setup allows to acquire CARS hyper-spectra of specific
Raman bands (from 2750 to 2950 cm-1 into the lipid vibration region and from 2900 to 3500 cm-1
into the so called water vibration region) of small unilateral vesicles (i.e. liposomes) and cells
contemporary exposed to high amplitude ultra-short (microsecond or nanosecond duration) electric
pulses. Being the physico-chemical bases of electropermeabilization still debated, our setup allows
the experimental assessment of the role of water molecules and phospholipids during and after the
delivery of the electric pulses.
The cells (human adipose mesenchymal stem cells HaMSC, murine fibroblasts DC-3F, and
erythrocytes) and liposomes visibility between the central and lateral (ground) electrodes of the
ground closed coplanar waveguide (GCCPW) was assured by the use of a transparent substrate
(quartz or CaF2) (see Fig. 1 for the experimental arrangement). The inter electrodes gap was fixed
to 0.5 mm in order to allow the CARS pump and Stokes laser beams (diameters of around 100 µm)
alignment and focusing into the biological solution in a controlled way. The biological solution was
contained in a PDMS holder with a total volume of 30 µL. The illumination scheme followed a non-
phased-matched geometry as suggested in [1]. The pump laser axis was kept parallel to the
microscope objective axis, while the Stokes beam axis was tilted by 7° (Fig.1) to efficiently attenuate
the non-resonant signal. An inverted microscope (Zeiss Axiovert 200) was used and the sample was
imaged by means of a 50x, NA=0.55 objective. Image of the observation plane was formed on an
intensified CCD camera (PIMAX 3 Camera, Roper Technologies, Sarasota, Florida), as shown in
Fig.1. The bandpass filters (F1 CVI-CP-AG-540 and F2 Semrock, FF01-534/42 respectively) were
placed in order to reflect the pump and Stokes beams right after interacting with the sample of
interest. Additionally, a further bandpass filter (F3 = F2) was placed at the exit of the objective in
front of the CCD camera (F3 in Fig.1) to block residual light. The camera was triggered by the laser
pulse using a delay generator (DG 545, Princeton instruments) with a temporal gate of 40 ns.
To verify the possibility of focusing the pump and Stokes beams between the gold electrodes of the
CCPW (see Fig. 1), CARS spectra of a phosphate buffer saline (PBS) solution (used to suspend
cells and liposomes) were analyzed at first. Spectra of PBS solution were acquired under
unexposed and exposed conditions (8 electric pulses of 100 µs and a field amplitude of 1200 V/cm
at a repetition frequency of 1 Hz). These spectra were retrieved from a set of 146 images recorded

358
by scanning the Stokes frequency between 2915 and 3453 cm-1. Different pump and Stokes
energies were used to establish precise operational limits for the subsequent analyses.
Results
the CARS spectra of the PBS solution in unexposed and exposed conditions are presented. The
CARS spectra of unexposed PBS shows a main peak at 3215 cm-1 while the exposed buffer
presents a slightly shifted main peak at 3135 cm-1. This difference can be related to slightly different
organization of water molecules at the interface of the setup substrate. Energy for pump and Stokes
that guarantees the preservation of the gold GCCPW electrodes are below the 300 µJ and 200 µJ
for the two beams respectively. Thanks to these results, CARS images and spectra acquisitions of
unexposed and exposed HaMSC, DC-3F, and erythrocytes have been acquired. Liposomes
solutions in exposed and unexposed conditions were also successfully imaged by our CARS setup.
Cells and liposomes were exposed to both microsecond (100µs) and nanosecond (10 ns) electric
pulses with different amplitudes, pulse number, and repetition frequencies. Final results, presented
during the workshop entitled “Advanced technologies for analysis of biological effects of pulsed
electric fields” will provide further light on basic mechanisms of electropulsation in the microsecond
and nanosecond pulse regimes.
Acknowledgments
This project has received funding from the European Union’s Horizon 2020 Research and
Innovation Program under Marie Sklodowska-Curie Action grant agreement No. 661041 OPTIC
BIOEM. Work performed also in the frame of the LEA (European Associated Laboratory) EBAM
(Pulsed Electric Field Applications in Biology and Medicine).
References
1. A. Silve et al., J Raman Spectroscopy, 2012, 43, 644-653.
2. C. Merla et al., IEEE Proceeding of EuCap 2017 Conf., Paris 19-24 March, 2017.
3. Koroteev, N. I.; Endemann, M.; Byer, R. L. Physical Review Letters 1979, 43, 39
Figures

359
 Figure 1. The experimental setup: a schematic representation.

W3-8 [17:40]
Terahertz spectroscopy to monitor the leakage of molecules from live cells exposed to
pulsed electric fields: comparison to fluorescent video-microscopy experiments
Tomás García-Sánchez1, Antoine Azan1, Guilhem Gallot2 & Lluis M. Mir1
1Vectorology and Anticancer Therapies, UMR 8203, CNRS, Univ. Paris-Sud, Gustave Roussy, Université
Paris-Saclay, 94805 Villejuif, France
2Laboratoire d’Optique et Biosciences, Ecole Polytechnique, UMR7645, CNRS, Route de Saclay, 91128,
Palaiseau, France
For the first time, a Terahertz-Attenuated Total Reflection (THz-ATR) experimental device was used
to investigate the effects of electroporation electric fields on live cells. The THz signal, generated
and detected by a classical Terahertz time-domain spectroscopy (THz-TDS) setup, was compared
to fluorescence microscopy data (Yo-Pro1 uptake).
Abstract:
For the first time, a Terahertz-Attenuated Total Reflection (THz-ATR) experimental device was used
to investigate the effects of electroporation electric fields on live cells. The interaction of this type of
radiation, between the microwave and the mid-infrared (10 GHz) < f < 10 THz), with cells can bring
interesting information about the changes in composition of cells after electric field pulse application.
The THz signal was generated and detected by a classical Terahertz time-domain spectroscopy
(THz-TDS) setup. The THz signal can be used as a label-free biomarker of the amount of proteins

360
and other molecules inside the cells. The THz and fluorescence signals of MDCK cells were
recorded before and after the delivery of the electric pulses.
In our experiments we showed how the THz signal was sensitive to the presence of cells (MDCK
cell line). The relative THz signal, calculated with respect to pre-pulse conditions, decreased after
the delivery of the electric pulses. This relative signal decrease was associated to the leakage of
molecules, ranging from small molecules to large ones, such as proteins, from the cell inside due to
the membrane electropermeabilization. In order to compare the THz and the fluorescence
techniques, time-lapse fluorescent images of MDCK cells, exposed under the same conditions to
the same electric pulses, were acquired in the presence of Yo-Pro-1. The uptake of Yo-Pro-1 into
cells was assessed in order to quantify the dynamics of the electropermeabilization process. The
evolution of the normalized THz signal after the delivery of the electric pulses was fitted to a
mathematical model based on an exponential function. Similarly, the dynamics of fluorescence
microscopy results were also fitted to an exponential model.
Our results indicate that the detection threshold of the THz setup was lower than the detection
threshold of fluorescence at low electric fields. When increasing the electric field magnitude, a dose
effect on the THz and fluorescence results was noticed. Some discrepancies were found in the
dynamics obtained from both methods.

361
 Session: FA
Student Flash Poster Session A
Monday June 25, 2018 • 13:30 - 14:30
Emerald
Chairs: Christopher Portier & Niels Kuster

FA-1 [13:30]
STUDENT PAPER
Extremely low frequency-electromagnetic radiation deteriorated the sleep reduction in
Drosophila melanogaster under heat stress
Xiaomei Huang1, 2, Ziyan Zhang1, Hongying Zhang1, 2, Chuanjun Yang1, Yongyan Sun1, 2, Chao Tang1 & Hui
Yu1
1Key Laboratory of Urban Environment and Health, Institute of Urban Environment (IUE), Xiamen, China,
361021
2Chinese Academy of Sciences, Beijing, China, 100049
Keywords: Behavioural, ELF/LF, Completed (unpublished)
Presented by: Xiaomei Huang
The coupling effects of heat stress and extremely low frequency-electromagnetic field (ELF-EMF, 50
Hz, 3 mT) on sleep and activity were studied using Drosophila (wild-type Canton-Special (CS) and
mutant W1118 flies) as a model. The mRNA expressions of circadian clock genes, neurotransmitters
related genes, as well as important neurotransmitters were analyzed. Result showed that ELF-EMF
deteriorated the sleep reduction in Drosophila under heat stress.
Published under Poster Session A: PA-3

FA-2 [13:33]
STUDENT PAPER
Numerical determination of the risk of cardiac ventricular fibrillation in humans from body
currents with frequencies up to 1 MHz using outcomes of animal studies
Pia Schneeweiss1, Tobias Theiler1, Kai Jagielski1 & Dominik Stunder1
1Research Center for Bioelectromagnetic Interaction (femu), RWTH Aachen University, Aachen, Germany,
52074
Keywords: Dosimetry (computational), All Frequencies, Work in Progress
Presented by: Pia Schneeweiss
Cardiac ventricular fibrillation (CVF) is a potential cause of death following electrical accidents
occurring when the electrical activity of the heart is disrupted. In order to determine thresholds for
CVF, cadaver and animal studies were carried out in the past, which are however limited e.g. in
tissue’s electric properties and the utilized probes properties. Currently executed simulations with
high resolution anatomical body models traversed by body currents with frequencies up to 1 MHz
will be used for the derivation of CVF thresholds and can therefore help to prevent injuries caused
by electrical currents in the future.
Published under Poster Session A: PA-21

FA-3 [13:36]
STUDENT PAPER
Industrial indoor massive MIMO human EM-exposure evaluation
Sergei Shikhantsov1, Arno Thielens1, 2, Gunter Vermeeren1, Emmeric Tanghe1, Piet Demeester1, Guy Torfs1,

362
Luc Martens1 & Wout Joseph1
1Department of Information Technology, Ghent University/IMEC, Ghent, Belgium, 9000
2Department of Electrical Engineering and Computer Sciences, University of California Berkeley, Berkeley,
CA, USA, 94704
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Sergei Shikhantsov
We present a numerical method of estimating human exposure to the electromagnetic fields (EMF)
of a massive MIMO base station (BS) in an industrial indoor environment e.g., an assembly line or a
warehouse. The method relies on a massive MIMO channel model derived from ray-tracing
simulations with stochastically generated environments. Incoming rays at the receiver location are
used as the input for Finite-Difference Time-Domain plane-wave simulations with a realistic human
phantom to obtain EM-exposure estimate.
Published under Poster Session A: PA-23

FA-4 [13:39]
STUDENT PAPER
MRI patient exposure - characterization and sequence-comparisons
Jennifer Frankel1, Kjell Hansson Mild1 & Jonna Wilen1
1Department of Radiation Sciences, Umeå University, Umeå, Sweden, 90187
Keywords: Dosimetry (measurements), ELF/LF, Completed (unpublished)
Presented by: Jennifer Frankel
The unique mixture and intensity of magnetic fields experienced by the patient during an MRI exam
are quite complex from an exposure perspective. We measured the RF- and gradient magnetic
fields inside a 3T MRI scanner and extracted several exposure parameters to give a comprehensive
description of the exposure of each sequence. We found significant differences between sequences
for each of the exposure parameters studied. We also found within-sequence variation. This means
that MRI exposure is not only sequence-specific but also specific to the individual patient and exam
occurrence, a complexity which must be kept in mind when performing exposure assessment for
epidemiological studies on MRI.
Published under Poster Session A: PA-37

FA-5 [13:42]
STUDENT PAPER
Measurement and health assessment of electromagnetic fields by electric vehicles during
road traffic operation using driving characteristics
Kai Jagielski1, Pia Schneeweiss1, Dominik Stunder1 & Thomas Kraus1
1Research Center for Bioelectromagnetic Interaction (femu), RWTH Aachen University, Aachen, Germany,
52074
Keywords: Dosimetry (measurements), Pulsed, Work in Progress
Presented by: Kai Jagielski
This contribution introduces a measurement system which enables magnetic field measurements in
the frequency range between 1 Hz and 400 kHz on electric vehicles during vehicle operation. The
additional recording of the driving style, with acceleration sensor, gyroscope and GPS module,
allows a detailed investigation of the influence of various vehicle components on the magnetic field
spectrum. First measurement results of electric cars and the evaluation possibilities are presented.
In future, measurements will be carried out for example at wireless charging systems or electric
buses.

363
Published under Poster Session A: PA-41

FA-6 [13:45]
STUDENT PAPER
Enhancement of in vitro gene electrotransfer using gold nanoparticles
Tjasa Potocnik1, Tina Batista Napotnik1, Tamara Pezić1, Matej Reberšek1 & Damijan Miklavčič1
1University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia, 1000
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Tjaša Potočnik
Gold nanoparticles when added to cell suspension prior to electroporation significantly increased
transfection rate using monopolar pulses in vitro. When bipolar pulses were used for gene
electrotransfer, the presence of gold nanoparticles had no effect on transfection rate. Our results
show the possibility of providing sufficient transfection rate with lower voltage applied to cells in
presence of gold nanoparticles.
Published under Poster Session A: PA-65

FA-7 [13:48]
STUDENT PAPER
Repeated long term exposure of Mesenchymal Stem Cells to electric fields using a new pulse
generator
Shirmone Botha1, Borja Lopez2, Óscar Lucía2, Héctor Sarnago2, Alejandro Naval2, José-Miguel Burdio2,
Adeline Muscat1, Tomás García-Sánchez1, Franck Andre1 & Lluis M. Mir1
1Laboratory of Vectorology and Anticancerous Therapeutics UMR8203, Institut Gustave Roussy, Paris,
France, 94805
2Department of Electronic Engineering and Communications, University of Zaragoza, Zaragoza, Spain,
50018
Keywords: Human, Pulsed, Work in Progress
Presented by: Shirmone Botha
In mesenchymal stem cells (MSCs), numerous studies describe the spontaneous oscillations of
cytosolic calcium at various concentrations. We demonstrate that the application of one high voltage
micro second pulse disrupts natural calcium oscillations making it possible to insert an additional
synthetic calcium oscillation pattern over several hours. This study strengthens the perception that
pulsed electric fields can be used as a tool to manipulate calcium oscillations in human adipose
derived MSCs using a novel pulse generator as the one specifically developed. These
manipulations might be used as a tool to improve MSC engineering.
Published under Poster Session A: PA-73

FA-8 [13:51]
STUDENT PAPER
A novel method for estimation of brain tissue electrical conductivity: from in silico to in
clinico results
Andres Carvallo1, Julien Modolo1, Pascal Benquet1 & Fabrice Wendling1
1Univ Rennes, INSERM, LTSI - U1099, Rennes, France, 35000
Keywords: Human, ELF/LF, Completed (unpublished)
Presented by: Andres Carvallo

364
OBJECTIVE: Estimating brain tissue electrical conductivity using low-intensity pulse stimulation.
METHODS: We derived an analytical model of the electric potential generated by SEEG electrodes.
We coupled the electric potential with an electrode-electrolyte interface model to i) derive an
analytical expression of brain tissue response to biphasic pulses and ii) estimate conductivity.
RESULTS: We validated our biophysical model using i) saline solutions calibrated for electrical
conductivity, ii) rat brain tissue, and iii) intracerebral electrophysiological data recorded in epileptic
patients during pre-surgical evaluation. CONCLUSIONS: Rapid and reliable brain tissue electrical
conductivity estimation is achieved using the proposed method.
Published under Poster Session A: PA-75

FA-9 [13:54]
STUDENT PAPER
Local thermal dosimetry applied to in vitro studies at millimetre waves
Rosa Orlacchio1, Maxim Zhadobov1, Stanislav Alekseev2, Ronan Sauleau1, Yann Le Page3 & Yves Le
Dréan3
1Institut d'Électronique et de Télécommunications de Rennes (IETR), University of Rennes 1, Rennes,
France, 35042
2Institute of Cell Biophysics, Russian Academy of Sciences, Pushchino, Russian Federation, 142290
3Institute for Research on Environmental and Occupational Health (IRSET), University of Rennes 1, Rennes,
France, 35042
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Rosa Orlacchio
This study investigates the role of thermal convection arising from high SAR gradients during in vitro
exposure to millimetre waves (MMW). The impact of the liquid volume and exposure duration on
initiation of convective currents is investigated. Water and a convection-free medium with water-
equivalent electromagnetic properties have been exposed to continuous wave (CW) and amplitude
modulated MMW. Continuous and pulsed heating were recorded using a microthermocouple. Our
results suggest that convection plays an important role in MMW-induced heating in in vitro
experiments, and its initiation depends both on the volume of the exposed liquid and the duration of
exposure.
Published under Poster Session A: PA-91

FA-10 [13:57]
STUDENT PAPER
Effect of high frequency electric field on growth factor of cultured osteoblasts
Shunsuke Yamaguchi1, Masatake Akutagawa1, Hiromichi Yumoto2, Kouji Hirao2, Takahiro Emoto1, Hiroo
Tarao3, Toshihiko Tominaga4, Toshitaka Ikehara5 & Yohsuke Kinouchi1
1Department of Electrical and Electronic Engineering, Tokushima University , Tokushima, Japan, 770-8506
2Department of Periodontology and Endodontology, Tokushima University , Tokushima, Japan, 770-0042
3Department of Electrical & Computer Eng., National institute of technology Kagawa colllege, Takamatsu,
Japan, 761-8058
4Tominaga Dental Clinic, Naruto, Japan, 771-0360
5Institute for Health Sciences, Tokushima Bunri University, Tokushima, Japan, 770-8514
Keywords: In vitro, IF, Work in Progress
Presented by: Shunsuke Yamaguchi
In this experiment, high frequency AC voltage was applied to cultured osteoblasts. The voltage
waveform used is a sinusoidal wave of 500 kHz 20 V. Temperature, voltage, osteoblastic VEGF,

365
VEGFmRNA and ALP were measured. VEGF and VEGFmRNA were measured 1 day after voltage
application and ALP was measured 7 days after. Durations of voltage application were 30, 60, 90
and 120 s. As results of them, it is obvious that it affects the osteogenesis of osteoblasts by
electrical factors. Although it depends on the initial state (cycle) of the cells, VEGF was significantly
increased at 90 seconds and VEGFmRNA at 120 seconds. Change of ALP by application of the
elecrical field was not observed in our experiment.
Published under Poster Session A: PA-99

FA-11 [14:00]
STUDENT PAPER
Effect of long-term RF-EMF exposure on microglia activation in 5xFAD mice
Ye Ji Jeong1, 2, Hyung-Do Choi3, Jeong-Ki Pack4, Nam Kim5, Yun-Sil Lee6 & Hae-June Lee1
1Division of Basic Radiation Bioscience, Korea Institute of Radiological and Medical Sciences, Seoul, Korea
2Department of Life Sciences, Korea University, Seoul, Korea
3Department of EMF Research Team, ETRI, Daejeon, Korea
4College of Engineering, Chungnam National University, Daejeon, Korea
5School of Electrical and Computer Engineering, Chungbuk National University, Cheongju, Korea
6Graduate School of Pharmaceutical Sciences, Ewha Womans University, Seoul, Korea
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Ye Ji Jeong
To explore the long-term RF-EMF effect on Alzheimer’s disease, we performed 6 months RF-EMF
exposure to 5xFAD mice (SAR 5W/kg, 2 h/day, 5 days/week) and investigated behavioral changes
and neuroinflammatory response. 5xFAD mice exposed to long-term RF-EMF showed mild
decrease in Aβ deposition, memory improvement, and decrease in expressions of Iba-1 and
microglia regulator genes compared to sham exposed group. Our finding suggests possible
mechanism of beneficial effect of long-term RF-EMF on AD mice.
Published under Poster Session A: PA-109

FA-12 [14:03]
STUDENT PAPER
Steepest-entropy-ascent quantum thermodynamic approach to scaling the electric field
parameters and criticality related to cell signaling in electrically perturbed cells:
Experimental evidence and rationale
Ishan Goswami1, Scott S. Verbridge2 & Michael R. von Spakovsky1
1Mechanical Engineering, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA,
24061
2Biomedical Engineering and Applied Mechanics, Virginia Polytechnic Institute and State University,
Blacksburg, Virginia, USA, 24061
Keywords: Mechanistic/Theoretical, Pulsed, Work in Progress
Presented by: Ishan Goswami
A successful coupling of standard cancer treatment modalities with electric field-based tumor
ablation techniques requires a general understanding of the mechanism of action of each modality
to evaluate synergistic effects for better treatment efficacies. The caveat, however, is that the
electric field parameter space and the molecular pathways are so large that understanding the
mechanism of action of electric fields via exploring all possible combinations is prohibitive from a
cost and time standpoint. To solve this parameter space problem, this work introduces a new
computational approach that takes advantage of thermodynamic scaling laws applicable to bio-

366
membranes. We report experimental evidence and the rationale behind this approach.
Published under Poster Session A: PA-121

FA-13 [14:06]
WITHDRAWN

367
 Session: FB
Student Flash Poster Session B
Tuesday June 26, 2018 • 13:30 - 14:30
Emerald
Chairs: Christopher Portier & Niels Kuster

FB-1 [13:30]
STUDENT PAPER
Development and evaluation of microwave antenna for transcatheter denervation
Shohei Matsuhara1, Kazuyuki Saito1, Hiroshi Kawahira1, Nobuyoshi Takeshita2 & Tomoyuki Tajima3
1Chiba University, Chiba, Japan
2National Cancer Center, Kashiwa, Japan
3The University of Tokyo Hospital, Tokyo, Japan
Keywords: Clinical (therapy), RF/Microwaves, Work in Progress
Presented by: Shohei Matsuhara
In this study, a catheter, which ablates extravascular nerve by the thermal effect of microwave, is
introduced. In addition, heating characteristics of the catheter are evaluated by the numerical
calculation and the experimental investigation. As a result, temperature distributions around the
catheter were revealed and the usefulness of a denervation catheter using microwave was
suggested.
Published under Poster Session B: PB-10

FB-2 [13:33]
STUDENT PAPER
Numerical analysis of a three-compartment head model subjected to variation of input
parameters
Anna Šušnjara1, Mario Cvetkovic1, Dragan Poljak1 & Hrvoje Dodig2
1Department of Electronics and Computer Science, University of Split, Split, Croatia, 21000
2Department of Naval Electronics and Information Technology, University of Split, Split, Croatia, 21000
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Anna Šušnjara
A stochastic framework for the assessment of the induced electric field in the three-compartment
model of human head is presented. The relative permittivity and conductivity of scalp, skull and
brain are modelled as random variables. The statistical moments are calculated by means of a
stochastic collocation method. The sparse grid interpolation in the multidimensional random space
resulted in a small number of simulations. The sensitiy analysis of input parameters indicate the
higher overall impact of relative permittivity over the tissue conductivity on the induced field in the
brain. When considering the induced field along head axis, the results show the impact of
parameters’ variability to be distributed unevenly.
Published under Poster Session B: PB-30

FB-3 [13:36]
STUDENT PAPER
Investigation of E-field strength from mobile phone base stations and transmitted power
from 4th generation mobile phones
Shota Kurosaki1, Masaki Hagiwara1, Masao Taki1, Atsuko Aimoto1, Miwa Ikuyo1, Kaoru Esaki1 & Kanako

368
Wake2
1Department of Electrical and Engineering, 1 - 1 Minami-osawa, Hachioji-shi, Japan, 192-0397
2Electromagnetic Compatibility Laboratory, National Institute of Information and Communications Technology,
Tokyo, Japan, Japan, 184-8795
Keywords: Dosimetry (measurements), RF/Microwaves, Review, Commentary, Recommendation,
Evaluation
Presented by: Shota Kurosaki
We measured 4G Tx power and the total downlink E-field strengths during drive around Tokyo and
Chiba area in Japan for about 8 hours. The association between 4G Tx power and the downlink
E-field strengths was investigated. The total downlink E-field strength was high and 4G Tx power is
low in the Center of the Capital while the total downlink E-field strength was low and the 4G Tx
power is high in residential/rural area.
Published under Poster Session B: PB-48

FB-4 [13:39]
STUDENT PAPER
Flexible conductive polymer microelectrode arrays for electropulsation, neurostimulation
and electroporation in vitro and in vivo
Gerwin Dijk1, Hermanus Ruigrok2 & Rodney P. O'Connor2
1Bioelectronics, Ecole des Mines de Saint-Etiennes - Panaxium, Gardanne, France, 13120
2Bioelectronics, Ecole des Mines de Saint-Etiennes, Gardanne, France, 13120
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Gerwin Dijk
Organic conductive polymers are highly versatile materials that have been widely adopted in
commercial electronics as display technology, transistors, LEDs and solar cells. Our research is
exploring their use for interfacing biology with electronics. We have developed microelectrode array
devices using the conductive polymer PEDOT:PSS for electropulsation, neurostimulation and
electroporation investigations in vitro and in vivo. Here we will present our work using plastic
bioelectronics to develop multiwell electrode arrays for in vitro studies using live cell imaging and
flexible, implantable electrodes that are suitable for studying electropulsation effects with in vivo
imaging and electrophysiology.
Published under Poster Session B: PB-68

FB-5 [13:42]
STUDENT PAPER
Cell electroporation enhanced by conductive nanoparticles
Amina Ghorbel1, Lluis M. Mir1 & Tomás García-Sánchez1
1UMR 8203, CNRS, Gustave Roussy, Villejuif, France, 94805
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Amina Ghorbel
In order to optimize the outcome of Electroporation , a novel approach based on the use of
conductive nanoparticles, is proposed to enhance locally the efficacy of the electric field intensity in
the targeted biological sample.
Published under Poster Session B: PB-70

369
FB-6 [13:45]
STUDENT PAPER
Does Extremely Low Frequency Magnetic Fields stimulations of the vestibular system
modulate postural control in humans?
Nicolas Bouisset1, 2, Sebastien Villard1, 3, Daniel Goulet7, Michel Plante7, Martine Souques6, François
Deschamps5, Genevieve Ostiguy7, Jacques Lambrozo6 & Alexandre Legros1, 2, 3, 4, 8
1Human Threshold Research Group, Lawson Health Research Institute, London, ON, Canada
2Department of Kinesiology, Western University, London, ON, Canada
3Department of Medical Biophysics, Western University, London, ON, Canada
4Department of Medical Imaging, Western University, London, ON, Canada
5RTE, Département Concertation et Environnement, Paris-La Défense, France
6Service des Études médicales , EDF, Levallois-Perret Cedex, France
7Hydro-Québec,, Montréal, Québec, Canada
8EuroMov, Université de Montpellier, Montpellier, France
Keywords: Human, ELF/LF, Work in Progress
Presented by: Nicolas Bouisset
Exposure to extremely low frequency magnetic fields (ELF-MF) induces electric fields and currents
within the human body that modulate nervous system functions. Although still a matter of debate,
there are evidences that the vestibular system could respond to ELF-MF. It is well established
however that specific electrical stimulations of the vestibular system trigger specific postural
outcomes. This work investigates the vestibular response to ELF-MF and alternating electrical
currents through their impact on human postural sway.
Published under Poster Session B: PB-84

FB-7 [13:48]
STUDENT PAPER
Alpha band analysis in resting EEG after exposure to radiofrequency signal (900 MHz):
EEG/MEG study combined with an anatomic MRI
Jasmina Wallace1, 2, Lydia Yahia-Cherif3, 4, Laurent Hugueville3, 4, Christophe Gitton3, 4 & Brahim
Selmaoui1, 2
1Department of Experimental Toxicology, Institut National de l'Environnement Industriel et des Risques
(INERIS), Verneuil-en-Halatte, France
2Peritox Laboratoire de Périnatalité & Risques Toxiques, Unité mixte INERIS, Université de Picardie Jules
Verne, CURS, Amiens, France
3Centre De NeuroImagerie De Recherche - CENIR, Institut Du Cerveau Et De La Moelle Épinière - ICM,
Paris, France
4Inserm U1127, CNRS UMR 7225, Institut Du Cerveau Et De La Moelle Épinière - ICM, Paris, France
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Jasmina Wallace
With the extensive use of mobile phones (MP) several studies have been realized to understand the
effects of radiofrequency exposure on brain activity. The results show changes in the alpha band
spectral power. To better understand the cortical structures involved in these changes after MP
exposure (900 MHz), we carried out electroencephalography and magnetoencephalography
recording followed by the anatomical magnetic resonance imaging on healthy volunteers.
Participants were selected according to inclusion criteria. They were healthy males or females, aged
between 18 and 35, non-smokers and with regular sleep habits. Data analyses of 10 subjects are
still under process and ready results will be presented at the conference.

370
Published under Poster Session B: PB-90

FB-8 [13:51]
STUDENT PAPER
The role of NF-kB and p-SAPK/JNK inflammatory responses of RFR- exposed colon cancer
cells
Fatih Senturk1, Elcin Ozgur1, Gorkem Kismali2, Tevhide Sel2 & Goknur Guler1
1Biophysics Department, Gazi University, Ankara, Turkey, 06500
2Biochemistry Department, Ankara University, Ankara, Turkey, 06110
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Fatih Senturk
The aim of this study is to investigate the effects of 900, 1800, 2100 MHz mobile phone radiation
exposure during 1 hour and 4 hours on NF-kB and p-SAPK/JNK activation levels on HCT116 colon
cancer cells. Results showed that 1h exposure to 900, 1800 MHz caused to statistically significant
increase, whereas 2100 MHz decreased NF-kB levels, but 4 hours of exposure lead to increase the
level of NF-kB for all exposure groups.On the other hand, 1-hour exposure of 2100 MHz leads to
increase in SAPK/JNK levels. However, no difference was found in the other groups with respect to
sham groups.To conclude that 2100 MHz RFR exposure may modulate NF-kB and SAPK/JNK
expression in HCT116 cells.
Published under Poster Session B: PB-102

FB-9 [13:54]
STUDENT PAPER
Physical chemical analysis of tomato seeds (Solanum lycopersicum L.) under magnetic
treatment
Javier Torres Osorio2, Laura Maria Echeverry-Cardona1, Elisabeth Restrepo-Parra1, Nelson Ceballos
Aguirre3 & Diana Carolina Galvez-Coy4
1Departamento de Física y Química, Universidad Nacional de Colombia, Manizales, Colombia, 170002
2Departamento de Física, Universidad de Caldas, Manizales, Colombia, 170002
3Department of Agricultural Production, University of Caldas, Manizales, Colombia, 170002
4Tecnoacademia, Servicio Nacional de Aprendizaje SENA, Manizales, Colombia, 170002
Keywords: In vitro, Static, Completed (unpublished)
Presented by: Laura Maria Echeverry-Cardona
Physical chemical analysis of magnetic treatment tomato seeds that exhibited reduction in the mean
germination time (MGT) was carried out for knowing the mechanisms that generate biophysical and
biochemical modifications of the seeds because the magnetic field. The experiment was performed
using a set of dipolar toroidal magnets, 8 doses of stimulation and a control for germination tests.
The seeds with lower MGT were chosen. The chemical and the surface morphology characterization
was performed. Seeds with lower MGT showed an increase of functional groups of lipid molecules,
increase in the free water for reactions and expressions of potassium ions of the seed coat, that can
be related to the increment in the membrane transport.
Published under Poster Session B: PB-110

FB-10 [13:57]
STUDENT PAPER
Effects of RF-EMF on APP-processing and cell death in mouse hippocampus cell line

371
Kyeonghee Yoon1, Sojeong Choi1, Nam Kim2, Hyung Do Choi3, Hae June Lee4 & Yun-Sil Lee1
1College of Pharmacy & Graduate School of Pharmaceutical Sciences , Ewha Womans University , Seoul,
Korea
2School of Electrical and Computer Engineering, Chungbuk National University, Chungju, Korea
3Departmet of EMF Research Team, Radio and Broadcasting Technology Laboratory, ETRI, Seoul, Korea,
139-706
4Division of Basic Radiation Bioscience, Korea Institute of Radiological and Medical Sciences, Seoul, Korea,
01693
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Kyeonghee Yoon
In vitro experiments were performed after exposure of LTE RF-EMF (1745 GHz) because currently,
the spread of LTE mobile communication network has increased. The LTE type RF-EMF was
exposed to the cells for 24 hours at an intensity of 8 W/kg, which is the maximum intensity of our
exposure system and Amyloid precursor protein (APP) processing and cell death effects were
examined. RF-EMF exposure showed no significant changes of ADAM10 and BACE1 as well as
expression of CTFβ and CTFα. RF-EMF, no significant cell death was induced. Even though more
detailed experiments is needed, unlike to IR, exposure of 8 W/kg LTE type RF-EMF for 24 hours did
not affect APP processing and cell death in HT22 hippocampus cells.
Published under Poster Session B: PB-114

FB-11 [14:00]
STUDENT PAPER
Magnetic field effects on H2O2 and possible biological implications including cancer
Sahithi Kandala1, Rodolfo Bruzon1, Hakki Gurhan1 & Frank Barnes1
1Electrical, Computer and Energy Engineering, University of Colorado, Boulder, Colorado, USA, 80309
Keywords: In vivo, Static, Work in Progress
Presented by: Sahithi Kandala
The links between Reactive oxygen species, Electromagnetic Fields and Biological Implications are
explored. Experimental data is used as background to show the changes that are observed due to
static magnetic fields on fibrosarcoma cells.
Published under Poster Session B: PB-118

FB-12 [14:03]
STUDENT PAPER
Cellphone radiofrequency signal induced inflammatory response oxidative stress and effect
on memory in rat brain
Kumari Vandana1, 2, Jayprakash Nirala1, 2, Ranjan Nanda2 & Paulraj Rajamani1, 2
1School of Environmental Sciences, Jawaharlal Nehru University, New Delhi, India, 110067
2Translational health group, International Center for Genetic Engineering and Biotechnology, New Delhi,
India, 110067
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Kumari Vandana
The exponential worldwide growth in the use of mobile phones has raised concerns regarding
potential harmful effect of exposure to radiofrequency modulated electromagnetic field radiations. Of
particular concern has been the effect on brain and hippocampus in specific. Present study was
undertaken to determine if long term exposure to RF cause inflammatory cytokines mediated

372
oxidative stress in brain and if there is any functional consequence of it in terms of memory
impairment.
Published under Poster Session B: PB-122

FB-13 [14:06]
STUDENT PAPER
Design of a triple band monopole antenna using parasitic elements with SAR reduction for
WLAN and WiMAX Applications
Hanul Bong1, Niamat Hussain1, Min-Joo Jeong1, Ji Woong Park1, Seungwoo Lee2 & Nam Kim1
1Chungbuk National University, Cheongju, Korea, 28644
2Korea Electric Power Corporation, Naju, Korea, 58217
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Hanul Bong
A triple band antenna consists of a dipole, a monopole and parasitic elements is presented. The
aimed frequency band is WLAN (2.4/5.8 GHz) and WiMAX (3.5 GHz) applications. The parasitic
elements around feed line is introduced for improving the performance of the antenna. The antenna
shows acceptable radiation pattern along with - 10 dB return loss at all three bands. Moreover, the
proposed antenna with reflector got reduction of 85.12 %, 50.06 % and 36.87 % in 1g SAR value at
2.4 GHz, 3.5 Ghz and 5.8GHz, respectively when compared with international standard value.
Published under Poster Session B: PB-136

FB-14 [14:09]
STUDENT PAPER
Design of a triple band folded dipole antenna with low Specific Absorption Rate (SAR) for
LTE/WLAN applications
Ji Woong Park1, Min-Joo Jeong1, Niamat Hussain1, Hanul Bong1, Seungwoo Lee2 & Nam Kim1
1Chungbuk National University, Cheongju-si, Korea, 28644
2Korea Electric Power Research Institute, Cheongju-si, Korea, 58217
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Ji Woong Park
This paper presents a triple band folded dipole antenna with a reflector for 1.8 GHz LTE and 2.4/5.8
GHz WLAN applications. The antenna is composed of a monopole, via, two folded dipoles, and a slit
etched on the folded dipole. To obtain the optimized parameters, we used the simulator, Sim4Life
and the antenna shows an acceptable radiation pattern along with -10 dB return loss at all three
bands. The maximum value of the SAR1g (1 g-averaged SAR) for 1.8 GHz, 2.4GHz, and 5.8 GHz
band was 0.076 W/kg, 0.071 W/kg, and 0.0381 W/kg, respectively. These values are much lower
than the international standard values of SAR1g (1.6 W/kg), even though the distance between
antenna and head is zero.
Published under Poster Session B: PB-144

373
 Session: PA
Poster Session A
Monday June 25, 2018 • 14:30 - 16:00
Mediteranea

PA-1 [14:30]
The influence of electromagnetic field from microwave range on honeybees – acoustic and
thermal analysis
Pawel Bienkowski1, Pawel Cala1, Urszula Libal2, Piotr Nowakowski3, Przemyslaw Plaskota4, Krzysztof
Rudno-Rudzinski4 & Bartlomiej Zubrzak1
1Department of Telecommunications and Teleinformatics, Wroclaw University of Science and Technology,
Wroclaw, Poland, 50-370
2Department of Signal Processing Systems , Wroclaw University of Science and Technology, Wroclaw,
Poland, 50-370
3Institute of Animal Breeding, Wroclaw University of Environmental and Life Sciences, Wroclaw, Poland,
50-375
4Department of Acoustics and Multimedia, Wroclaw University of Science and Technology, Wroclaw, Poland,
50-370
Keywords: Behavioural, RF/Microwaves, Work in Progress
Presented by: Pawel Bienkowski
The paper presents the results of research on the influence of the electromagnetic field from
microwave range on the population of honeybees, analyzing changes in the sounds from hives and
changes in temperatures in the hive.
Introduction
At the turn of the last decades, we experience a significant reduction in the honeybee population in
the world. The number of bee colonies has decreased in recent years in Europe by 10-30 percent,
by 30 percent in the United States and up to 85 percent. in the Middle East. Specialists cannot
unambiguously answer what is causing this state of affairs. It is assumed that responsible for this
are factors such as pesticides, viruses, climate change, GMOs or parasites. The convergence of
bee colony extinctions with the development of telecommunications and the emergence of base
stations of cellular telephony has led to the thesis that such a factor may also be the
electromagnetic field (EMF) originating from mobile telephony base stations.
Electromagnetic signals are both positive and negative for insects life. There is species identity
through analysis of flight buzzing of social insects [1]. Bees cannot hear the pressure fluctuations of
sounds but they are sensitive to airflows associated with sounds. Most of these airflows are caused
by vibrating wings and bodies[2]. Positive communication is the noise generated by honeybee
colonies in the course of swarming [3] while mobile phones affect negatively the honeybees` life
system [4].
Since honeybees are shown to be sensitive to the earth's magnetic field and are likely to use it for
navigation, many studies on bees have been carried out [5] in which various reactions were tested
exposing them to EMF from different frequency bands, starting from a constant magnetic field
through the field from power transmission lines (50 / 60 Hz)[6], from GSM base stations (typically
900 and 1800 MHz) [7,8] and a higher frequency (eg 2.4 GHz or 10 GHz)[9]. Reviewing the results
of the conducted research it can be concluded that there are no unambiguous results. The authors
of the studies indicate links between honeybee behavior and EMF presence, but at the same time
similar experiments show their lack. In many cases, the research methodology adopted is
debatable, and the most controversial is the way in which bees are exposed to electromagnetic
fields, where the exposure system are close surroundings of mobile phone base stations, cell
phones, DECT telephone bases or field generators verified by non-adapted devices are used.

374
Methods
The experiment was conducted on popular honeybee species (Apis mellifera carnica), in one of the
Wroclaw University of Science and Technology laboratory during 6 days period. Because the
laboratory conditions slightly differed from the natural environment of bees, first three days were the
adaptation period. The temperature in the room was about 24°C and humidity was in the range from
30 to 40%. There was no artificial illumination - light conditions in the laboratory depended on the
light outside. The research was performed at the turn of spring and summer, so the day was
relatively long. It allowed to provide environmental lighting conditions close to natural. Honeybees
were placed in three wooden micro-hives with two glass walls. Insects were obtained from same
family and same hive frame in number of approx. 200 per micro-hive. Each micro-hive was
equipped in two scaled feeders, contained water and the food (aqueous sugar solution in 1:1
proportion). Food and water consumption was monitored at specific intervals, as well as insect
mortality. Micro-hives were placed at such a distance from each other to minimize mutual influence
and also to provide similar environmental conditions. First micro-hive (exposed) has been subjected
to an electromagnetic field from the 900 MHz frequency, pulse-modulated with parameters similar to
the electromagnetic field parameters from the mobile phone base station (GSM900 system). Second
hive contained the honeybee control group (not exposed to EMF). The antenna used in the
experiment have been designed to ensure uniform exposure of the micro-hive simultaneously
reducing the unwanted emission towards other directions and was presented in [10]. Applied field
strength values reflected the levels that could be encountered in the surroundings of real mobile
phone BS, which was the field strenght of 7 V/m. Exposure time was controlled by timer system with
settings: 10 minutes with exposure, 50 minutes without. During the experiment hive temperatures
were monitored and the acoustic signals from the hives were recorded. The temperature in the room
was measured, and also the acoustic signals was recorded to eliminate from further analysis the
informations, which could be caused by sudden sounds in the experiment environment. The
schematics of the research system setup was presented in figure 1.
Results
During the experiment a large amount of data was obtained. Some of the data was excluded from
the results, as it was corrupted by unwanted and accidental sounds from the experiment
environment, as it was not perfectly silenced room. Exemplary spectrum of audio recording (in this
case - escited bees) is presented in figure 2 and figure 3 present the temperature graph during 48
hours of experiment. Both temperature graph and audio recordings showed a repetitive life cycle of
micro-hive. Audio recording analysis methodology was aimed to estimate the honeybee acoustic
activity (for example: worker piping or excitation ofthe bees) in relation to exposure to
electromagnetic field. In the time domain, acoustic pressure graphs were analysed (10 minutes
before exposure and 10 minutes during exposure in different time of the day and night) and
compared with the graphs recorded in the unexposed micro-hive. In the frequency domain, spectral
analysis was performed for the same audio samples.
Thermal analysis was done using statistical tests of significance of differences between the
temperature samples measured before turning the EM field and during the EM field emissions. The
tests were performed for pairs of samples from the 10-minute periods preceding the field inclusion
and from the immediately following 10-minute periods in which the field was included. The paired-
sample t-test [11] on significance level 5% did not reject null hypothesis that difference of the two
samples comes from a normal distribution with mean equal to zero. The obtained p-values are as
follows: 0.1390 for null hypothesis that there is no difference between temperature in exposed hive
with applied EMF and the temperature withoutapplied EMF, 0.0797 - in control hive with EMF and
without EMF.
Discussion
Preliminary research presented in this work showed, that there is no clear evidence of the influence
of electromagnetic field from microwave range with parameters similar to the GSM900 mobile
telephony base station on honeybees that can be noticed based on a basic audio analysis of sounds

375
recorded in micro-hives and analysis of temperature changes in the micro-hives.
As the recorded sound presents source of data with greater potential, more sophisticated methods
need to be applied in further research. Taking into account the modern possibilities in the field of
acoustic signal analysis, where there are really many methods, it would be necessary to plan
research based on possibly precise working hypotheses. In this case it would be possible to
optimize the selection of research methods.
Conclusion
As part of the experiment organization and the EMF parameters used, no change in acoustic or
thermal activity related to EMF exposure was found. The result cannot be generalized in any way, it
may be caused by too low sensitivity and resolution of the recording method and should be a
contribution to further research and analysis.
References
1. Gradišek A., Gašper Slapničar G., Sorn J., Luštrek M., Gams M., Grad J. (2017) "Predicting
species identity of bumblebees through analysis of flight buzzing sounds" Bioacoustics, 26 (1):
63–76.
2. Michelsen A. "Mechanical Signals in Honeybee Communication" Animal Signals and
Communication 3; Studying Vibrational Communication Cocroft R.B., Gogala M., Hill P. S. M.,
Wessel A. (Editors), Elseviere: 333-348.
3. Eskov E.K., Toboev V.A. "Changes in the Structure of Sounds Generated by Bee Colonies
During Sociotomy". Entomological Review, 91 ( 3): 347–353.
4. Favre D. "Disturbing Honeybees’ Behavior with Electromagnetic Waves" Methodology Journal of
Behavior 2 (2): 1010.
5. Friesen M "Review of the literature on the biological effects of wireless radiaton on insects: a call
for more studies on honey bees" October 31st, 2014 Entomological Society of Manitoba, 70th
Annual Meeting
6. Bindocas, V P; Gauger, J R; Greenber, G B "Mechanism of biological effects observed in honey
bees (Apis mellifera, L.) hived under extra-high-voltage transmission lines: implications derived
from bee exposure to stimulated intense electric fields and shocks". Bioelectromagnetics. 9(3):
285-301.
7. Sainudeen Sahib S. "Impact of mobile phones on the density of honeybees" Journal of public
administration and policy research Vol. 3(4) pp. 131-117 April 2011
8. Ritu Ranjan Taye, Mukul Kumar Deka, Ataur Rahman and Manha Bathari "Effect of
electromagnetic radiation of cell phone tower on foraging behaviour of Asiatic honey bee, Apis
cerana F. (Hymenoptera: Apidae)" Journal of Entomology and Zoology Studies 2017; 5(3):
1527-1529
9. Atli, E., & Unlü, H. "The effects of microwave frequency electromagnetic fields on the
development of Drosophila melanogaster" International Journal of Radiation Biology, 82(6),
435–441.
10. Bieńkowski P., Cała P., Wyszkowska J., Zubrzak B.: "Układy ekspozycyjne PEM w badaniach
biomedycznych / Exposure systems in biomedical research", Przegląd Telekomunikacyjny,
Wiadomości Telekomunikacyjne. 2015, R. 88, nr 4, s. 510-514,
11. Welch B.L., "The generalisation of student's problems when several different population
variances are involved". Biometrika, Vol. 34, No.1-2, pp.28–35, 1947.
Figures

376
Figure 1. Schematics of the research setup.

377
Figure 2. Exemplary spectrum of recording from the hive - bees excitation.

Figure 3. Temperature graph form 48 hours of experiment with indication when the EMF was
applied.

378
PA-3 [14:30]
STUDENT PAPER
Extremely low frequency-electromagnetic radiation deteriorated the sleep reduction in
Drosophila melanogaster under heat stress
Xiaomei Huang1, 2, Ziyan Zhang1, Hongying Zhang1, 2, Chuanjun Yang1, Yongyan Sun1, 2, Chao Tang1 & Hui
Yu1
1Key Laboratory of Urban Environment and Health, Institute of Urban Environment (IUE), Xiamen, China,
361021
2Chinese Academy of Sciences, Beijing, China, 100049
Keywords: Behavioural, ELF/LF, Completed (unpublished)
Presented by: Xiaomei Huang
The coupling effects of heat stress and extremely low frequency-electromagnetic field (ELF-EMF, 50
Hz, 3 mT) on sleep and activity were studied using Drosophila (wild-type Canton-Special (CS) and
mutant W1118 flies) as a model. The mRNA expressions of circadian clock genes, neurotransmitters
related genes, as well as important neurotransmitters were analyzed. Result showed that ELF-EMF
deteriorated the sleep reduction in Drosophila under heat stress.
Both green-house effect and electromagnetic radiation received increasing public health concern.
However, their coupling effects and the mechanism are still unclear. In our study, Drosophila was
applied as a model to assess the coupling effects on the locomotion and the sleep of flies by heat
stress and extremely low frequency-electromagnetic field exposure (ELF-EMF, 50 Hz, 3 mT). The
mRNA expressions of circadian clock genes (cry, per, tim, and clk), neurotransmitters related genes
(ddc, pale, trh, tph and gad1), as well as important neurotransmitters (serotonin, dopamine, and
glutamic acid) were analyzed. Results showed that single exposure to ELF-EMF did not affect the
locomotion and the sleep rhythms of flies while heat stress significantly disturbed the rhythms. The
activity was decreased by heat stress and ELF-EMF can alleviate the decrease. The quality of sleep
was notably decreased by heat stress and it can be deteriorated by ELF-EMF. Second, heat stress
and ELF-EMF disturbed the circadian rhythm of flies via affecting the expression of tim, cyc and clk.
Third, the expression of pale was noticeably decreased by heat stress, but not affect by ELF-EMF.
The expression of ddc was considerably increased by ELF-EMF, but not affected by heat stress.
The expression of gad1 and the content of Glu were significantly decreased by heat stress but not
ELF-EMF. Overall, combined exposure to heat stress and ELF-EMF could affect the expression of
the circadian clock and neurotransmitter related genes, as well as the synthesis of
neurotransmitters, resulting in circadian rhythm disorders and sleep quality decline.
Figures

379
Figure 1. Effects of heat stress and ELF-EMF exposure on drosophila activity and sleep.
(A) Locomotion rhythm of CS; (B) Locomotion rhythm of w1118; (C) Sleep rhythm of CS; (D)
Sleep rhythm of w1118; (E) Activity percent in the first hour; (F)Activity counts; (G)The total
sleep time of CS and w1118; (H) The number of sleep episodes of CS and w1118; (I) Mean
sleep episode duration of CS and w1118.

380
Figure 2. Heat stress and ELF-EMF exposure changed the expressions of clock genes
and the neurotransmitter related genes in drosophila. (A) Expression of tim; (B)
Expression of cyc; (C) Expression of clk; (D) Expression of magr; (E) Expression of trh; (F)
Expression of tph; (G) Expression of pale; (H) Expression of ddc; (I) Expression of gad1.

381
 Figure 3. Overview of the results. Red arrows and red short lines correspond to the results of
the groups exposed to 35°C compared with those at 25°C. Black arrows and black short lines
indicate the results in the groups exposed to ELF-EMF compared with those without ELF-EMF.
Upward arrows represent increasing trend, downward arrows show decreasing trend and
horizontal short lines indicate no significant difference. (TST: total sleep time; NSE: the
number of sleep episodes; MSED: mean sleep episode duration.)

PA-5 [14:30]
Biological tissue characterizations by HF band electromagnetic field
Kazuyuki Saito1 & Takashi Yakubo1
1Chiba University, Chiba, Japan, 2638522
Keywords: Clinical (diagnostics), IF, Work in Progress
Presented by: Kazuyuki Saito
Characteristics of biological tissues were considered by measurement of their electrical properties in
the high frequency (HF) band. “Detection of tissue coagulation” and “inspection of tissue freshness”
were introduced as examples.
Introduction
We have been studying biological tissue characterization by measurement of its electrical
properties. In this investigation, selection of measurement frequency is important. Recently, along
with the expansion of wireless communication by the radio wave, studies on the electrical properties
of the biological tissue more than the microwave band have been processed. However, according to
our previous studies, measured electrical constants of the microwave band had not reflected the
tissue conditions. On the other hand, around the high frequency (HF) band, the electromagnetic
energy transmits in the biological tissue as an electrical current. Therefore, the electrical current
passes are different by the cellular condition. As a result, the cellular condition reflects to impedance

382
of the biological tissue. In addition, it means variation of the electrical properties of the tissue. In this
time, two examples such as “detection of tissue coagulation” and “inspection of tissue freshness”
are shown as characterization by the electrical properties measurements.
Methods
In order to measure the electrical properties of the biological tissue, an LCR meter (Waynee Kerr
Electronics 6530B) was employed. Samples for the measurements are filled into the cylindrical
sample folder which has two electrodes at both ends. Figure 1 shows photograph during the
measurement.
In the investigation of “detection of tissue coagulation”, temperatures of the sample folder are
changed, and the electrical properties in the folder are measured at each temperature. The sheet-
shaped heater which is wrapped up the sample folder was employed for temperature control. In
addition, in order to investigate the “tissue freshness”, variation of the tissue electrical properties
with time transition is recorded.
Results and discussion
Figure 2 shows temperature dependent of relative permittivities and conductivities of swine liver
tissue. Around 70 ○C or more, the tissue is coagulated. Consequently, the electrical properties of the
tissue have changed drastically at the temperature. This phenomenon may be used for detection of
tissue coagulation in surgical operation.
Figure 3 is variation of the electrical properties of swine liver tissue with time. The same point color
indicates the same sample. From the results, it can be observed that as time proceeds, the relative
permittivity decreases and the conductivity increases. Moreover, the values of differences are
relatively large. So, it may be applied to any physics and chemistry researches.
Conclusion
In this study, biological tissue characterizations by the HF band electromagnetic field have been
considered. From some results, it can be said that the electrical properties of the biological tissue
reflect their situation. As a further study, simplification of the measurement will be considered.
Figures

Figure 1. Measurement system (LCR meter with sample folder).

383
 Figure 2. Temperature dependent electrical properties of swine liver tissue.

Figure 3. Variation of electrical properties of swine liver tissue with time.

PA-7 [14:30]
Is it biologically plausible to use millimetre and terahertz waves for therapeutic applications?
Mats-Olof Mattsson1, Olga Zeni2 & Myrtill Simko1
1SciProof International AB, Östersund, Sweden, 83132
2CNR-IREA, Naples, Italy, 80124
Keywords: Clinical (therapy), THz, Completed (unpublished)
Presented by: Mats-Olof Mattsson

384
There are many diagnostic and therapeutic situations where presently available tools (physical,
chemical, biological) are insufficient. The search for improved approaches includes investigations of
the usefulness of millimetre (MMW) and terahertz waves (THz). Here we investigated if published in
vivo and in vitro studies provide support for the use of these waveforms in therapeutic applications,
during conditions that do not cause tissue heating. Our conclusion is that there is some evidence for
that both MMW and THz can influence biological systems in a manner that is not obviously driven by
tissue heating. However, the number of relevant studies is very low which limits the drawing of any
far-reaching conclusions.
Introduction
Health care institutions are challenged to develop additional instruments that can improve on
diagnosis and/or therapy, and at the same time being affordable to purchase, easy to use, and have
reasonable running costs. From that perspective, significant interest has appeared for developing
approaches that are based on the millimetre wave (MMW) and terahertz wave (THz) parts of the
electromagnetic spectrum. Before such approaches will have any meaningful clinical impact, a
number of requisites have to be fulfilled. These include to understand the interactions between
MMW or THz on one hand and biological structures on the other; to understand how biological
effects occur and which they are if a therapeutic use is intended; to have a relevant data base of
experimental studies for design of further studies; and finally to perform controlled clinical studies of
sufficient size.
Until presently, MMW, that lie in the frequency region from 30–300 GHz (wavelength: 1–10 mm),
have been reported to produce therapeutic effects in certain conditions, whereas diagnostic use
seems to be less well investigated. THz, lying in the region from 300 GHz to 10 or even 20 THz
(wavelength: 100 μm to 1 mm), on the other hand has received interest since it can be used for
making imaging possible in locations not accessible to relevant investigations with conventional
techniques, and with reasonable resolution. Only few studies have addressed the therapeutic
potential of THz waves.
The primary requisite for any therapeutic action of MMW or THz waves is that exposure to these
electromagnetic fields induces biological effects at exposure levels that are not damaging to tissues.
Such detrimental effects will occur at high power density levels due to tissue heating, which may be
desirable in certain clinical settings (e.g. killing of tumour cells). However, for the majority of clinical
situations such drastic outcomes are not warranted. Thus, the question, which we address in this
paper, is if experimental studies (in vivo and in vitro) provide evidence for “non-thermal” biological
effects. Such effects would not cause tissue heating and associated damage and are the ones that
can be exploited for therapeutic medical use.
Methods
In order to answer the question if MMW and THz exposures can provide “non-thermal” biological
effects, which would be a prerequisite for many of the suggested therapeutic applications based on
these parts of the EM spectrum, we have scrutinized published in vivo and in vitro studies for
evidence of effects, taking also study quality criteria into consideration. Epidemiological studies were
not considered since they are presently not available.
Results
In total, around 120 relevant papers were identified and analysed regarding exposure conditions,
biological endpoints and outcomes, and study quality. Importantly, analysed papers were considered
to be performed under conditions that would allow “non-thermal” effects to occur, either according to
the data provided in the description of the study or according to a statement to that effect by the
study authors.
When analysing the details in the studies, it is apparent that the relatively modest number of
identified papers cover many different exposure conditions and endpoints. Thus, any conclusions
regarding possible health relevant effects are not possible to draw based on independent
replications or strong similarity between studies.

385
Conclusions
The investigated studies provide some evidence for that both MMW and THz can influence
biological systems in a manner that is not obviously driven by tissue heating. However, the number
of relevant studies is very low which severely limits the drawing of any far-reaching conclusions.
Furthermore, the studies have not addressed specific interaction mechanisms and do not provide
hints for future mechanistic studies. Also, the studies do not indicate any specific importance
regarding power density. levels, frequencies, or exposure duration. It is also unclear if any specific
biological end-points are especially sensitive.
Any therapeutic potential of MMW or THz has to be evaluated based on future high quality studies
dealing with physical, bio-physical, and biological aspects that have specific health-related
perspectives in mind.

PA-9 [14:30]
WITHDRAWN

PA-11 [14:30]
QGIS calculation method for evaluation of ELF electromagnetic field exposure of general
public due to overhead power lines
Wout Joseph1, Kris Vanhecke1, Christophe Geuzaine2, Leen Verloock1, Matthias Van Den Bossche1, Mart
Verlaek3, Michel Goethals3 & Luc Martens1
1INTEC, Ghent University\IMEC, Ghent, Belgium, 9052
2Electrical Engineering and Computer Science, University of Liege, Liege, Belgium, 4000
3Bureau for Environment and Spatial Development – Flanders, Departement Omgeving, Brussels, Belgium,
1000
Keywords: Dosimetry (computational), ELF/LF, Completed (published)
Presented by: Wout Joseph
A calculation model (applying Gmsh and GetDP) and QGIS plug-in for simulating magnetic field
contours of overhead power lines is presented. This model enables the estimation of the ELF
exposure of the general population, relevant to environmental and urban planning, and can take into
account future changes in the power grid. Simulations are validated with measurements and
demonstrated for the region of Flanders, Belgium.
Introduction
Infrastructure for the transmission of electricity is a source of non-ionizing radiation. In recent years,
the power grid has undergone changes to accommodate renewable and decentralized energy
sources (wind, solar installations, biomass, etc.). Therefore, investments in the transmission
infrastructure are required.
Epidemiological studies showed that time-weighted average exposure to power-frequency magnetic
fields above 0.4microT is associated with a small increase in the absolute risk of leukaemia in
children (Ahlbom et al. 2001). The International Agency for Research on Cancer (IARC) evaluated
extremely low frequency (ELF) magnetic fields as being “possibly carcinogenic to humans”
(classification 2B) (IARC, 2002). Currently, no mechanism has been found that explains possible
health risks of extreme low-frequency (ELF) radiation. When installing new infrastructure for the
transmission of electricity, governments want to be able to determine geographical areas (i.e.
contours surrounding the overhead lines) where the field strength exceeds a given magnitude and
account for uncertainties mentioned above by applying precautionary measures. A calculation
model enables governments and researchers to make field estimations for environmental and urban
planning. A GIS (geographical information system) calculation model is also valuable for
communication to the general public.

386
The goal of this paper is to develop a GIS calculation model for magnetic field contours of
transmission lines and enabling an estimation of the exposure of the population. The GIS model is
validated with on-site measurements in proximity to overhead power lines and with data from the
transmission system operator. Magnetic field exposure levels for the Flemish region will be
presented.
Materials and Methods
Tower geometry and conductor positions for single or double three-phase circuits are modelled
using the open source mesh generator Gmsh (Dular et al. 1998). This is a 3D finite elements grid
generator with build-in CAD system. Field calculations are performed using the multi-physics finite
element solver GetDP (Geuzaine & Remacle 2009). Gmsh and GetDP are integrated in the
graphical environment ONELAB (Geuzaine et al. 2012) (http://onelab.info). Parameters can be set
interactively (on-the-fly) and automatic calculation of 3D conductor profile is performed. Fig.1 shows
an example of the calculation of the magnetic field of overhead power lines using Gmsh and GetDP.
The arrangement of the electrical circuit (i.e. the arrangement of the phases, and in particular
whether there is transposition when there are two circuits) and the maximum sag are accounted for
in the calculation.
A plugin was created for QGIS 2.x to calculate the field contours and to report geographical areas
where the population is exposed to a magnetic field in excess of a given magnitude. The inputs of
the plugin are (i) geospatial vector data files with the location of towers and overhead lines, (ii) tower
geometry information, and (iii) user input such as the maximum allowed field, custom scenarios with
regard to mean load, transposition and maximum sag of the conductors. The processing part of the
plug-in converts these inputs into parameters for Gmsh and GetDP, and imports the output of GetDP
(coordinates and field strength) into QGIS. The plugin generates geographical areas, visualized by
polygons, where the field strength exceeds the maximum field level under consideration. A reporting
module then determines the number of people that reside in areas that overlap with these polygons.
Fig.2 shows the flow of the methodology of the QGIS plugin’s calculation model. Grid points where
the magnetic field exceeds 0.4 microT are indicated by red dots.
For the validation of the plugin, measurements were performed in proximity to 150-kV and 380-kV
overhead power lines in the city of Lokeren, Belgium, in December 2017. The measurement
procedure of Joseph et al., (2009) was used. The fields were measured by an electric- and
magnetic-field analyzer of type PMM EHP-50C (accuracy magnetic field 6%). The current load was
monitored during 24 hours by the operator; e.g., a current of 467A was recorded for the 150-kV line
at 4:02pm. The magnetic field variation is proportional to the relative current variation. Average and
maximum field values (over 24 hours, a year, etc.) can then be calculated by averaging the time-
varying current in the conductors.
Results and Discussion
No phase transposition at the time of measurement was applied. The area where the measured
value exceeds 0.4 microT, indicated in orange in Fig. 3(b), extends outward for approximately 55
meters and 25 meters from the 150-kV and 380-kV lines respectively. This agrees reasonably well
with the simulation.
Fig. 3(a) shows the simulation of the magnetic fields in and around the city of Lokeren, Belgium, and
Fig. 3(b) shows a more detailed view of the area where the measurements were performed. The 0.4
microT contours are indicated in orange. By default, the plugin assumes that each transmission line
carries a mean load of 25% of its nominal current. The geospatial vector data does not include
transposition information. This was disabled to account for a worst-case scenario. A grid size of 10
meters was considered to be sufficiently accurate for the large area being calculated.
The QGIS plugin can perform simulations for both large regions (with a strong tradeoff between grid
resolution and calculation time) and smaller, specific areas with increased detail. A simulation of the
entire region of Flanders on a 10-meter grid took approximately 14 hours on an Intel i7-4770
desktop processor. The grid size is the determining factor of the simulation time: selecting a 5-meter

387
grid quadruples the processing time. The simulation runs in a single-threaded process, so on a
powerful computer a large area can be broken up into multiple smaller sections and simulated in
parallel. On the i7-4770, a processor with eight virtual processing cores, the simulation time can
thus be reduced by a factor of eight, if care is taken to divide the processing load evenly across all
cores.
The plugin includes a scenario feature that allows government agencies and researchers to
examine the potential impact of (proposed) changes to the electric power grid. For example,
upgrading a transmission line to allow a higher current could result in additional urban areas where
the magnetic field exceeds a given allowed magnitude.
Conclusions
In this paper, an easy to use calculation model (applying Gmsh and GetDP) and QGIS plugin for
simulations of magnetic field contours of overhead power lines is presented. The tool enables us to
make an estimation of the ELF exposure of the population and make field estimations for
environmental and urban planning. Calculations are validated with on-site measurements of
overhead lines and good agreement is obtained. Magnetic field simulations are performed for the
region of Flanders, Belgium and are demonstrated through the plugin.
Acknowledgement
This work was financially supported by the department Environment, Flemish Government. We
thank ELIA for the data of the power lines.
References
Ahlbom A., Cardis E., Green A., Linet M., Savitz D., and Swerdlow A., “Review of the epidemiologic
literature on EMF and health,” Environ. Health Persp., vol. 109, pp. 911–933, 2001, (Suppl 6).
Dular P., Geuzaine C., Henrotte F. and Legros W., “A General Environment for the Treatment of
Discrete Problems and its Application to the Finite Element Method”, IEEE Transactions on
Magnetics, 34(5), pp. 3395-3398, 1998.
Geuzaine C. and Remacle J.-F., “Gmsh, a three-dimensional finite element mesh generator with
built-in pre- and post-processing facilities”, International Journal for Numerical Methods in
Engineering 79(11), 1309-1331, 2009.
Geuzaine C., Henrotte F., Remacle J.-F., Dular P. and Sabariego R., ONELAB: Open Numerical
Engineering LABoratory , Proceedings of the 7th European Conference on Numerical Methods in
Electromagnetism (NUMELEC2012), July 3-5, 2012
IARC, “Static and extremely low frequency electric and magnetic fields,” IARC Monographs Eval.
Carcinogenic Risks Humans, vol. 80, pp. 429–430, 2002, Int. Agency Res. Cancer
Joseph W., Verloock L., and Martens L., “General Public Exposure by ELF Fields of 150-36/11kV
Substations in Urban Environment”, IEEE Transactions on Power Delivery, pp. 642-649, April 2009.
Figures

388
Figure 1. Calculation of B-field in microT of 400 kV overhead lines using GetDP and Gmsh.

Figure 2. Methodology of QGIS plugin (from left to right): overhead line shape file, narrowing
down the considered area, dividing into grid points, calculation of 0.4 microT contour, and
visualization of 0.4microT contour and grid points.

389
 Figure 3. Calculation of 0.4 microT for (a) Lokeren, Belgium and (b) zoomed in on the area
where the measurements were performed. Purple dots indicate the measurement locations.

PA-13 [14:30]
Dielectric parameters for THz study of biological entities
Alireza Lajevardipour1, 2, Zoltan Vilagosh1, 2 & Andrew Wood1, 2
1Swinburne University of Technology, Melbourne, Australia, 3122
2Australian Centre for Electromagnetic Bioeffects Research, Melbourne, Australia
Keywords: Dosimetry (computational), THz, Completed (published)
Presented by: Alireza Lajevardipour
In order to accurately predict THz energy absorption in skin or microbial entities, well-known
simulation method FDTD can be employed. The accuracy of results is depend on precision of
dielectric parameters in the range 0.1 – 100 THz.The Gabriel database uses best-fit parameters in
up to four Debye terms, to predict dielectric parameters up to 0.1 THz. Above taht, the Debye
approach to predicting dielectric values is inadequate.In the case of lipid, there are no specific data

390
sets available, however, analogous data for fat and oil is available. The lack of experimental data for
lipids, is exacerbated by some of absorption data being reported in arbitrary units.
The present work deals with difficulties to select proper dielectric parameters of biomaterials in THz
frequency range. Those parameters are essential to predict energy absorption of biological entities
via FDTD simulations. Although experimental data is freely available for water in this range, there
are lack of experimental reports about dielectric parameters of biomaterials like protein and lipids.
Along with this deficiency of experimental data, some reported absorption data are in arbitrary unit
that need manipulations to scale them for an indirect usage.
Figures

Figure 1. Available experimental data for lipid resembled materials.

391
 Figure 2. The graph of unified experimental data for lipid resembled materials to avoid
arbitrary unit.

PA-15 [14:30]
An investigation of key factors for the simplified compliance assessment of a millimeter-
wave base station
Young Seung Lee1, Haeng-Seon Lee2, Sangbong Jeon1 & Hyung Do Choi1
1Radio & Satellite Research Division, Electronics and Telecommunications Research Institute, Daejeon,
Korea, 34129
2Department of Electronic Engineering, Sogang University, Seoul, Korea, 04107
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Sangbong Jeon
This study presents several key factors for the convenient compliance test of a millimeter-wave
(mmWave) base station (BS). By using the ray-tracing simulation, we analyzed the power density
(PD) output profiles from a mmWave BS installed at an urban area. Several important points which
have a significant influence on the maximum PD levels are found from numerical simulations. The
derived key factors indicate that the convenient compliance assessment of a mmWave BS can be
performed based on the simplified model of the installation site instead of the present mandatory
procedure at the actual complex environmental profile.
Introduction
With the growing demand in mobile traffic and numbers of wireless devices in the communication

392
networks, many agencies now develop the next fifth-generation (5G) technology to realize a higher
capacity in order to support the increase in data consumption. However, this future 5G network
should require the denser installation of base stations (BSs) such as a small cell deployment
because of the larger path loss characteristic at this mmWave band [1], resulting in drastic increase
in the cost of the BS compliance assessment as the 5G service will start. Hence it is of fundamental
importance to simplify the assessment procedure of a mmWave BS mandatory for all installations at
an actual site and this study provides several important key factors for this purpose.
Materials and Methods
We used the ray-tracing method as a numerical simulation of a mmWave BS installation. The
mmWave 28GHz frequency was assumed and installations at urban hot spot areas are considered.
The BS antenna has the effective isotropic radiated power (EIRP) = 54.6dBm, horizontal beamwidth
(BW) = 10˚, vertical BW = 15˚, horizontal scan range [-40˚, +40˚], vertical scan rage [-10˚, 0˚], and
mainly the vertical polarization. The power density (PD) outputs are employed as the exposure
metric since this is used as the compliance reference level (RL) at mmWave band [2]. We conduct
numerical studies for various BS installation cases with different environmental and BS
electromagnetic (EM) characteristics. The observation heights corresponding to the human body
area are selected from the current BS compliance guideline for reasonable modeling.
Results and Discussion
Figure 1 presents its representative example of numerical simulations. Both BSs are installed at the
identical environment except the primary spot directly illuminated by the main beam peak (i.e., the
edge and the face of a building in Figure 1(a) and (b), respectively). From these similar output PD
distributions and maximum (worst-case) levels, it is clearly seen that the diffraction mechanism has
little effect on the BS compliance for mmWave frequency. In addition, we can know from additional
simulations that the EM properties (the dielectric constant and conductivity) of the building have a
significant effect on the maximum PD outputs, but its ground EM properties do not play a prominent
role in contrast. Also the vertical BW is also an important part for the BS compliance while the PD
outputs hardly change as varying the horizontal BW variation. There would be no PD dependencies
on the complex environmental geometry such as the square, junction, street or rooftop, on the other
hand strong output PD dependencies on its relative simple vertical conditions are shown, i.e., the
downtilt angle and the installation heights, etc. This could mainly come from the highly directional
characteristic of a mmWave BS antenna. From these results, we could derive several key factors
that greatly affects the maximum PD output and these factors could enable us to perform the more
convenient BS compliance test based on the simple assessment model than the present procedure
mandatory for all installed BSs at the actual environmental site.
Conclusions
In this study, we provided several key factors for the compliance assessment of a mmWave BS
based on the ray-tracing numerical simulations. The PD levels show a great dependence on these
factors extracted from the essential points of the output level characteristic. These results could be
of great help to us to perform the simpler compliance test of a mmWave BS than the assessment
procedure described in the current guideline.
Acknowledgements
This work was supported by Institute for Information & communications Technology Promotion(IITP)
grant funded by the Korea government(MSIT) (2017-0-00961, Study on the EMF Exposure Control
in Smart Society)
Figures

393
Figure 1. Power density distributions with different primary spots (a) the edge (b) the face of a
building.

394
PA-17 [14:30]
Numerical evaluation of SAR and temperature increase due to dipole antenna in human fetus
models representing actual fetal growth
Tomoaki Nagaoka1 & Soichi Watanabe1
1Electromagnetic Compatibility Laboratory, National Institute of Information and Communications Technology,
Tokyo, Japan, 184-8795
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Tomoaki Nagaoka
In this study, we estimate the specific absorption rate (SAR) and temperature increase in human
fetuses exposed to EM waves from a half-wavelength dipole antenna at 900 MHz and 2 GHz
irradiated around the abdomen of women using pregnant female models including fetuses
representing actual fetal growth.We confirmed that the peak 10g-averaged SAR in the fetus tends to
slightly increase with fetal growth and the peak 10g-averaged SAR in the fetus does not exceed the
SAR in the maternal body. We also found that the core temperature increase in the fetus is
sufficiently smaller than the peak fetal temperature increase.
Introduction
With the progress of wireless communication technology, the range of use of wireless
communication equipment has diversified, and the electromagnetic (EM) environment surrounding
us has changed considerably. Under this kind of environment, the numerical dosimetry for fetuses
during pregnancy is one of the most important issues in EM field safety, because smartphones or
tablet computers are used by women during pregnancy. Fetal size and the anatomical structures of
gestational tissues including fetal tissues vary with fetal growth during pregnancy. Moreover, the
abdominal body structure of pregnant females also changes with fetal growth. Therefore, in this
study, we estimate the specific absorption rate (SAR) and temperature increase in human fetuses
exposed to EM waves from a half-wavelength dipole antenna at 900 MHz and 2 GHz irradiated
around the abdomen of women using pregnant female models including fetuses representing actual
fetal growth [1].
Method
We used pregnant female models at 20, 26 and 29 weeks of pregnancy in this study [1]. In the
models, fetal and gestational tissues were based on fetal MRI data at these weeks of pregnancy,
which were acquired from a healthy volunteer. The pregnant female models consist of over 70
different tissue types. Each pregnant female model contains 20 different fetal and gestational tissue
types, including amniotic fluid, placenta, umbilical cord, brain, cerebrospinal fluid, eyeball, lens, lung,
stomach, liver, heart, kidney and intestines. The placentas of these models were also on the back
side of the abdomen. We placed a half-wavelength antenna at 900 MHz and 2 GHz in front of the
abdomen of the maternal body. We calculated SAR by the finite-difference time-domain (FDTD)
method. Furthermore, we solved the bioheat transfer equation and estimated the temperature
increase. Figure 1 shows the antenna positions for the abdomens of the pregnant female models.
We set the reference position of the feeding point of the antenna to 15 mm from the tip position of
the abdomen of the maternal body. The antenna was located at the reference position and its
vicinity as shown in Figure 1. We preformed the calculations with the antenna placed at 15
positions. The output power of the antenna was assumed to be 1 W.
Results and Discussion
Peak 10g-averaged SARs were calculated using a SAR averaging scheme based on the IEEE
standard guideline [2]. The SAR in the mother at 2 GHz is significantly higher than that at 900 MHz,
owing mainly to the difference in penetration depth between 900 MHz and 2 GHz. On the other
hand, the SAR characteristics of these models are similar at all stages of pregnancy. The results
indicated that pregnancy progression does not significantly affect the peak local SAR of the
maternal body. In contrast to the results from mothers, the SAR in the fetus differs slightly between

395
the models at different stages of pregnancy. The SAR in the fetus tends to increase with fetal
growth. The main reason is that the distance between the fetus and the antenna decreases with
fetal growth. Figure 2 shows the percentage differences in peak 10g-averaged SARs in fetuses to
the SARs in mothers. The SARs in fetuses are lower than those in mothers at all stages of
pregnancy. In addition, at 2 GHz, the fetal SARs are much lower than the maternal SARs. This
finding indicates that the peak 10g-averaged SAR in the fetus does not exceed the SAR in the
maternal body. The temperature increase in the fetus tends to be slightly larger at 2 GHz than at 900
MHz. However, the temperature increase in the fetus is not significantly different between the
models at different stages of pregnancy. From the results, pregnancy progression may not
significantly affect the peak temperature increase in the fetus. Figure 3 shows the percentage
differences in the peak core temperature increase in fetus to the peak fetal temperature increase.
The peak core temperature increase in the fetus is at most approximately half of the peak fetal
temperature increase. Therefore, the core temperature increase in the fetus is sufficiently smaller
than the peak fetal temperature increase.
Conclusions
We estimated the SARs and temperature increases in fetuses exposed to EM waves from a half-
wavelength dipole antenna at 900 MHz and 2 GHz irradiated around the abdomen of women during
pregnancy. We confirmed that the peak 10g-averaged SAR in the fetus tends to slightly increase
with fetal growth and the peak 10g-averaged SAR in the fetus does not exceed the SAR in the
maternal body. We also found that the core temperature increase in the fetus is sufficiently smaller
than the peak fetal temperature increase.
Acknowledgment
Part of this research is financially supported by the Ministry of Internal Affairs and Communications
of the Government of Japan.
References
[1] T. Nagaoka, T. Niwa, and S. Watanabe, “Specific Absorption Rate in Mothers and Fetuses in the
Second and Third Trimesters of Pregnancy,” International Journal Microw. Optical Tech., 9, 1,
January 2014, pp. 34-38.
[2] IEEE Standard C95.1, 2005.
Figures

Figure 1. Calculation models (antenna position)

396
 Figure 2. Percentage differences in peak 10g-averaged SARs in fetuses to SARs in mothers

Figure 3. Percentage differences in peak core temperature increase in fetus to peak fetal
temperature increase.

PA-19 [14:30]
Electromagnetic field exposure level using equivalent model of actual wireless power
transfer system
SangWook Park1, DongGeon Kim1 & BeomJin Choi1
1EMI/EMC R&D Center, Korea Automotive Technology Institute, Cheonan-si, Korea, 330-912
Keywords: Dosimetry (computational), ELF/LF, Work in Progress
Presented by: SangWook Park
Accurate dosimetry for a real wireless power transfer system requires an accurate description of the
field formed by the system. In particular, the electromagnetic field depends on factors such as the

397
construction of the transmitting and receiving coils, the circuit configuration, the input source of the
front end of the transmitting coil, and the input impedance of the rear end of receiving coil. However,
both circuit and electromagnetic simulations need to be performed to analyze the entire system,
which is a difficult task. In order to overcome this difficulty, a method using an equivalent circuit
model is proposed.
Abstract
A real wireless power transfer system consists of a resonance circuit part that relays energy, a
circuit that generates a high-frequency source in the front end, and another circuit that stores energy
in the rear end. This entire system should be considered to accurately simulate the nearby
electromagnetic fields. A three-dimensional electromagnetic simulation using a finite-difference time-
domain (FDTD) method and finite element method (FEM) is needed to analyze the electromagnetic
fields of the transmitting and receiving coils. The remaining circuit parts can be analyzed by a circuit
simulation such as the Simulation Program with Integrated Circuit Emphasis (SPICE). However,
there is no perfect numerical analysis method that can cover both the circuit and electromagnetic
simulations at once. Thus, the most convenient method is to perform an electromagnetic simulation
of the resonance part with an equivalent circuit configuration, which is created by replacing the front-
end circuit of the transmitting resonance part with voltage sourceand the rear-end circuit of the
receiving resonance part with load impedance. The latter operation is possible because the battery
may be treated as a resistance in an instantaneous circumstance.
The electromagnetic analysis of the wireless power transfer system was performed by using FEKO,
which is a commercial analysis tool developed on the basis of the method of moments (MoM). Since
the entire system is difficult to simulate, as mentioned above, the most approximate simulation to
the real system was implemented by setting voltage source for the transmitting port and load
impedance for the receiving port. When the simulation conditions of ouput power and battery voltage
are determined, battery current is determined; then, the voltage source is adjusted to obtain the
output current. In order to ensure the same simulation as the experimental conditions, current
flowing in the receiving coil and output power were measured in the experiment. Load impedance
was obtained from the relation; then, the simulation was carried out. If we want to simulate the
electromagnetic field generated from a WPT system without an experiment, the current values of
voltage source and receiving coil current can be determined by the results obtained from a circuit
simulation of the WPT system.
Acknowledgment
This work was supported by a “Development of wired quick charger (200 kW) and wired/wireless
dualuse charger (20 kW) for reducing charging time” grant from the Ministry of Trade, Industry and
Energyof Korea.

PA-21 [14:30]
STUDENT PAPER
Numerical determination of the risk of cardiac ventricular fibrillation in humans from body
currents with frequencies up to 1 MHz using outcomes of animal studies
Pia Schneeweiss1, Tobias Theiler1, Kai Jagielski1 & Dominik Stunder1
1Research Center for Bioelectromagnetic Interaction (femu), RWTH Aachen University, Aachen, Germany,
52074
Keywords: Dosimetry (computational), All Frequencies, Work in Progress
Presented by: Pia Schneeweiss
Cardiac ventricular fibrillation (CVF) is a potential cause of death following electrical accidents
occurring when the electrical activity of the heart is disrupted. In order to determine thresholds for
CVF, cadaver and animal studies were carried out in the past, which are however limited e.g. in
tissue’s electric properties and the utilized probes properties. Currently executed simulations with

398
high resolution anatomical body models traversed by body currents with frequencies up to 1 MHz
will be used for the derivation of CVF thresholds and can therefore help to prevent injuries caused
by electrical currents in the future.
Introduction
The effects of electrical current on humans and livestock are dealt with in the international standard
IEC/TS 60479-1 (IEC 2005). They depend to a large extent on the frequency, signal form, intensity
and duration of the current flow, but also on the current phase of the cardiac cycle, the current path,
the skin and inner body impedance. Together, all these parameters influence the body current.
Body current arises when a grounded person touches a conductive object on another electrical
potential and thus establishes an electrical connection between two different potentials. This leads
to a current flow through the human body. Low-frequency currents (< 100 kHz) lead to the direct
excitation of electrically conductive, excitable structures (e. g. muscles, blood vessels, peripheral
and central nervous system and sensory organs). High-frequency currents (> 100 kHz) heat the
tissue up to necroses and burns, as well as coagulation and secondary damage such as
thrombosis, embolisms, hemorrhages and edema, which may occur with a considerable time delay
(Zschiesche 2010).
A life-threatening hazard of body currents arise when the electrical current density in the heart
muscle interferes with the cardiac excitation-contraction and causes cardiac ventricular fibrillation
(CVF). CVF manifests itself in uncoordinated contractions of the entire heart musculature and
corresponds hemodynamically to a cardiovascular standstill. In the case of flickering ventricles, no
coordinated heart contractions are recorded, only a tremor of the heart muscles is detected, blood
pressure drops and blood flow is suspended. If it is not possible to stimulate the heart to coordinated
contractions through targeted countermeasures, death occurs within a few minutes (Spörri and
Stünzi 1969).
In order to determine thresholds for CVF, cadaver studies were carried out in the past, which are
however limited in tissue’s electric properties and the utilized probes’ size, relative position and
sensor sensitivity (Sam 1966). Additionally, in vivo-experiments were conducted with animals of
different body shapes and sizes, such as rats, guinea pigs, dogs, pigs, calves, etc.(Ferris, King et al.
1936). From various measurements e.g. by Jacobsen, Buntenkötter et al. (1974), Sugimoto, Schaal
et al. (1967) or Kouwenhoven and Milnor (1958) data was collected to determine the CVF threshold
for certain frequencies, flow durations, signal shapes, current levels and current paths for animals.
Since, from a cardiovascular and other physiological and pathophysiological point of view, there is a
high degree of agreement between humans and pigs, it is possible to obtain information on the
hazard of humans in different flow paths, especially from pigs (Jacobsen, Buntenkötter et al. 1974).
By implementing these results from in vivo-studies into mathematical models, the results were used
in in silico-studies, i.e. numerical simulations (Brucher and Meyer-Waarden 1985, Biegelmeier,
Kieback et al. 2003). From today's point of view, however, the models used in the studies carried out
in the past were very coarsely resolved and therefore inaccurate (distances between the calculation
points up to 16 mm).
An extendable and adaptable validated numerical body model will be developed for computational
body current dosimetry by combining results from in vivo-studies with today’s high-resolution
simulations of pig models and human models for DC to 1 MHz. This new data helps to prevent
injuries caused by electrical current of various signal forms, amplitudes and frequencies.
This conference paper presents the methodology of the outlined project.
Materials and Methods
Numerical Computations
All computations were carried out using the Sim4Life V 3.4.3.2565 simulation platform (SPEAG,
Zurich, Switzerland). The Quasi-Static Electromagnetic Solver (P-EM-QS) is ideal for the calculation
of body currents in the low frequency range. It applies the finite element method on graded voxel
meshes. The computational domain was discretized using an adaptive graded voxel mesh. The
399
computational results were evaluated with respect to the current path distribution as well as the
electric field.
Anatomical Body Models
Two different high resolution models were used to extrapolate the pigs’ results to humans: Male Pig
(version 1) from the Animal Models and Duke (version 3.1.1) from the Virtual Population (Gosselin,
Neufeld et al. 2014). As the model of the domestic pig contains numerous errors due to the
generation of the model out of MRI data, blocks with the dielectric properties of air were added, for
instance to separate the limbs (Figure 1). The body models consider skeletal muscle isotropy and
skin’s dependency on frequency and touch voltage. The models’ segmented tissues were linked to
the IT’IS tissue electric properties database version 3.0 (Hasgall, Di Gennaro et al. 2015), which
mainly refers to Gabriel et al. (1996).
Exposure Scenarios
In a first step, the dependency of body current on the current path and frequency was investigated.
All 16 IEC current paths were simulated with the 7 contact electrodes (100 x 100 mm²) positioned as
shown in Figure 2. The body current scenario was implemented as a voltage driven problem,
meaning the voltage level (touch voltage Ut) of the touched conductive object (here the contact
electrodes) was set to a fixed value and the resulting total body current (IBody) was measured with 4
surface integrals over the numerically derived internal electric field.
In a second step, the already mentioned animal studies are recreated in computational models to
reproduce and validate the published CVF’s. Subsequently, the derived CVF’s are extrapolated onto
humans with verification based on literature.
Results
Presently, a set of computational results is evaluated which will be presented at the conference as
well as an estimation of the risk of cardiac ventricular fibrillation depending on the current path and
frequency.
Acknowledgements
This work was supported by the German Social Accident Insurance Institution for the Energy,
Textile, Electrical, and Media Products Sectors (BG ETEM).
References
Biegelmeier, G., et al. (2003). Schutz in elektrischen Anlagen - Band 1: Gefahren durch den
elektrischen Strom. Berlin, VDE Verlag GmbH.
Brucher, R. and K. Meyer-Waarden (1985). Rechnergestützte Ermittlung gefährlicher Gesamtströme
im menschlichen Körper, Inst. zur Erforschung Elektr. Unfälle.
Ferris, L. P., et al. (1936). "Effect of Electric Shock on the Heart." AIEE Transactions 55(5): 498-515.
Gabriel, C. (1996). Compilation of the dielectric properties of body tissues at RF and microwave
frequencies, King's College London.
Gosselin, M.-C., et al. (2014). "Development of a New Generation of High-Resolution Anatomical
Models for Medical Device Evaluation: The Virtual Population 3.0." Phys Med Biol 59(18):
5287-5303.
Hasgall, P. A., et al. (2015, September 01, 2015). "IT’IS Database for thermal and electromagnetic
parameters of biological tissues, Version 3.0." Retrieved February 06, 2018.
IEC (2005). Effects of current on human beings and livestock - Part 1: General aspects; IEC/TS
60479-1: 2005-07, International Electrotechnical Commission.
Jacobsen, J., et al. (1974). "Beitrag zur Übertragbarkeit der Gefährdung durch elektrische Ströme
vom Modelltier Schwein auf den Menschen." Dtsch Tierarztl Wochenschr 81(9): 214-220.
Kouwenhoven, W. B. and W. R. Milnor (1958). "The Effects of High-Voltage, Low-Capacitance
Electrical Discharges in the Dog." IRE Trans Med Electron PGME-11: 41-45.
400
Sam, U. (1966). "Untersuchungen über die elektrische Gefährdung des Menschen bei
Teildurchströmungen, insbesondere bei Arbeiten in Kesseln, Behältern und Rohrleitungen (1. Teil)."
Elektromed Biomed Tech 11(4): 193-212.
Spörri, H. and H. Stünzi (1969). Pathophysiologie der Haustiere: Unter Mitarbeit zahlreicher
Wissenschaftler des In- und Auslandes. Berlin und Hamburg, Verlag Paul Parey.
Sugimoto, T., et al. (1967). "Factors determining vulnerability to ventricular fibrillation induced by 60-
CPS alternating current." Circ Res 21(5): 601-608.
Zschiesche, W. (2010). "Stromunfälle am Arbeitsplatz - Gefährdungen, gesundheitliche
Auswirkungen, ärztliche Maßnahmen." Arbeitsmed Sozialmed Umweltmed 45(4): 164-169.
Figures

Figure 1. Male Pig anatomical body model with 103 different tissues; pink blocks separate the
limbs and have the dielectric properties of air

401
Figure 2. Anatomical body model of Duke with the 7 contact electrodes (right/left hand
(RH/LH), right/left foot (RF/LF), posterior thorax (PT), anterior thorax (AX) and (posterior) seat
(PS)) and 4 surface integrals for the total body current (IBody) measurements

402
PA-23 [14:30]
STUDENT PAPER
Industrial indoor massive MIMO human EM-exposure evaluation
Sergei Shikhantsov1, Arno Thielens1, 2, Gunter Vermeeren1, Emmeric Tanghe1, Piet Demeester1, Guy Torfs1,
Luc Martens1 & Wout Joseph1
1Department of Information Technology, Ghent University/IMEC, Ghent, Belgium, 9000
2Department of Electrical Engineering and Computer Sciences, University of California Berkeley, Berkeley,
CA, USA, 94704
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Sergei Shikhantsov
We present a numerical method of estimating human exposure to the electromagnetic fields (EMF)
of a massive MIMO base station (BS) in an industrial indoor environment e.g., an assembly line or a
warehouse. The method relies on a massive MIMO channel model derived from ray-tracing
simulations with stochastically generated environments. Incoming rays at the receiver location are
used as the input for Finite-Difference Time-Domain plane-wave simulations with a realistic human
phantom to obtain EM-exposure estimate.
Introduction
In the next generation wireless access networks, increased requirements on data rates and the
number of simultaneously served users need to be met. One of the most promising sub-6GHz
technologies being developed is massive multiple-input multiple-output (MMIMO). The number of
antennas in a MMIMO base station (BS) is much larger than the number of serviced users. The high
amount of antennas allow spatial focusing of the transmitted signal at the intended user location [6].
In favorable propagation conditions [5], MMIMO provides increased capacity [9] and energy
efficiency [3].
There exist methods to assess exposure in far-field conditions. These methods use multiple plane-
wave Finite-Difference Time-Domain (FDTD) simulations with known statistics of the propagation
environment, such as the number of incident waves, their angular and power distribution [10].
However, no valid stochastic MMIMO channel model currently exist. Moreover, the channel statistics
vary significantly between users depending on the BS resource allocation. Therefore, the current
methods are not suitable for estimation of exposure to MMIMO. At the same time, certain progress
was made in deterministic MMIMO channel modeling using the ray-tracing method [4].
The problem of electromagnetic (EM) exposure estimation of MMIMO has not yet been addressed in
the literature, to the author’s knowledge. The development of a numerical technique for EM-
exposure estimation and evaluation of human EM-exposure in an indoor industrial scenario are the
main goals of this paper.
Methods and materials
We utilize stochastic geometry to generate indoor environments for ray-tracing simulations. The
environment consists of a single cuboid room with scatterers inside it, as shown in Figure 1. The
dimensions of the room are fixed at 20-by-30-by-6 m3. The material assigned for walls is a dielectric
type with ε=7, ρ=0.015 (concrete). A scatterer is a cuboid of fixed width and length (2-by-0.2 m2)
and height sampled from uniform random distribution from 2 to 4m. We applied Poisson Disc
sampling algorithm [1] to evenly distribute scatterers in the environment without intersecting each
other. In addition, all scatterers are rotated around the z-axis at a uniform random angle in [0, π].
Scatterers use perfect electric conductor material properties.
The environment contains one massive MIMO BS and a grid of receivers. The BS is a 10-by-10
rectangular array of 3.7 GHz λ-separated isotropic antennas, resulting in the overall size of 0.7-
by-0.7 m2. The BS is positioned at the 3m height and transmits with 1 W (30 dBm) of power. Each
receiver in the grid is positioned at the height of 1.5m and equipped with a single isotropic antenna.

403
Isotropic antennas are used to simplify further analysis.
The calculated channel is described with complex channel matrix H, which elements has the form
where Pnk is the normalized power at the kth receiver when only nth BS element is
radiating, and θis the phase of the voltage across a matched load at that receiver.
To calculate the received power, REMCOM Wireless InSite ray-tracing software
combines E-fields of all received rays at a given antenna coherently, according to
where λ is the carrier wavelength, Z0 is the impedance of free space and Ei
is the complex electric field of the ith ray [8].
In this work, we only consider the case of Maximum Ratio Combining
(MRC) precoding. MRC precoding matrix is defined as W = αHH, where
normalization coefficient αis chosen such that W is unitary [4]. If the BS transmits with the total
power of P to the kth receiver, then its received power is
Using equation (2) we find the total electric field magnitude at the location of
kth receiver.
To estimate the exposure of an adult man at the location of kth receiver we modeled and simulated
the exposure as multiple plane-waves incident on Duke phantom (a 34 years old adult male from the
Virtual Population v.3.1) [2] using an FDTD solver. Frontal exposure was assumed (φ in [0,π],θ = 0)
with 18 plane-waves having equal angular spacing. Equal power densities were assigned to all 18
plane-waves, such that their algebraic sum equals to the power density at the receiver’s location
found with the ray-tracing method. Phases of the waves are adjusted for constructive interference at
the center of the domain. The center of the phantom’s head is aligned with the center of the domain.
The described approach is simplified. It approximates the angular distribution of the incident rays,
which a real massive MIMO system would have, by a combination of several plane waves with fixed
wave-vectors and the spatial focusing is created artificially. However, massive MIMO spatial
multiplexing gain is accounted for.
Results
The presented method was applied to calculate the exposure of a user 15 m away from the BS, and
only experiencing non line-of-sight signals (see Fig. 1). The received power was found to be around
5.4mW. From (2), sum E-field magnitude is around 88V/m, or 4.9V/m per plane-wave.
In the free-space FDTD simulation, the RMS E-field magnitude in the hot-spot (Figure 2, left) was
approximately 63V/m. Simulation with the phantom yielded peak-spatial Specific Absorption Rate in
10g (psSAR10g) of around 76mW/kg. The resulting exposure is just below ICNIRP basic restrictions
for head and trunk [7] (80 mW/kg). Its relatively high value, given that the BS radiates 30 dBm, can
be by part explained by focusing of electromagnetic energy in a narrow spatial region by MMIMO
system.
Conclusion and future work
In this paper we introduced a numerical approach for human EM-exposure estimation to MMIMO
system, that combines ray-tracing and FDTD numerical methods. It was applied to assess the
exposure of an adult male in an industrial indoor environment with a single MMIMO BS.
The ongoing research is aimed at including direction of arrival, amplitude and phase properties of
the incoming rays in FDTD simulations. Large number of environments having similar propagation
characteristics will be simulated in order to evaluate statistical properties of a human EM-exposure.
References
[1]D.Dunbar and G.Humphreys. A spatial data structure for fast poisson-disk sample generation.
ACM Transactions on Graphics (TOG), 25 (3): 503–508, 2006.
[2]M.-C. Gosselin, E.Neufeld, H.Moser, E.Huber, S.Farcito, L.Gerber, M.Jedensjö, I.Hilber,

404
F.DiGennaro, B.Lloyd, etal. Development of a new generation of high-resolution anatomical models
for medical device evaluation: the virtual population 3.0. Physics in medicine and biology, 59 (18):
5287, 2014.
[3]E.G. Larsson, O.Edfors, F.Tufvesson, and T.L. Marzetta. Massive MIMO for next generation
wireless systems. IEEE Communications Magazine, 52 (2): 186–195, 2014.
[4]D.Loschenbrand, M.Hofer, and T.Zemen. Ray-tracer based channel characteristics for distributed
massive MIMO. 2017.
[5]T.L. Marzetta. Noncooperative cellular wireless with unlimited numbers of base station antennas.
IEEE Transactions on Wireless Communications, 9 (11): 3590–3600, 2010.
[6]T.L. Marzetta. Massive MIMO: an introduction. Bell Labs Technical Journal, 20: 11–22, 2015.
[7]I.C. onNon-Ionizing RadiationProtection etal. ICNIRP guidelines for limiting exposure to time-
varying guidelines for limiting exposure to time-varying electric, magnetic and electromagnetic fields.
Health Phys, 74: 494–522, 1998.
[8]REMCOM. Wireless InSite 3.2.0 Reference Manual. 2017. URL http://www.remcom.com/-
WirelessInSite.
[9]E.Telatar. Capacity of multi-antenna gaussian channels. Transactions on Emerging
Telecommunications Technologies, 10 (6): 585–595, 1999.
[10]G.Vermeeren, W.Joseph, and L.Martens. Statistical multi-path exposure method for assessing
the whole-body SAR in a heterogeneous human body model in a realistic environment.
Bioelectromagnetics, 34 (3): 240–251, 2013.

Figures

Figure 1. An example of a ray-tracing simulation environment. The MMIMO BS is supplied with

1 W of power. The power at each node of the receiver grid is shown and intended receiver is
indicated.

405
 Figure 2. E-filed RMS magnitude in free space (left) and in presence of the phantom’s head
(right). The location psSAR10g peak-cube shown in green coincides with the free-space E-field
area of focus.

PA-25 [14:30]
Influence of anatomical skin thickness of children on the whole-body SAR for future 5G-
frequencies
Richard Überbacher1 & Stefan Cecil1
1EMC & Optics Division, Seibersdorf Laboratories, Seibersdorf, Austria
Keywords: Dosimetry (computational), RF/Microwaves, Completed (published)
Presented by: Richard Überbacher
In this work we investigate the influence of the anatomical skin thickness on the whole body SAR of
children regarding to the ICNIRP 1998 (gen.pub.) guidelines. Especially for the new upcomming 5G-
frequencies, a closer realistic look and anatomical modelling of the border layers (e.g. skin and SAT)
is recommended.
Introduction
In this work we investigate the influence of the anatomical skin thickness of children on the whole
body SAR (WB-SAR) regarding the ICNIRP 1998 guidelines [1]. As known from former publications
[2] [3] the reference levels for children are in some mobile communication frequencies (DCS-1800
and UMTS-2100) exceeded due the smaller physical size. The new upcoming 5G generation
according to the GSA (Global mobile Suppliers Association) will be operated additionally to the
former analog radio frequencies from the so called “digital dividend” also in higher ranges, so e.g. in
Europe in the 3400-3800 MHz band. In the US the 5G will be operated in 3100-3550 MHz and
3700-4200 MHz, and in Japan the 3600-4200 MHz band is intended. Because of the lower
penetration depth of the electromagnetic waves in higher frequencies more energy is deposited near
the surface structures. An anatomical realistic modeling of the tissues close to the surface (e.g. skin
and SAT (subcutaneous adipose tissue) becomes more importance for the dosimetry. The main
focus in this work is the impact of the skin thickness difference from children compared to adults.
Materials and Methods
In a former project with ITIS (CH) we were involved together in the segmentation of the first

406
generation of the Virtual Family (v1.0) for the BFS [2]; in concrete the 5 years old girl “Roberta” and
the 8 years old boy “Dizzy”. At that time they was segmented with uniformly 2 pixel of skin outside,
corresponding to 1.736 mm what’s a acceptable mean value for adults. It has to be said in
advanced, that the skin in the original MR pictures is not to identify clearly because of the small
dimensions, the low grey contrast, and also the artifacts from movement (e.g. breathing) of the body
during the scans (Figure 1). It’s known now (i.e. [4] or [5]) that children typically have a thinner skin
than adults, so that 2 pixel of skin may be to thick, and only 1 pixel of skin (corresponds to 0.868
mm) may be anatomically more realistic; apart from the fact that the skin thickness varies over the
entire body.
The study of Seidenari et al. [4] reports mean values for the skin thickness for a group of 42 children
in the age from 2-13 years measured on 8 different points. He found the lowest value with 0.89
mm±0,14(SD) for the volar forearm, and the highest value of 1.43 mm ±0,43(SD) for the forehead.
The measurements where carried out with 20 MHz ultrasound. Under the use of the original MR
slices from the both models the skin was segmented in new different layers thickness, to investigate
the impact on the whole body SAR. Since only discrete steps where possible (x/y resolution of the
MR pictures: 0,868 mm/pixel), a range from n=1 to 4 pixel was investigated. Additionally to the
variation of skin thickness the material parameter for the skin was varied in “wet skin” and “dry skin”.
As exposure as required from the ICNIRP guidelines a plane wave used with horizontal (TE) and
perpendicular (TM) polarization for the E-field, and frontal and left side as incident direction. For the
simulation the platform Semcad-X was used. The grid resolution was set to variable grid 0.5-1.5 mm
(grading ratio 1.3) for the frequencies up to 2100 MHz, and to variable grid 0.3-0.8 mm (grading ratio
1.3) for the frequencies above. The material parameter was set according to the Virtual Population
database from ITIS [6].
Results
The results for various typical mobile communication frequencies from 400 MHz up to 6 GHz are
shown in Figure 2 for the girl “Roberta” (5y) and the boy “Dizzy” (8y) regarding the ICNIRP 1998
(gen.pub) guidelines. The depicted values represent the maxima found from the different exposure
scenarios and polarisations.
For Roberta and Dizzy the maximum SAR-WB value (with a skin thickness of 2 pixels) was found for
the UMTS frequency at 2100 MHz. If the skin size changes to 1 pixel, for “Dizzy” the highest value
can found at the 5G band at 3500 MHz, while for “Roberta” the situation does not change. The
calculated SAR values according to the ICNIRP 1998 (gen.pub.) guidelines are shown in Figure 3
as relative percentage to the given limit for the whole body SAR. The maximal exceeded values for
a skin thickness of 1 pixel (corresponding to 0.868 mm) for Roberta was 148.3% for UMTS-2100
and for Dizzy 140.9% for 5G-3500. Compared to a segmentation with 2 pixel skin thickness
(corresponding to 1.736 mm) for Roberta the maximal SAR value is 11.6% lower, while for Dizzy the
difference is 28.4% for 5G-3500, what is a remarkable (and surprising) number.
Conclusion
The anatomical segmentation accuracy is a important parameter for the uncertainty budget in the
numerical dosimetry. This aspect should be treated with more attention when doing numerical
computations, especially for the upcoming new 5G generation and the therefore used higher
frequencies.
References
[1] ICNIRP, "Guidelines for limiting exposure to time-varying electric, magnetic and electromagnetic
fields (up to 300 GHz)," Health Physics, vol. 74, pp. 494- 522, 1998.
[2] A. Christ, G. Schmid, M. Zefferer, R. Überbacher, M. Lichtsteiner, E. Neufeld, et al., "Numerische
Bestimmung der Spezifischen Absorptionsrate bei Ganzkörperexposition von Kindern,"
Abschlussbericht zum Forschungsauftrag StSch 4517 des deutschen Bundesamtes für
Strahlenschutz, (Juli 2009), http://doris.bfs.de/jspui/handle/urn:nbn:de:0221-201004071342, 2009.
[3] Sven Kühn, Wayne Jennings, Andreas Christ, and N. Kuster, "Assessment of induced radio-
407
frequency electromagnetic fields in various anatomical human body models," Phys. Med. Biol. 54
(2009) 875–890, 2009.
[4] S. Seidenari, G. Giusti, C. Bertoni, and G. Pellacani, "Thickness and echogenicity of the skin in
children as assessed by 20-MHz ultrasound," Dermatology, vol. 201(3), pp. 218-222, 2000.
[5] J. G. B. Derraik, M. Rademaker, W. S. Cutfield, T. E. Pinto, S. Tregurtha, A. Faherty, et al.,
"Effects of Age, Gender, BMI, and Anatomical Site on Skin Thickness in Children and Adults with
Diabetes," PloS One, vol. 9(1): e86637, 2014.
[6] https://www.itis.ethz.ch/virtual-population/tissue-properties/overview/
Figures

Figure 1. “Roberta” (girl 5y) from the Virtual Family (v1.0); example slice in the torso region

408
Figure 2. Results for the SAR values for different values of skin thickness (1 to 4 pixel) for
”Roberta” and “Dizzy” according to the ICNIRP 1998 (gen.pub.) guidelines

Figure 3. Relative whole body SAR (SAR-WB) values regarding to the ICNIRP 1998
(gen.pub.) guidelines

409
PA-27 [14:30]
Numerical RF dosimetry in 2D and 3D in vitro skin cell cultures
Rosanna Pinto1, 2, 3, Stefania Romeo2, 3 & György Thuroczy3
1Division of Health Protection Technologies, ENEA, Roma, Italy
2Institute for Electromagnetic Sensing of the Environment , CNR, Naples, Italy
3Department of Non-Ionizing Radiation , National Public Health Institute, Budapest, Hungary
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: György Thuroczy
The new developments of wireless devices operate at higher radiofrequencies with respect to the
past therefore larger exposure of the human skin to such electromagnetic fields is expected. In the
present work, a numerical dosimetry analysis was carried out to identify the optimal exposure
conditions, in terms of efficiency and uniformity of SAR distribution of 2D and 3D in vitro skin cell
cultures to 1950 MHz and 2450 MHz. At 2450 MHz the RF exposures were simulated in Wire Patch
Cell (WPC) exposure system, at 1950 MHz a standard waveguide exposure system was used. Both
WPC and waveguide systems provide a good efficiency, in terms of induced SAR in the skin
samples to incident power.
Introduction
Due to the wide use of wireless technologies (UMTS, DECT, LTE, WiFi) based on radiofrequency
fields at higher frequencies than GSM 900 MHz systems and the forthcoming introduction of 5G
devices, which will operate at even higher frequencies, a larger exposure of the human skin to such
electromagnetic fields (EMF) is expected. In fact, as the frequency increases, the penetration depth
of the EMF decreases, and most of the electromagnetic power is absorbed by the superficial body
tissues. Therefore, a deep understanding of the interaction between RF and the human skin is
fundamental when addressing safety issues (Schmid and Kuster, 2015). Experimental studies
involving the exposure of in vitro skin cell cultures for the research of EMFs under well controlled
electromagnetic and environmental conditions are a fundamental step in this direction.
In the present work a numerical dosimetry analysis was carried out to identify the optimal exposure
conditions, in terms of efficiency and uniformity of SAR distribution of 2D and 3D in vitro skin cell
cultures to 1950 MHz and 2450 MHz, by means of different exposure systems.
Methods
Three types of cell cultures were used in the numerical models: primary human keratinocyte cell
cultures (2D monolayer), reconstructed human epidermis (RhE) model (3D) and full thickness (FT)
in vitro skin model (3D). For exposure at 2450 MHz (WiFi) a Wire Patch Cell (WPC) applicator was
considered, as originally designed by XLIM (France) and modified by ENEA (Italy) (Paffi, 2010).For
exposure at 1950 MHz (UMTS)a standard waveguide exposure system was used as previously
published by Calabrese et al (2006).
Simulations were performed using the commercial software CST MicrowaveStudio®, 2016, based
on the Finite Integration Technique, to evaluate scattering parameters as well as SAR distribution
inside the samples. The standard deviation over mean value was considered as a measure of the
SAR inhomogeneity inside the sample.
Numerical dosimetry at 2450 MHz in Wire Patch Cell (WPC)
The geometrical model of WPC was described in Paffi et al (2010). The Primary Human
Keratinocyte Cell Cultures (2D monolayer) were modelled as 2 ml of culture medium(er = 77, s = 2.2
S/m) in four 35 mm Petri dish. The estimation of the mean SAR in the monolayer of cells was
performed by a vertical extrapolation applied starting from the mean SAR values calculated in all
layers of liquid: third order polynomial functions were used for extrapolation.
The Reconstructed Human Epidermis (RhE, EPI-200) and full thickness (FT, EFT-300) 3D cell

410
cultures were modelled as a sample of skin (er = 38.8, s = 1.25 S/m from CST library) placed inside
a 12 mm inner perspex box inserted in 35 mm Petri dish: both Petri dishes and inner Perspex boxes
were filled with 1 ml of culture medium. A thickness of 200 mm and a diameter of 9 mm were
considered for EPI-200sample and a thickness of 1mm and a diameter of 9 mm were considered for
EFT-300sample.
In all cases, the centre of each Petri dishes was placed at 52 mm from the centre of the WPC, the
medium has a distance of 2,6 mm from the plate of WPC. Moreover 20 lines per wavelength were
set for the global mesh, whereas, inside the sample, a finer mesh was used to take into account the
thickness of the simulated samples. ‘Open’ boundary conditions were used to terminate the
calculation domain. Results will be provided in terms of efficiency and non-uniformity degree of SAR
(CV) in the samples.
Numerical dosimetry at 1950 MHz in standardized waveguide system
A short-circuited WR-430 waveguide was chosen as applicator to simulate skin model exposure
at1.95 GHz, by adopting the same strategy of previous works to optimize efficiency and SAR
uniformity inside the samples [Calabrese et al 2006; Romeo et al 2013]. The simulations were
performed on one, two, or four 35-mm Petri dishes filled with 2-ml samples in case of keratinocytes
in monolayer (εr = 77; σ= 2.2 S/m). In the case of EpiDerm 3D Skin Tissue Model (EPI-200, εr =
38.8; σ= 1.25 S/m) two or four samples were considered. In all cases, the sample aspect parallel to
the electric field in the empty waveguide was chosen, and the simulations were run by changing the
distance of the samples centres from the short circuit, as well as the relative vertical distance
between samples, in order to identify optimal exposure conditions. Results will be provided in terms
of efficiency and non-uniformity degree of SAR (CV) in the samples.
Results
In the WPC exposure system the simulated S11 at 2450 MHz of keratinocytes monolayer was -17,9
dB and in case of 3D cell cultures -11,9 dB respectively. The results of normalized SAR values to 1
W net input power at 2450 MHz are summarized in Table 1. The normalized SAR is the highest in
2D Keratinocytes monolayer (Paffi, 2010), while The SAR in case of 3D full thickness is 0,71
(W/kg)/(W).
Table1: Results of SAR values at 2450 MHz evaluated for 1 W of net power in
Wire Patch Cell (WPC)

Mean SAR
Cell culture CV [%]
[W/kg)/W)]

Keratinocytes monolayer* 1,10 29

EpiDerm 3D
0,88 68
Model (EPI-200)

Full Thickness 3D
0,71 54
Model (EFT-300)

*from Paffi et al (2010)

The normalized SAR to 1 W net incident power in the waveguide system at 1950 MHz depends on
the number of samples inserted into the chamber (Table.2)
Table2: Results of SAR values at 1950 MHz evaluated in samples for
1 W of net RF power in WR430 waveguide

411
 Mean SAR
Cell culture CV [%]
[W/kg)/W)]

Keratinocytes monolayer with Sample#1: 227 Sample#1: 32

2 samples Sample#2:229 Sample#2: 32

EpiDerm 3D
Skin#1: 178,5 Skin#1: 15
with 2 samples
Skin#2: 177,6 Skin#2: 15
Model (EPI-200)

Skin#1: 7,04 Skin#1: 29

EpiDerm 3D
Skin#2: 6,88 Skin#2: 29
with 4 samples
Skin#3: 5,75 Skin#3: 56
Model (EPI-200)
Skin#4: 5,63 Skin#4: 56

Conclusions
The numerical simulations of different in vitro skin cell culture were performed successfully. No
similar simulations of SAR in 3D skin cell cultures were performed in the literature up to now. Both
the WPC and waveguide system guarantee a good efficiency of exposure, therefore we don’t need
too high RF amplifier in the experimental study. A good degree of uniformity was obtained in the
case of the 2D skin culture, but not in the case of 3D skin cultures. The preliminary data on 3D skin
cultures (both EPI-200 and EFT-300) exposed in WPC operating at 2450 MHz present the
coefficient of variation beyond the expected 30%, therefore future simulations will be carried out to
optimize the exposure conditions. The numerical SAR dosimetry in WPC at 1800 MHz chamber was
also performed, but in this condition the efficiency was too low for experimental realization. In the
WR430 waveguide system the numerical dosimetry was carried out in order to assess the optimum
orientation and positioning from the waveguide termination of the circular Petri dishes inside the
waveguide chamber, obtaining very satisfactory performances of the system when the electric field
is parallel to the sample. In the 3D cell culture model (EPI-200) satisfactory results were obtained
with up to two samples, while the efficiency and the CV were reduced with four samples.This study
was done for the ANSES project called “Cellular response to co-exposure of radiofrequency (RF)
and solar ultraviolet (UV) radiation in human in vitro skin model (SKIN-RF)”.
References
Calabrese ML, d’Ambrosio G, Massa R, and Petraglia G (2006) A High-Efficiency Waveguide
Applicator for In Vitro Exposure of Mammalian Cells at 1.95 GHz IEEE Transactions on Microwave
Theory and Techniques, 54(5): 2256-2264.
Paffi A, Liberti M, Lopresto V, Merla C, Lodato R, Lovisolo GA, Apollonio F (2010) A wire patch
cellexposuresystemforin vitro experiments at Wi-Fi frequencies, IEEE Transactions on Microwave
Theory and Techniques, 58(12): 4086-4093.
Romeo S, D’Avino C, Pinchera D, Zeni O, Scarfi MR, Massa R. (2013) A waveguide applicator for in
vitro exposures to single or multiple ICT frequencies, IEEE Trans. Microwave Theory and
Techniques, 61(5): 1994-2004.
Schmid G. and Kuster N. (2015) The Discrepancy Between Maximum In Vitro Exposure Levels and
Realistic Conservative Exposure Levels of Mobile Phones Operating at 900/1800 MHz,
Bioelectromagnetics, 36:133-148.

412
PA-29 [14:30]
Microenvironmental RF exposimetry in kindergartens
Peter Necz1, Noemi Nagy1 & György Thuroczy1
1Department of Non-Ionizing Radiations, National Public Health Institute, Budapest, Hungary, 1221
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: György Thuroczy
The aim of present study was to determine RF exposure the level in the microenvironment of
kindergartens in Hungary. Our assumption was that the RF exposures emitted by indoor wireless
devices (DECT, WiFi) in kindergartens or in nurseries’ (office) rooms are usually higher than the RF
radiation from base stations. Because the Wi-Fi routers and DECT telephones are usually placed in
the offices, the exposure coming from wireless sources in nurseries’ room are significantly higher
than the RF exposure from outdoor sources. The level of RF exposure in children’s room is similar
to the indoor residential exposure of general public if the indoor wireless devices are placed out of
their room.
Introduction
The aim of this study was to measure the indoor microenvironmental level of RF exposure in
kindergartens. Our main hypothesis was that the RF components of indoor exposure emitted by
household wireless devices (e.g. DECT, WiFi, Bluetooth, etc.) are already higher than the exposure
from downlink frequency bands of mobile base stations. We also supposed that the indoor exposure
levels are much lower in the children’s room than in the nurseries’ office room where the typical
indoor wireless devices are usually placed.
Methods
In this study MVG EME Spy 121 exposimeter was used to measure and recording the ambient RF
electric field in 12 predefined RF channels between 88 MHz and 2500 MHz: FM (88-108 MHz), TV3
(174-223 MHz), TETRA (380-400 MHz), TV4&5 (470-830 MHz), GSM Tx (880-915 MHz), GSM Rx
(925-960 MHz), DCS Tx (1710-1785 MHz), DCS Rx (1805-1880 MHz), DECT (1880-1900 MHz),
UMTS Tx (1920-1980 MHz), UMTS Rx (2110-2170 MHz), WiFi (2400-2500 MHz). The selected
exposimeter has an isotropic probe. The lower and upper detection limits were 0,05 V/m and 10 V/m
respectively. The sample rate can be selected between 4s – 255s. The data recording was
performed in 15 kindergartens in the capital of Hungary (Budapest) during 5 days. The exposimeter
was placed for 24 hours in the nurseries’ office room then for 24 hours in each (2 or 3) children’s
room. The sampling rate of the recording was 60s.The measurements were taken in Budapest the
capital of Hungary between January and May 2017.
Results
The recorded mean values of total RF exposure were less than 0.25 V/m which is several times
lower than the public limits of RF exposure (41-60 V/m). Because the Wi-Fi routers and DECT
telephones are usually in the nurseries’ office room, the exposure coming from indoor sources in the
mentioned rooms is significantly higher than the exposure from outdoor sources’. The RF exposure
in the children’s room is equivalent to typical the indoor town residential exposure of general public.
Discussion
The ratio between the exposure of downlink and non-downlink frequency bands is significantly
depends on the site and the microenvironment inside the building. Since the indoor wireless RF
sources work with much less power than the base stations thus the RF field strength decrease
rapidly with the distance.
The exposure from indoor sources (Wi-Fi, DECT, GSM uplink) is much lower in children’s room
when these sources are out of these rooms. If the indoor wireless sources are placed in another
room where the children use to be staying the RF exposure is similar to typical residential level.
Therefore in kindergartens and nurseries it is recommended to operate these devices only in places
where children are not present (e.g. in nurseries’ office room).

413
References
Péter Juhász, József Bakos, Noémi Nagy, Gábor Jánossy and György Thuróczy, RF personal
exposimetry on employees of elementary schools, kindergartens and day nurseries as a proxy for
child exposures, Progress in Biopysics and Molecular Biology 2011/1-7
György Thuróczy, Ferenc Molnár, , Gábor Jánossy, Noémi Nagy, József Bakos and Judit Szabó,
Personal RF exposimetry in urban area, Ann. Telecommunication 2008 63:89-96
Figures

Figure 1. Recording of RF exposure presented by EME Spy 121’s software. The recorded
frequency bands can be seen on the right, the measured electric field on the left.

414
 Figure 2. The distribution of indoor and non-indoor sources at one of the examined
kindergarten. In this box-plot figure in the rectangle the median is inside, the 25% 75%
percentiles are the edges. The shoes show the minimum and maximum or the P25-IQR and
P75+IQR. The points show the extreme if different.

PA-31 [14:30]
Study on measurement method for human body exposure of electromagnetic wave in pulsed
signals
Dong Geun Choi1, Jongil Lee1, Judong Jang1 & Yongki Kwon1
1EMF,National Radio Research Agency, Naju, Korea, 58323
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Dong Geun Choi
A measurement method of the pulsed electromagnetic waves using a spectrum analyzer was
proposed to verify the compliance of the pulse type signal with the electromagnetic wave human
protection standards. For this purpose, various methods for measuring the electromagnetic wave of
the pulsed signals using a signal generator and a spectrum analyzer in a laboratory environment
were reviewed. The measurement method was suggested through comparison and verification even
in the operating environment of the radio station which generates the actual pulsed signal.
Recently, as human influence of electromagnetic waves generated in a pulse type high power radio
station has emerged as a social issue, it has been necessary to review the criteria and
measurement methods. Therefore, in this paper, we analyze the domestic and international human
body protection standards for pulsed electromagnetic waves and propose a measurement method
for measuring electromagnetic waves of human exposure using the zero span mode of the spectrum
analyzer. The high-power base station of pulsed signals is mainly used for detecting and tracking
objects such as ships or airplanes at long distances. The pulsed high power base station for this
purpose intermittently radiate the pulse signal, so that the average value is not high, but the peak
value is very high. In order to verify this, we measured and analyzed the electromagnetic field
strength considering the pulse signal characteristics such as pulse width, pulse repetition period,
and pulse interval in the laboratory environment and actual radio station operating environment.
415
Based on these results, we propose a method for evaluating human body exposure to
electromagnetic waves by applying the optimal spectrum analyzer setting(resolution bandwidth
(RBW), span, sweep time, trace, etc.) method which can judge whether the human body protection
standard of pulse type electromagnetic wave is complied with.
PA-33 [14:30]
A novel magnetic field sensor for the compliance with EMF Directive 2013/35/EU
Valerio De Santis1, Tommaso Campi1, Silvano Cruciani1 & Mauro Feliziani1
1Department of Industrial and Information Engineering and Economics, University of L'Aquila, Italy, Italy,
67100
Keywords: Dosimetry (measurements), ELF/LF, Work in Progress
Presented by: Valerio De Santis
In July 2016, the Directive 2013/35/EU has become in force posing new requirements on European
Union (EU) employers to evaluate the exposure of workers to electromagnetic fields (EMFs).
Current instrumentation used for the EMF evaluation is insufficient to fully demonstrate compliance
with exposure limits. This is particularly true in the low frequency (LF) range, where existing
commercial field sensors have large probe dimensions making the measurement of near-field
magnetic field sources not reliable. The aim of this study is to develop a novel magnetic field sensor
to facilitate the compliance assessment against the EMF Directive limits in the LF range.
Summary
Ubiquity of information and new emerging technologies, such as Wireless Power Transfer (WPT)
and Long Term Evolution (LTE), are posing human beings always more exposed to electromagnetic
fields (EMFs). In order to prevent the general public from excessive EMF exposure, the International
Commission on Non-Ionizing Radiation Protection (ICNIRP) has established several guidelines
which are non-binding to the several Countries [1], [2]. However, since July 2016 every employer
within the European Union (EU) will have to evaluate the exposure of their workers in accordance
with the new requirements of Directive 2013/35/EU [3], named EMF Directive later on. This is a
legally-binding and very challenging task for employers, who may lack the resources and expertise
required to show compliance with EMF limits.
For a large number of workplaces at industries utilizing EMFs, such as in the power electronics,
automotive and metal fabrication, the current instrumentation used for the EMF evaluation is
insufficient to fully demonstrate compliance with exposure limits [4]. This is particularly true in the
low frequency (LF) range (0-100 kHz), where localized or highly non-uniform in space and time field
sources are often encountered. Indeed, existing commercial field sensors have large probe
dimensions making the measurement of near-field magnetic field sources not reliable.
The aim of this study is to develop a novel magnetic field sensor to facilitate the compliance
assessment against the EMF Directive limits in the LF range. Specifically, a very small probe of
diameter equal to 2 cm (see Fig. 1) has been realized thanks to 3D printing technology. A
measurement setup with a Helmholtz coil magnetic field source has then been carried out to
calibrate the proposed sensor. Finally, some measurement tests have been performed with realistic
LF sources, such as a WPT system to recharge electric vehicles (EVs). In the final version, more
details about the sensor and measurement results for different LF sources and workplaces will be
provided.
References
[1] International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines for limiting
exposure to time-varying electric, magnetic and electromagnetic fields (up to 300 GHz),” Health
Phys., vol. 74, no. 4, pp. 494–522, 1998.

[2] International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines for limiting
exposure to time-varying electric and magnetic fields (1 Hz to 100 kHz),” Health Phys., vol. 99,

416
 no. 6, pp. 818–836, 2010.

[3] EN Directive 2013/35/EU, “on the minimum health and safety requirements regarding the
exposure of workers to the risks arising from physical agents (electromagnetic fields) (20th
individual directive within the meaning of article 16(1) of directive 89/391/EEC) and repealing
directive 2004/40/EC,” 2013.

[4] V. De Santis, T. Campi, S. Cruciani, and M. Feliziani, “Novel sensor concepts for the compliance
with the EMF Directive 2013/35/EU,” 2016 IEEE Sensors Applications Symposium, 20-22 April
2016, Catania, Italy.
Figures

Figure 1. Left to right: design of the sensor, final 3D CAD model and realized sensor.

PA-35 [14:30]
Development of solutions to reduce workers exposure to ELF-EMF in live line conditions
Luis Domingues1, Athanasio Mpalantinos1, Carlos Barbosa1 & Paulo Oliveira1
1Transmission Lines and Equipment Department, Electric Energy Research Center - CEPEL, Rio de Janeiro,
Brazil, 21941-911
Keywords: Dosimetry (measurements), ELF/LF, Work in Progress
Presented by: Luis Domingues
There are situations, in the operation of an electrical transmission network, when, either routine or
emergency maintenance activities, must be performed without disconnecting the transmission line
(live line work). Those activities are necessarily executed at short distances from electrically
energized parts, where electric and magnetic fields are presumably high. This article presents the
sequence of studies - computational modeling, laboratory and field measurements - developed in
order to control occupational exposure of electricians performing maintenance activities on
energized transmission lines and substations.
In Brasil exposure to electromagnetic fields is governed by a National Law that determines
application of exposure limits endorsed by World Health Organization – WHO. In the present
moment (2018) The National Regulating Agency – ANEEL - enforces the application of ICNIRP
limits, complemented by IEEE Standard in the gaps of the first, such as limits for DC installations.
During the years 2010-2014 all transmission lines, substations and power plants in Brasilian Electric

417
Power Sector were verified, either by calculation or by measurement, and in those cases where
limits were violated, adequate measures were adopted to ensure compliance.
In synthesis it can be said that all cases of general public exposure are within prescribed limits, and
the vast majority of occupational exposure situations are satisfactory as well. There are still a few
cases of non-compliance to limits, all in the domain of occupational exposure in specific situations.
The vast majority of non-compliance situations are related to live line maintenance exposure. In
situations where workers are necessarily close to high voltage/current sources, high exposure is
difficult to eliminate and must be studied in order to develop solutions that help bring levels within
acceptable limits.
This work describes the ensemble of studies, simulations, measurements and development of
specific devices to help maintain exposure of electricians in acceptable levels during maintenance
activities in energized high voltage transmission lines. In short the following steps were taken:
Inventory of maintenance routines, especially displacement paths and working positions in all
transmission tower designs;
Mapping, by measurement, electric and magnetic fields at those critical positions (see figure 1);
Evaluation of modifications in routine activities in order to avoid critical zones;
Specification of zones and activities where special protective equipments must be used;
Design, using computational simulation, of special shielding devices, to allow work in certain
situations (see figure 2);
building of prototypes of protective devices, with the requisites of portability and lightness, for
field use, and laboratory tests to verify effective performance.
Figures

418
Figure 1. Electric Field mapping in different tower designs

419
 Figure 2. Conception of lateral shielding to reduce electric field exposure

PA-37 [14:30]
STUDENT PAPER
MRI patient exposure - characterization and sequence-comparisons
Jennifer Frankel1, Kjell Hansson Mild1 & Jonna Wilen1
1Department of Radiation Sciences, Umeå University, Umeå, Sweden, 90187
Keywords: Dosimetry (measurements), ELF/LF, Completed (unpublished)
Presented by: Jennifer Frankel
The unique mixture and intensity of magnetic fields experienced by the patient during an MRI exam
are quite complex from an exposure perspective. We measured the RF- and gradient magnetic
fields inside a 3T MRI scanner and extracted several exposure parameters to give a comprehensive
description of the exposure of each sequence. We found significant differences between sequences
for each of the exposure parameters studied. We also found within-sequence variation. This means
that MRI exposure is not only sequence-specific but also specific to the individual patient and exam
occurrence, a complexity which must be kept in mind when performing exposure assessment for
epidemiological studies on MRI.
The unique mixture and intensity of magnetic fields experienced by the patient during an MRI
(magnetic resonance imaging) exam are quite complex from an exposure point of view. In addition
to the strong static magnetic field, there is a pulsed radiofrequency (RF) field and a switched
gradient magnetic field contributing to the EMF exposure.
Acute effects related to this kind of exposure are well known and constitute the basis for current
safety guidelines for MRI. Long-term effects, however, are not well understood, and if they exist
epidemiological studies would be one path to uncovering them. A major challenge with
epidemiological research is adequate exposure assessment and the complex exposure environment
of an MRI exam has made this difficult.
Our aim with this study was to characterize the RF and gradient field exposure in a way that would
be useful in future epidemiological exposure assessment and to compare MRI sequences to see if
they differ in terms of exposure.
We measured the RF- and gradient magnetic fields inside a 3T MRI scanner during the scanning of
a few common MRI sequences and extracted several exposure parameters to give a
comprehensive description of the exposure of each sequence. The sequences were compared to
each other for each exposure parameter, to see if there were any significant differences.
420
Furthermore, MRI sequence settings, such as slice thickness and field of view, were varied to see if
that would affect the resulting exposure, thereby adding possible within-sequence exposure
variation to any existing between-sequence variation.
We found significant differences between sequences for each of the exposure parameters studied,
and different sequences ranked highest for different exposure parameters. For example, the
difference in gradient field dB/dtmax between the sequences with highest and lowest dB/dtmax
exposures was 50% (P = 0.01). For the gradient field Bmax the exposure of one sequence was
nearly 4 times higher than that of another (P = 0.02). Figure 1 shows the large variation in exposure
between 5 different sequences for another gradient field exposure parameter.
The difference in root mean square (rms) of the RF B1 field was 4-fold between the highest and
lowest ranking sequences (P = 0.01), and the variation in RF duty cycle among the sequences was
also large, as can be seen in Figure 1.
We also found within-sequence variation. In one sequence field of view-adjustments caused a 6-fold
exposure increase in a gradient field exposure parameter, and flip angle adjustments caused a
6-fold exposure increase in an RF field exposure parameter.
With the exposure parameters presented here the RF- and gradient field exposure of an MRI
sequence can be characterized in a way that can hopefully be useful in exposure assessment of
future epidemiological studies on long-term effects of MRI exposure.
There are differences in exposure between MRI sequences, as was illustrated by the results of the
small sample in this study, so all MRI exams are not the same from an exposure perspective.
Furthermore, within-sequence variation is possible, so a single sequence may result in different
amounts of exposure for two separate patients, as the sequence settings are adjusted to fit the
individual. This means that MRI exposure is not only sequence-specific but also specific to the
individual patient and exam occurrence, a complexity which must be kept in mind when performing
exposure assessment for epidemiological studies on MRI.
Figures

421
 Figure 1. Gradient field exposure parameter shown as dark bars on left y-axis: proportion of
total sequence time with gradient B-field above an arbitrary cut-off value of 8 mT, for five
different sequences. RF exposure parameter shown as light bars on right y-axis: proportion of
total sequence time with RF pulses on, for five different sequences. Sequence names given
on x-axis are GE-specific.

PA-39 [14:30]
Exposure assessment of commercially available Infrared (IR) Panels
Peter Gajšek1, Tomaž Trček1 & Blaž Valič1
1Institute of Nonionizing Radiation, Ljubljana, Slovenia, 1000
Keywords: Dosimetry (measurements), Optical, Completed (unpublished)
Presented by: Peter Gajšek
In this study, exposure assessment of different commercially available IR panels was done. We
have investigated exposure to power frequency magnetic field and far infrared radiation. The results
showed that the measured values were below the current ICNIRP guidelines. For the worst case
scenario, when standing closer than 0.5 m directly in front of the IR panel of 1000W and 1400 W,
respectively, and continuously staring at it for at least 1000 seconds at an outside temperature of 22
°C, the limit values for cornea and lens damage of 190 W/m2 according to ICNIRP guidelines were
exceeded.
Introduction
In the last few years, the infrared (IR) panels have expanded as a way of heating the spaces using
infrared light. Infrared heating is a fairly recent addition to the domestic and commercial heating
scene. IR radiation is emitted from the heater, it then travels unimpeded through the air until it hits
an object. The object absorbs the radiation and heat up. The IR panels are heated to operating
422
temperatures, which are usually lower than 100°C during operation. The radiation spectrum of a
body heated to 100°C reaches its highest value at a wavelength of 7.8µm in the IR-C radiation
range (Far Infrared). In addition, IR panels are also source of extremely low frequency magnetic
field. Due to compliance testing according to manufacturer's requirements we have measured power
frequency (50Hz) magnetic fields and far infrared radiation of several commercially available IR
panels.
Material and Method
In the laboratory, measurements of the power frequency magnetic fields and IR radiation of the
panels were measured in such a way that the IR panels were oriented vertically, and the
measurements were performed at different distances and at different angles. The important data to
consider when comparing the measured values in T 1 are the power of the panels and their power
to surface ratio.
T 1: Technical specification of the investigated IR panels.

Power Dimensions Surface Power/surface

Manufacturer Model
[W] [cm] [m2] [W/m2]

Panel 1 Sunlife 450 450 59,5×59,5 0,354 1271

Standard
Panel 2 Sunlife 700 59,5×119 0,708 989
700

Panel 3 Sunlife One 1000 1000 85 x 119 1,011 989

Panel 4 Ekosen 1400 1400 59,5×179 1,065 1314

Before performing the measurements, the IR panels were switched on for at least 30 minutes to
heat up to the final temperature. For each panel, the IR measurements were made at the centre of
the panel at a distance of 0.5 (1.), 1 (2.) and 2m (3.) from the panel, and at a distance of 0.5m from
the panel at an angle of 45 to right (4.), left (5.) and above (6.) the panel and at an angle of 90° to
the right (7.). One measurement was also performed for each panel at a distance of 0.5m (8.).
The measurements of artificial IR radiation were performed in the wider wavelength in range than
0.19 to 3µm according to EN 62471: 2008 [1]. Since the IR panels are radiating in the far IR-C
range, the International Commission on Non-Ionizing Radiation Protection (ICNIRP) clearly stated
that in such a situation the measurements should be performed also in beyond 3µm (the
measurement range was expanded up to 11µm). The IR radiation was measured with Coherent
FiledMax Top with probe PM10 (extended uncertainty ±28%).
The measurements of ELF magnetic fields were performed according to IEC 61786:2005 [2] at three
locations: near the power cable at the distance of 0.2m (1.), the highest value in front of the IR panel
at the distance of 0.2m (2.) and in front of the panel at the distance of 0.5m (3.). The ELF magnetic
fields were measured with W&G EFA-3 (extended uncertainty ±31%).
Results
In the vicinity of the panel at a distance of 0.5m, the power to surface ratio of the panel is particularly
important. In the vicinity of the panel, the smaller area is mostly contributing to the total irradiance,
therefore the size of the panel itself and the total power of the panel do not have such a big
influence. From the measurements at the measuring point 1 it can be seen that although the
measured values increase with the increase of the panel power, the increase (170, 183, 205 and
230W/m2) is significantly smaller than the increase of the power of the panel (450, 700, 1000 and

423
1400W). The values are related also to the power per surface ratio of the panel. At a distance of 2m
at the measurement point 3, the increase in the measured values (19, 22, 35 and 42W/m2) is
significantly more related to the power of the panel, which is also evident from Table T 2.
T 2: Measurement results for IR irradiation.

Measurement Irradiation Irradiation Irradiation Irradiation

distance
point [W/m2] [W/m2] [W/m2] [W/m2]

Panel 1 Panel 2 Panel 3 Panel 4

1 0,5m 170 183 205 230

2 1,0m 41 85 102 120

3 2,0m 19 22 35 42

4 0,5m 44 77 89 110

5 0,5m 41 71 68 100

6 0,5m 45 61 85 80

7 0,5m 12 5 2,5 10

8 0,5m 20 4 29 12

A maximum magnetic flux density value of 8.9μT was measured at the distance of 0.2m from the IR
panel No.4 (see T 3). The magnetic field exposures at the distance of 0.5 m from the panel are
below 0.4μT for all four panels.
T 3: Measurement results for magnetic fields of 50 Hz.

Measurement Magnetic flux Magnetic flux Magnetic flux Magnetic flux

point density B [µT] density B [µT] density B [µT] density B [µT]

Panel 1 Panel 2 Panel 3 Panel 4

1 0,15 0,23 0,52 0,60

2 2,00 4,00 7,80 8,90

3 0,16 0,30 0,38 0,28

Conclusion
In general, exposure assessment of the IR panels has clearly demonstrated that IR radiation in far
IR range and power frequency magnetic fields (50 Hz) did not exceed the current ICNIRP guidelines
424
[3, 4, 5] at the distances greater than 1m for all investigated panels.
For the worst case scenario, when standing closer than 0.5 m directly in front of the IR panel No.3
(1000 W) and No.4 (1400 W), respectively, and continuously staring at it for at least 1000 seconds
at an outside temperature of 22°C or higher, the limit values for cornea and lens damage of 190
W/m2 according to ICNIRP guidelines [3, 4] were exceeded. All other threshold limit values (skin) for
this worst-case exposure scenario were not exceeded.
At a distance greater than 0.5m from the panels, the limit values for IR radiation [3] and power
frequency magnetic fields [5] were not exceeded in any case.
Since research to determine if routine IR-C exposures that produce heat stress and some degree of
heat strain are without risk is ongoing, some recommendations for using the IR panels based on
reasonable precautions could be provided:
IR panels should be placed on the ceiling rather than on the walls.
IR panels should be positioned at least 1m away from the places where they could be viewed
directly and people stay longer. Avoid IR panels as mirrors.
Install two weaker IR panels instead of one stronger panel in one room. The two panels should
be away from each other as much as possible.
Painting IR panels with favourite images and motifs is not recommended, as the panel should be
placed as seamlessly as possible in the room, in order to reduce the possibility of direct viewing
in the radiation source.
Acknowledgement
This work was supported by Ekosen d.o.o., Ulica Eve Lovše 19, Maribor, Slovenia, info@ekosen.si
References
[1] EN 62471:2008 - Photobiological safety of lamps and lamp systems (IEC 62471:2006, modified).
[2] IEC 61786:1998 - Measurement of low-frequency magnetic and electric fields with regard to
exposure of human beings - Special requirements for instruments and guidance for measurements
[3] ICNIRP. ICNIRP Guidelines on Limits of Exposure to Incoherent Visible and Infrared Radiation.
Health Physics 105(1):74‐96; 2013
[4] ICNIRP. ICNIRP Statement on Far Infrared Radiation Exposure. Health Physics 91(6):630‐645;
2006
[5] ICNIRP, “Guidelines for Limiting Exposure to Time-Varying Electric, Magnetic, and
Electromagnetic Fields (Up to 300 GHz)”, 1998

PA-41 [14:30]
STUDENT PAPER
Measurement and health assessment of electromagnetic fields by electric vehicles during
road traffic operation using driving characteristics
Kai Jagielski1, Pia Schneeweiss1, Dominik Stunder1 & Thomas Kraus1
1Research Center for Bioelectromagnetic Interaction (femu), RWTH Aachen University, Aachen, Germany,
52074
Keywords: Dosimetry (measurements), Pulsed, Work in Progress
Presented by: Kai Jagielski
This contribution introduces a measurement system which enables magnetic field measurements in
the frequency range between 1 Hz and 400 kHz on electric vehicles during vehicle operation. The
additional recording of the driving style, with acceleration sensor, gyroscope and GPS module,
allows a detailed investigation of the influence of various vehicle components on the magnetic field
spectrum. First measurement results of electric cars and the evaluation possibilities are presented.

425
In future, measurements will be carried out for example at wireless charging systems or electric
buses.
Introduction
Electromobility is currently regarded as a possible solution for reducing air pollution in major cities
and the use of renewable energy sources in road transport. As such, electromobility is currently
experiencing a major upswing and is being promoted in a targeted manner. The electrification of
motor vehicles leads to an increasing occurrence of electromagnetic fields, e. g. the acceleration
currents of an electric vehicle of up to 500 A generate strong partly pulsed magnetic fields, which
affect both the driver and other passengers. In addition, there are innovative inductive charging
concepts that transmit power in the two-digit kW range via a magnetic field, usually at 85 kHz [1]. In
order to prevent health related effects, such as nerve stimulation or heating effects, the magnetic
field strength should not exceed exposure limits, for example, the reference levels of the
International Commission on Non-Ionizing Radiation Protection (ICNIRP) [2, 3]. Nevertheless,
testing the electromagnetic compatibility of newly-approved electric vehicles in accordance with the
current UN ECE Regulation No. 10 Rev. 5 [4] does not require a health-related check of the
magnetic fields in the passenger compartment. Even if the vehicle manufacturers have an interest in
compliance with limiting values, such as occupational safety [5], there is no legal obligation and no
independent verification is carried out. As a result, most measurements are not public and only
known to the manufacturers. Due to the large number of newly introduced vehicle models, current
scientific publications on the emission of magnetic fields by fully electrically powered vehicles are
still rare, while measurements at hybrid powered vehicles are out-of-date and hardly comparable to
newer models. The requirements for measuring magnetic fields on electric vehicles in all operating
conditions are extensive. A test stand, for example, must be able to transfer power to the vehicle
during deceleration in order to measure the influence of the recuperation brake. However, a
conventional car test bench cannot perform this. Therefore, measurements have already been
carried out during on-road operation in road traffic, although the reproducibility and comparison of
different vehicle types is difficult due to different driving styles. For these reasons, a measuring
system was developed which records not only the magnetic field exposure but also the
characteristics of the driving style which enables more detailed analyses.
Methods
Within the scope of this research project, a mobile measuring system is being set up which is
battery-powered and can record time signals in the frequency range from 1 Hz to 400 kHz.
Therefore, a calibrated Narda ELT 400 is used as a magnetic field meter, either with a 100 cm² or 3
cm² isotropic magnetic field probe. The analog output of the ELT 400 with a data rate of 1MHz for
each orthogonal axis is read out with a calibrated digital oscilloscope, Picoscope 5444B, and
transferred to a laptop. The recorded time signal can be analyzed live during the measurement and
evaluated for example with the weighted-peak method in the time domain as well as recorded for
further evaluation with the data of the driving style characteristics.
To record the driving style, accelerometer, gyroscope, GPS module and barometer of an Android
smartphone (Samsung Galaxy S7, SM-G930F) are used. The sensor chip (K6DS3TR), which
contains both the accelerometer and the gyroscope, has sufficient accuracy to record the course of
the route. This enables the evaluation of the vehicle acceleration, braking power or steering during
the measurement. This can be used to obtain further information on the use of vehicle components,
such as the use of the recuperation brake and the resulting impact on magnetic field measurement.
The GPS module is used to document the route and to compare it with map data. From the location
data of, for example, power pylons and transformers, it is possible to determine when magnetic field
measurements are significantly distorted by external influences. The joint evaluation of all collected
data will enable the best possible reproducibility of magnetic field measurements and thus contribute
to a better understanding of the fields encountered on electric vehicles.
Results and Prospects

426
In this contribution, the measuring system and method is presented. First measurement results of
electric cars and possibilities of the measuring system are presented and discussed at the poster. In
the future, measurements will be carried out on electrically powered buses, e-bikes and various
charging systems for electric vehicles, like wireless power transfer (WPT) systems or wired
charging. Applications of the measuring system in trams and trains are also possible.
Acknowledgements
This work is funded by the German Ministry of Innovation, Higher Education and Research of North
Rhine-Westphalia within the interdisciplinary research project NRW Forschungskolleg ACCESS!.
References
1. DIN - German Institute for Standardization, DIN IEC/TS 61980-2:2017-10 - Electric vehicle
wireless power transfer (WPT) systems - Part 2: Specific requirements for communication
between electric road vehicle (EV) and infrastructure with respect to wireless power transfer
(WPT) systems (IEC 69/480/CD:2017), 2017.
2. International Commission on Non-Ionizing Radiation Protection, Guidelines for limiting exposure
to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz), Health Phys,
1998, 74(4): pp. 494-522.
3. International Commission on Non-Ionizing Radiation Protection, Guidelines for limiting exposure
to time-varying electric and magnetic fields (1 Hz to 100 kHz), Health Phys, 2010, 99(6): pp.
818-36.
4. United Nations Economic Commission for Europe - UN ECE, Regulation No. 10 Rev. 5 - Uniform
provisions concerning the approval of vehicles with regard to electromagnetic compatibility,
2014.
5. European Parliament, Directive 2013/35/EU on the minimum health and safety requirements
regarding the exposure of workers to the risks arising from physical agents (electromagnetic
fields), 2013.

PA-43 [14:30]
Spatial variations of personal exposure to radiofrequency electromagnetic fields at the
University of Castilla-La Mancha (Spain)
Alberto Juárez-Castillo1 & Alberto Nájera1
1Medical Sciences Department, University of Castilla-La Mancha, Albacete, Spain, 02005
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Alberto Nájera
Summary: We measured personal exposure to RF-EMF in 4 different microenvironments of 7
buildings of the University of Castilla-La Mancha in Albacete (Spain). Exposure to WiFi 5G was
predominant in all the buildings. Mobile telephone signals represented a small percentage of the
total exposure.Regarding the correlation study, only the average exposure of some centres was
statistically significant, but in all cases less than 0.550. Exposure varies from each
microenvironment and building.
There are numerous studies that have described personal exposure to radio frequency
electromagnetic fields (RF-EMF) using personal exposimeters. These studies have characterized
exposure by microenvironments or even in entire cities.
In this work, we measured personal exposure to RF-EMF in 4 different microenvironments of 7
buildings of the University of Castilla-La Mancha in Albacete (Spain) (Figure 1) in order to evaluate
possible statistically significant differences. The buildings studied were the following: Faculty of
Medicine (FM), Benjamín Palencia (BP), Faculty of Pharmacy (FP), Melchor de Macanaz (MM),
Simón Abril (SA), Infante Don Juan Manuel (IDJM) and Manuel Alonso Peña (MAP).
For this purpose, a personal exposimeter EME Spy 140 was used to measure 90 minutes per
building (every 5 seconds), considering the mobile telephone frequency bands (DCS, GSM, UMTS

427
uplink and downlink) and WiFi wireless networks (2G and 5G). 4 different environments were
identified: classes, corridors, cafeteria and library. Statistical tests of normality were carried out for
the analysis of the data, as well as Mann-Whitney U test, Kruskal-Wallis and Spearman correlation.
The exposure to WiFi 5G was predominant in all the buildings (Figure 2). Mobile telephone signals
represented a small percentage of the total exposure. The highest average value (Table 1) of
exposure was recorded in the IDJM building with 21.202 µW/m2, with an average value in the class
of 56.551 µW/m2 that contrasts with the classroom average value of FP of 0.366 µW/m2. The
average values for the rest of buildings ranged between 2.771 µW/m2 of the MM building and 8,306
µW/m2 of FM. The contrast tests indicated statistically significant differences in most cases.
Regarding the correlation study, only the average exposure of some centres was statistically
significant, but in all cases less than 0.550.
In conclusion, the personal exposure recorded from mobile phone base stations and terminals and
WiFi bands in the buildings and microenvironments studied turned out to be generally low and very
variable between rooms and buildings. Personal exposure does not exceed the maximum limit
legally established in Castilla-La Mancha of 100,000 µW/m2.
Figures

Figure 1. Albacete (Spain) Campus map.

428
Figure 2. Frequency band contributions at each building.

Figure 3. Descriptive statistics per building and microenvironment.

429
PA-45 [14:30]
Exposure to radiofrequency electromagnetic fields (RF-EMF) assessment: comparison
between spot and personal exposimeter measurements in Taracón (Spain)
Alberto Nájera1, Jesus Alpuente-Hermosilla2, Jesus Gonzalez-Rubio1, Rocio Sanchez-Montero2 & Pablo
Luis2
1Medical Sciences, University of Castilla-La Mancha, Albacete, Spain, 02008
2Signal Theory and Communications, University of Alcalá de Henares, Alcalá de Henares, Spain, 28801
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Alberto Nájera
Summary: In this paper, we compare RF-EMF exposure levels determined by this two equiment: a
broadband isotropic probe (NARDA EMR-300) and personal exposimeter (Satimo EME Spy 140) in
38 different locations for 6 minutes covering the whole urban area of Tarancón (Spain).
Personal exposure to radiofrequency electromagnetic fields (RF-EMF) has substantially increased
as new wireless technologies are introduced. Society has become more concerned about the
possible effects of RF-EMF on human health in parallel to this increase in their exposure. Over the
last 15 years, several studies have characterized personal exposure to RF-EMF with personal
dosimeters or spot measurements.
Characterization of exposure with personal exposimeters (PEM) provides individualized and detailed
temporal and spatial information, but they are costly in terms of time and resources. On the other
hand, estimation of time-weighted averages (TWAs) based on broadband spot measurements could
simplify the work.
In this paper, we provide a comparison of the RF-EMF exposure levels determined by these two
approaches: using an electric field monitor with a broadband isotropic probe (NARDA EMR-300) in
the range from 100 kHz to 3 GHz and a PEM (Satimo EME Spy 140) in the range from 88 MHz to
5.8 GHz. We have measured in 38 different locations (Figure 1) for 6 minutes covering the whole
urban area of Tarancón (Spain) that has a population 14,834 (2017). We have used simultaneously
two PEM, placed on either side of hip of the volunteer, configured to measure every 5 seconds, so
mean value at each location was determined using 72 measurements for each exposimeterand one
isotropic probe adjusted for six minutes using the root-mean-square mode of the device and placed
in a non-metallic tripod at 1,5 m height. During the 6 minutes measurement, volunteer carrying the
PEMs was continuously moving around the static probe.
A statistical and spatial analysis of the measurements is presented. Broadband measurement mean
value was 0.438 V/m (IC95%: 0.328-0.548) and we determined 0.281 V/m for the PEM
measurements (IC95%: 0.213-0.350) (Figure 2). Broadband and PEM measurements showed a rho
Spearman correlation of 0.948** (p<0.01). PEMs underestimate measurements, so we estimated a
correction factor of 1.6. Taking this value into account, PEM exposure mean value should be 0.450
V/m (IC95%: 0.340-0.561). We also perform a U-Mann-Whitney test (spot and corrected PEM
measurements) that conclude no statistically significant differences (p>0.05). We also provide a
kriging spatial analysis comparing spot, PEM, and mean exposure value differences (Figure 3).
We conclude that exposure assessment based on spot measurements could be a feasible proxy to
rank personal RF exposure, either using a static probe or a moving around PEM. The methodology
developed provides a reliable and rapid characterization with which to monitor personal exposure
and detect possible changes over time.
Figures

430
Figure 1. 38 measurement locations.

Figure 2. Narda and Satimo Field (V/m).

431
 Figure 3. Kriging analysis. A) Narda. B) Satimo. C) Differences.

PA-47 [14:30]
A study on the verification of the exposure system and the SAR values of the exposed cells
in in vitro experiments
Jaesung Lee1, Hyungsang Ahn1, Junoh An1, Sangbong Jeon2 & Jeong-Ki Pack3
1Institute for Future Radio Engineering, Seoul, Korea, 03175
2Electronics and Telecommunications Research Institute, Daejeon, Korea, 34129
3Chungnam National University, Daejeon, Korea, 34134
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Jaesung Lee
This study was conducted to verify the SAR values of the exposed cells in an in vitro experiment. In
this study, the signal output power and the temperature change in the Petri Dish were measured,
and the exposure level of the electromagnetic wave was derived based on the measurement data.
The results show that power of 3.09W is estimated as the appropriate source input for 1W/kg of 1g-
averaged SAR and 24.72W for 8W/kg of 1g-averaged SAR.
Introduction
Monitoring and verification of the exposure level of an exposure system used in in vivo or in vitro
experiment is required in order to secure the reliability and objectivity of the results of biological and
human impact studies on electromagnetic waves. In this study, the exposed SAR values were
verified by measuring the temperature change according to the signal source power of cell exposure
system for WCDMA frequency.
Measurement methods
The modulated WCDMA signal (center frequency 1950MHz) is supplied to the exposing device by
an antenna via a controllable attenuator, an amplifier (maximum output 40W) and an RF coupler.
The exposure level and the exposure time schedule can be set up by a computer.
The measurement set up and the exposure system used for verification is shown in Figure 1. Rohde
& Schwarz’s ESU EMI Test Receiver (20 Hz-8 GHz), Aglient's power meter (E4417A), Aeroflex's
Fixed Coaxial Attenuators (model 53), Luxtron's Two Channel Fiber Optic Thermometer (Luxtron
812) temperature probe were used.

432
In order to monitor the exposure level of the cell exposure system, the output power of the signal
source was measured, and the SAR values were evaluated by measuring the temperature change
of the culture medium at five points in the Petri dish.
SAR value is related to the temperature change as follows:

- Cp: 3,841 [J/kg℃] as the static specific heat at constant pressure of the culture medium.
- dT/dt: temperature rise per unit time
Results
Figure 2 shows the measurement points in the Petri dish and Figure 3 shows the change in
temperature over time at a point in the Petri dish when exposed to 40 W of power input. The SAR
values were calculated using the results of the part in which the temperature rise is approximately
linear. Table 1 shows the evaluated SAR values and the SAR values at 1W input.
Therefore, to obtain the desired SAR values in the cell exposure system, it is estimated that the
output of the signal source need to be set using the SAR conversion table summarized in Table 2.
Table 1. Measurement data and SAR values per unit input for the cell exposure system in WCDMA
band

SAR per unit

Measurement input power
Input power[W] ΔT/Δt[℃/sec] SAR [W/kg]
location [W/kg/W]Calculated
at 40W

1 40 0.00344 13.213 0.330

2 40 0.00290 11.139 0.278

3 40 0.00342 13.136 0.328

4 40 0.00371 14.250 0.356

5 40 0.00338 12.983 0.325

Average 40 0.00337 12.944 0.324

Table 2. SAR Conversion Table
SAR [W/kg] Set power [W] SAR [W/kg] Set power [W]

1 3.09 5 15.45

2 6.17 6 18.54

3 9.27 7 21.63

4 12.36 8 24.72

Acknowledgement
This work was supported by Institute for Information & Communication Technology Promotion(IITP)
grant funded by the Ministry of Science and ICT(MSIT) (2017-0-00961, Study on the EMF Exposure
Control in Smart Society)

433
Figures

Figure 1. Mesurement set up and the exposure system

Figure 2. 5 Measurement points in the Petri dish (Dish diamete: 60mm)

434
 Figure 3. Temperature change at a point in the Petri dish over time

PA-49 [14:30]
The response of magnetotactic bacteria suspensions to radiofrequency in the UHF & SHF
ranges: A prospective study
Simona Miclaus1, Cristina Moisescu2, Ioan I. Ardelean2, Lucian Barbu-Tudoran3, Ioan Ardelean4, Lourdes
Farrugia5, Charles Sammut5 & Paul Bechet1
1Department of Technical Sciences, "Nicolae Balcescu" Land Forces Academy, Sibiu, Romania
2Department of Microbiology, Institute of Biology, Bucharest, Romania
3National Institute for R&D of Isotopic and Molecular Technologies, Cluj-Napoca, Romania
4Physics and Chemistry Department, Technical University of Cluj-Napoca, Cluj-Napoca, Romania
5Department of Physics, University of Malta, Msida, Malta
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Simona Miclaus
As biogenic magnetite nanoparticles encapsulated in a membrane forming magnetosomes inside
magnetotactic bacteria are excellent competitors of magnetic nanoparticles obtained by chemical
synthesis, their study over a wide frequency band in the upper part of radiofrequency spectrum may
provide knowledge for new technological applications. On the other hand, impact of power loss
production and deposition in magnetite crystals due to the magnetic component of the absorbed

435
field is not yet studied in case of various communication devices worn near the head (case of
tissues contain magnetite). In this regard a set of experiments were carried out to follow the
energetic response of magnetite at hundreds-thousands of MHz.
1. Introduction
Starting from 1996 biogenic magnetite present in human or animal tissues received specific
attention in connection to both thermal and non-thermal effects due to exposures at frequencies in
the UHF and SHF ranges. In 2003, looking for the mobile phone bioeffects via magnetite crystals
present in the brain, the mechanism of ferrimagnetic transduction was checked for the first time
using magnetotactic bacteria (MTB). In the last decades, chains of magnetosomes (MS) extracted
from MTB where shown to have a large number of applications in: microbiology, biophysics,
biochemistry, nanotechnology and biomedicine. In this regard a number of 35 applications are
mentioned in various references. Among them, magnetic field cancer therapy gained much
attention, mainly because it was proved that the biogenic magnetite from MS is much more efficient
than any other synthetic magnetic nanoparticles solutions used in such treatments.
Much of the research connected to MTB/MS was made in the frequency range (100-300) kHz, the
same frequencies as the ones used for cancer treatment by magnetic fluid hyperthermia. However,
ranges of frequency in which many of everyday life technologies emit electromagnetic field (EMF)
cover UHF and SHF ranges. Considering that biogenic magnetite in MS is very similar to the one
existent in the human brain, and that its role is still unknown, prospective studies following both
thermal and non-thermal effects are needed. In this regard, present work proposes the use of MS in
water-based solutions for checking their properties and response to microwaves (MW) stimuli.
2. Materials, methods and results
Magnetospirillum gryphiswaldense (strain DSM 6361, Schleifer et al., 1991; Schuler and Kohler,
1992) was used in this study. The cells were cultured under microaerobic conditions as described
previously (Moisescu et al., 2008; 2011) and harvested when they reached the exponential growth
phase. The cells were fixed in 1% formalin solution and subjected to a series of investigations
aiming at observing the energetic effects of MW exposure.
A. TEM images of magnetosomes
A field-emission-gun scanning transmission electron microscope (STEM), 80-200 kV, with
secondary electron imaging capability model Hitachi HD2700 was used for obtaining images of the
MS chains and of the bacteria (Fig. 1). Intact bacterial cellular bodies with magnetite chains of cube-
octahedral nanocrystals were identified in the prepared suspensions and the dimensional histogram
revealed that MS were in the range of (11-60) nm, with almost 72% in the range of (31-50) nm.
B. Dielectric and magnetic properties of MTB suspensions
Complex permittivity measurements of MS solutions were carried out over the frequency band (500
MHz-50 GHz). Two techniques were applied: the open-ended reflection technique and the
rectangular waveguide method. Since the second procedure enables both transmission and
reflection measurements, the complex magnetic permeability has been measured for the time being
in the frequency band (8.2 GHz – 12.4 GHz). These data will enable further dosimetric computations
for both components of the EMF.
C. RF field exposure of suspensions at 440MHz: thermal impact assessment
MS solutions were exposed in the near field of a monopole antenna belonging to a portable
transceiver fed at 5 W (usually used in front of the face/brain). Glass containers with MS solutions
were irradiated for 350 sec. at a distance D=5 cm from the transceiver. Incident field components in
air were measured by miniature field probes connected to a spectrum analyzer. The average rms
strengths were Erms=200 V/m and Hrms= 0.21 A/m. Excitation signal was a digital voice at
f=440MHz. The slopes of the heating curves of the solutions were extracted from temperature
recordings during exposure (the samples were thermally insulated), by using a fluoroptic probe
(Luxtron One with TrueTemp3 software). The temperature increase due to RF field exposure was in

436
the sensitivity range of the probe. Shielding of the E-field component is practically cancelling even
minute temperature increase, showing the H-field component do not heat up the samples (Fig. 2).
D. Non-thermal effect: MTB/MS suspensions inspected by photometry pre- and post-
exposure
Spectrophotometric investigations were carried out on MTB solutions prior to and after a 15 min.
session of MW exposure in the conditions mentioned at paragraph C. An Able JASCO V-530
spectrophotometer was used for transmittance (T %) determination in the range (200-900) nm. The
reference cuvette contained the suspension liquid and the sample cuvette contained the MTB/MS
suspensions (successively). Near infrared spectrum of exposed versus not-exposed sample showed
a negative leap of transmittance of max. (-1.3 %) in the range (728-740) nm, and a progressive
increase of transmittance in the range (740 – 900) nm for the exposed versus not-exposed sample
(Fig. 3). At 900 nm the relative increase of T% was of 3.2%. At the end of the exposure the sample’s
temperature was elevated with 0.5 °C due to thermal radiation of the transceiver itself (non-isotropic
thermal expansion of suspensions should not be excluded for the observed effect).
monopole antenna belonging to a portable transceiver fed at 5 W (usually used in front of the
face/brain). Glass containers with MS solutions were irradiated for 350 sec. at a distance D=5 cm
from the transceiver. Incident field components in air were measured by miniature field probes
connected to a spectrum analyzer. The average rms strengths were Erms=200 V/m and Hrms= 0.21
A/m. Excitation signal was a digital voice at f=440MHz. The slopes of the heating curves of the
solutions were extracted from temperature recordings during exposure (the samples were thermally
insulated), by using a fluoroptic probe (Luxtron One with TrueTemp3 software). The temperature
increase due to RF field exposure was in the sensitivity range of the probe. Shielding of the E-field
component is practically cancelling even minute temperature increase, showing the H-field
component do not heat up the samples (Fig. 2).
E. NMR relaxometry of MTB suspensions
Nuclear Magnetic Resonance (NMR) relaxation measurements using a low-field instrument (Bruker
MINISPEC MQ20) were performed on samples prepared with two concentrations of MTB. A
progressive decrease of the transverse relaxation time with the increase of MTB concentration was
observed. Compared with the host-solution, the MTB suspensions presented two orders of
magnitude lower relaxation time. These results demonstrate the potential use of MTB as effective
contrast agents in magnetic resonance imaging.
3. Conclusion
Energetic response of biogenic magnetite produced by magnetotactic bacteria is investigated by
different methods in order to classify and to understand the thermal / non-thermal effects that may
appear in ferrimagnetic nanocrystals suspensions when they are exposed to the very high frequency
range of the radiofrequency spectrum.
Acknowledgement: This research work was partially supported by COST Action BM1309 (COST
EMF-MED).
References
1. Moisescu, C., Bonneville, S., Staniland, S., Ardelean, I., and Benning, L. G. (2011). Iron uptake
kinetics and magnetosome formation by Magnetospirillum gryphiswaldense as a function of pH,
temperature and dissolved iron availability. Geomicrobiol. J. 28, 590–600. doi:
10.1080/01490451.2011. 594146
2. Moisescu, C., Bonneville, S., Tobler, D., Ardelean, I., and Benning, L. G. (2008). Controlled
biomineralization of magnetite (Fe3O4) by Magnetospirillum gryphiswaldense. Mineral. Mag. 72,
333–336. doi: 10.1180/minmag.2008.072.1.333
3. Schleifer KH, Schuler D, Spring S, Weizenegger M, Amann R, Ludwig W, Kohler M. (1991). The
genus ¨ Magnetospirillum gen. nov. Description of Magnetospirillum gryphiswaldense sp. nov.
and transfer of Aquaspirillum magnetotacticum to Magnetospirillum magnetotacticum comb. nov.
Syst Appl Microbiol 14:379–385

437
 4. Schuler D, Kohler M. (1992). The isolation of a new magnetic spirillum. Zentralbl ¨ Mikrobiol
147:150–151.
Figures

Figure 1. STEM image of MTB cells containing chains of MS (left) and an enlarged image of a
MS chain of magnetite nanocrystals (right).

Figure 2. Heating curve slope of a MTB suspension during field exposure at 440 MHz: when
both E and H field components are present (left); when only H field component is present
(right).

438
 Figure 3. Transmittance of the MTB suspension before and after exposure in the UV-VIS-NIR
ranges.

PA-51 [14:30]
Record of non-ionizing radiation levels, by a pocket portable dosimetric personal device for
epidemiological studies
Marco Rivera-Gonzalez1, Nazario Felix-Gonzalez1 & Ceferino Maestu1
1Laboratorio de Bio-electromagnetismo, Unviversidad Politécnica de Madrid, Madrid, Spain, 28223
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Ceferino Maestu
In recent years, increase in electromagnetic,EM, radiation generated by artificial means is in the
focus of attention as possible promoter for biological alterations. The increase of people affected by
this EM fields gives rise to the need for a personal dosimetry system, low cost and accessible to the
entire population with homologated measures; so, the population is aware of the radiation dose
received daily by means of easy access to their data, with a historical record of the measures taken;
where, the registered data of the entire population are stored in a server; this allows to create a
database big enough for the development of epidemiological studies that can be considered
relevant and to generate radioelectric maps in surface.
Introduction
The increase of the radiation levels generated by different base stations emitting EMF, jointly with
the variability of its intensity, and lack of knowledge of its characteristics, such as the exact location
of these EMF base stations; have generate different reports of people affected by pathologies
whose origins are unknown by medicine, such as Electromagnetic Hypersensitivity. The vast
majority of these patients are referred to medical specialists in the field of rheumatology, neurology,
even psychology, because the symptoms that they present are very similar to traditional patients,
but they are not get better of their symptoms with medical treatments. Therefore, they have founded
associations and groups of people affected by Electromagnetic Hypersensitivity, who have this type
of pathology and demand a solution to their chronic problem.
Concerns about the possible effects that these fields have on human beings are growing between
different population sectors. Electromagnetic radiations in the radiofrequency range are considered
non-ionizing and those are on the portion of the electromagnetic spectrum whose energy is not
capable of breaking molecular covalent bonds, even at high intensities. However, these radiations

439
can penetrate living tissues and yield with enough energy to produce thermal effects (heating) and
non-thermal effects like resonance and ion reorientation mechanisms.
The absorption of the energy of EMF is measured according to international radiological protection
regulations, with the Specific Absorption Rate (SAR) that is the number of watts per kilogram of
exposed tissue (W / Kg); and they consider that values above 4 W / kg on average for the entire
body, can cause disease; these values can be found at several meters away from a EMF emitter
system; and it is important to consider that a suitable measure of SAR is obtained adding all sources
of emission of electromagnetic fields.
The biological effects produced by exposure to Electromagnetic Fields, contemplate behavior
changes at the cellular level, producing variations of biochemical reactions of nerve cells; in the
same way it can generate alterations in biological rhythms and even reach harmful levels such as
the promotion of cancer and tumors.
These biological effects have been published in several journals; however the lack of large-scale
epidemiological studies makes the results inconclusive. Therefore, for the design and development
of more precise epidemiological studies, an individual evaluation of exposure to electromagnetic
radiation by RF is necessary. The exposure measurement should be effective, separating the
exposures of the possible RF sources and providing information of the length of time to which the
subject is exposed at all times.
The need to control radioelectric emissions has led to the creation of specific regulations linked to
the exposure of EMF. Specifically, the recommendation of the European Union, articulated as a Real
Decreto y Orden Ministerial, in Spain (Art. 8 7d of Royal Decree 1066/2001), and of scientific
committees with radiation levels requires that emissions be kept below predetermined thresholds.
The increase of people affected by electromagnetic radiation gives rise to the need for a personal
dosimetry system, low cost and accessible to the entire population with homologated measures; so,
the population is aware of the radiation dose received daily by means of easy access to their data,
with a historical record of the measures taken; where, the registered data of the entire population
are stored in a server; this allows to create a database big enough for the preparation of
epidemiological studies that can be considered relevant. Currently, there is no database that allows
researchers to know the levels of electromagnetic radiation which the general population is subject.
Serious epidemiological studies are needed to establish the causality of certain diseases associated
with artificially produced EMF radiation.
Keywords
Electromagnetic fields (EMF), Dosimetry, Radiofrequency, Epidemiological study, Biological
interactions with EMF fields
Objetive
Design and development of a portable and pocket-sized device to register in a personalized and
continuous way levels of non-ionizing radiation corresponding to the band between 50 and 6000
MHz. This device will allow the elaboration of an extensive database for the preparation of
epidemiological studies related to radiation generated by the antennas and base stations of
telecommunications.
Theory and methods
To cover the range of frequency bands between 50 and 6000 MHz, the dosimeter has several
antennas that are connected to a non-reflected frequency switch, controlled by the processor of the
device.
The device has a PLL signal synthesizer that allows a sweep in frequency between 50 to 6800 MHz,
and is controlled by the processor. The signal generated by the PLL enters to a mixer together with
the signals registered by the antennas. The mixer is used to reduce the frequency of the signal
captured to a fixed frequency of 315 MHz, where it corresponds to the following equations:

440
The signal registered by the antennas (VA), is a signal with amplitude (AA), frequency (wA) and
phase (θA), this signal is multiplied in the mixer with the signal generated by the PLL (VP) with
amplitude (AP), frequency (wP) and phase (θP). The final result (mixer output = VA VP), is a signal
with a double frequency component: the first one is the sum, while the second one is the difference
between both frequencies. The signal from equation 3 is amplified by a K value due to the LNA and
then pass through a band pass filter, where the rise component is filtered, resulting (wA- wP) = 315
MHz. Therefore equation 3 becomes:

The output signal of the mixer enters into the detector, where the signal is rectified and detects the
maximum amplitude value. Therefore, the output signal of the detector corresponds to equation 5.

The output signal of the detector gives the amplitude of the signal, which is related to the power of
the signal, which is the value of interest.

According to equation 7, the power of the signal varies according to the amplitude of the signal
captured by the antenna (AP); and the signal is recorded in the processor, and it is stored in non-
volatile memory, and then processed by the computer. Figure 1 shows the block diagram of the
prototype.
The signal generated by the PLL corresponds to the value of the X axis (frequency); while the signal
captured by the detector corresponds to the Y axis (Power). The device includes a geolocation
system (GPS) that allows registering the place where the device is located in each of the records
made and is stored in non-volatile memory.
The device has USB communication to communicate with the computer, where can download the
records made by the dosimeter. The software processes the information and displays the records at
different location points on a map; which can access to the power graph by a click; In addition, the
software sends the entire registry over the network to a database that allows epidemiological studies
of the population and monitoring of each user; Only the administrators have access to this database.
Results
Figure 2 A) shows a frequency sweep from 50 to 6000 MHz, take with a spectrum analyzer,
configured with maximum retention. In Figure 2 B), C) the response of the mixer is observed,
locating at 315 MHz with a known signal, by a signal generator.
In Figure 2, the mathematical theory is corroborated with the design of the prototype, generating a
known signal to be measured at 2400 MHz frequency and -25.1 dBm amplitude, as shown in Figure

441
2B) and an output signal is obtained from the 315 MHz mixer and with an amplitude of -50 dBm, as
shown in Figure 2 C).
According to the software in figure 3, the geographic location of the sample is shown on the map,
where a graph with the powers of electromagnetic radiation is displayed.
Discussions and Conclusions
As seen in figure 1, a procedure for capturing electromagnetic radiation emissions corresponding to
the exposed formulation has been developed. In Figure 2, the results obtained with this prototype
are shown; therefore the principle of system operation is validated. The captured power varies
according to the amplitude of the signal captured by the antennas.
The software allows managing the data acquired by the prototype, storing and visualizing the
records captured during the day. Additionally, this set of data is sent to a database that allows
creating a database large enough to carry out exhaustive epidemiological studies. This data also
allows the development of a EMF radio map of the city, allowing to visualize the moments in which
there is a higher amount of radiation generated by different sources of EMC.
The prototype allows continuous monitoring 24 hours seven days a week, so it includes an internal
lithium polymer battery. Furthermore, its small size is closely to credit card dimensions. This device
is ideal to work as a personal dosimeter for non-ionizing radiation; however, it is necessary to add a
system for measuring low frequency signals, generated by the electric network (50/60 Hz), to
contemplate the most relevant frequencies in the studies.
References
Fields at Work, 2015. ExpoM-RF UserManual (Version 2.1)
(http://www.fieldsatwork.ch/uploads/Downloads/Expom_RF_Manual-EN_v2.1.pdf).
Parliamentary Assembly. Doc. 12608. 06 May 2011, The potential dangers of electromagnetic fields
and their effect on the environment.
Gryz, K., Zradzinski, P., Karpowicz, J., 2015. The role of the location of personal exposimeters on
the human body in their use for assessing exposure to the electromagnetic field in the
radiofrequency range 98–2450 MHz and compliance analysis: evaluation by virtual measurements.
Biomed. Res. Int. 2015 (272460). http://dx.doi.org/10.1155/2015/272460.
Neubauer G, Röösli M, Feychting M, Hamnerius Y, Kheifets L, Kuster N, Ruiz I, Schüz J,
Überbacher R, Wiart J (2006), Study on the feasibility of epidemiological studies on health effects of
mobile telephone base stations—final report. http://www.mobile-research.ethz.ch
/var/pub_neubauer_pref14.pdf
Thuróczy G, Jánossy G, Nagy N (2000) Radiofrequency RF exposure of mobile communications in
Hungary and evaluation relevant to EU and national standard: base stations and handy devices. In:
Klauenberg BJ, Miklavcic D (eds) Radiofrequency radiation dosimetry and its relationship to the
biological effects of electromagnetic fields, NATO-ARW Series. Kluwer, Norwell, pp 531–539
Henderson SI, Bangay MJ (2006) Survey of RF exposure levels from mobile telephone base
stations in Australia. Bioelectromagnetics 27:73–76
Thuróczy G, Molnár M, Szabó J, Jánossy G, Nagy N, Kubinyi G, Bakos J (2006) Public exposure of
RF from installed sources: site measurements and personal dosimetry, Proceedings of the first
European Conference on Antennas and Propagation (EuCAP 2006), Nice, France, 6–10 November

Figures

442
Figure 1. Prototype: A) Electronic diagram; B) Software diagram; C) Electronic PCB

Figure 2. Frequency Response: A) PLL Swept from 50 to 6000 MHz; B) Input signal to be
measured 2400 MHz; C) Output signal at 315 MHz.

443
 Figure 3. Power graph related to geographical position

PA-53 [14:30]
EMF measurement probes are prone to errors for modulated signals
Antonio Sarolic1 & Pawel Bienkowski2
1Chair of Applied Electromagnetics - EMLab, FESB, University of Split, Split, Croatia, HR-21000
2Technical University of Wroclaw, Wroclaw, Poland, 50-370
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Antonio Sarolic
Electromagnetic field (EMF) measurement probes are commonly used for performing health hazard
assessments. EMF waveforms are rarely purely sinusoidal, leading to measurement errors of diode-
based EMF probes. The existence of such errors is known to researchers, but their accurate
magnitude with respect to the signal waveform is less known. This study shows that the error
magnitude depends on the individual probe, and can be surprisingly significant.
EMF probes are commonly based on diode-loaded dipoles or monopoles. Considering that diode
I-U characteristic changes its behavior for varying signal strengths, the resulting device is very
sensitive to the amplitude of the input signal, i.e. the field strength. The amplitude dynamics is
declared by the manufacturers, and commonly covers two orders of magnitude (100 – 102 V/m). The
field strength in RF range is expressed and compared to guidelines as RMS value. However, the

444
probe characteristics for modulated signals will vary considerably throughout this range, generating
field measurement error. The measurement result will differ from the actual RMS value. This has
been known and partly analyzed previously, and is also being periodically revisited [1]-[5], as the
actual error magnitude still has not been analyzed quantitatively for all common signal waveforms.
The error can be expressed in dB as Error[dB] = 20 log (Emeasured/Erms), where Emeasured is the
measurement results as displayed by the probe, and Erms is the actual RMS value of the incident
field.
Figure 1 shows an example of two different commercial probes, both of them having similar
construction. The error was measured for the pulse-modulated RF signal having duty cycle of 1/8,
RMS electric field of 50 V/m (in the middle of their dynamic range), carrier frequencies well within
the probes frequency ranges (far from their cut-off frequencies), varying the pulse duration as shown
in figure. The error character depends on the pulse duration, and its magnitude is surprisingly
significant for certain combinations of waveform parameters. It is worth noting that none of these
probes was defective.
The magnitude of measurement error is generally unknown to the measurer, thus invalidating
hazard assessments based on such measurements.
Acknowledgements:
This study was performed within the research projects "Measurements in EMC and EM health
effects research" and "Measurements in Bioelectromagnetics (M-BEM)" supported by Ministry of
Science, Education and Sports of the Republic of Croatia, and within the framework of COST Action
BM1309 (COST EMF-MED).
References:
[1] A. Sarolić, B. Modlic, "Measurement of electric field probe response to modulated signals using
waveguide setup”, IEEE Antennas and Wireless Propagation Letters, vol. 9, pp. 1041-1044, Oct
2010.
[2] A. Sarolić, V. Roje, B. Modlic, "Measurement of Electric Field Probe Error for Pulsed Signals",
2006 IEEE International Symposium on Electromagnetic Compatibility, Portland, 2006, vol.2,
pp.244–248
[3] P. Bienkowski, A. Sarolić, H. Trzaska, "Electromagnetic field meters checking for labor safety and
environment protection purposes", BioEM 2016, Ghent, 5-10 June 2016, pp. 662-663
[4] J. Nadakuduti, S. Kuhn, M. Fehr, M. Douglas, K. Poković and N. Kuster, "The Effect of Diode
Response of Electromagnetic Field Probes for the Measurements of Complex Signals," in IEEE
Transactions on Electromagnetic Compatibility, vol. 54, no. 6, pp. 1195-1204, Dec. 2012.
[5] I. Exposito, M. G. Sanchez and I. Cuinas, "Uncertainty in Field Level Measurements of LTE
Signals Associated with User Load," in IEEE Antennas and Wireless Propagation Letters, vol. PP
(2018), no. 99, pp. 1-1.
Figures

445
 Figure 1. Measurement errors for two different models of commercial EMF probes

PA-55 [14:30]
ELF measurements around powerlines for validation with operator data
Fabian Schneider1, Philipp Bachmann1, Marco Zahner1, 2 & Jürg Fröhlich1
1Fields at Work, Zürich, Switzerland, 8006
2Institute of Electromagnetic Fields (IEF), ETH Zurich, Zürich, Switzerland, 8092
Keywords: Dosimetry (measurements), ELF/LF, Work in Progress
Presented by: Fabian Schneider
Field strength measurements of specific configurations of powerlines at different maximum current
patterns are difficult to estimate in reality. Therefore, different measurement methods are proposed
and compared to an analytical field calculation based on actual grid data.
With the long term measurement results changes in the field strength due to load regulation can be
clearly recognized. The ground measurements show a good agreement between the simulation and
the measurements. Aerial measurements using a Quadro copter are a viable solution. However, due
to movement artifacts, position uncertainties and simulation inaccuracies a direct comparison of the
two is difficult to achieve.
Introduction
Powerlines generate magnetic and electric fields which need to comply to limiting values defined by
national regulations. One approach is to use simulations to estimate the resulting field strengths.
However, intricate simulation parameters, e.g. soil properties, line sag, powerline geometry and

446
conductor arrangement are difficult to incorporate and are therefore often not modelled in detail.
This leads to uncertainties regarding the simulation results. For validation and uncertainty
estimation, comparisons with actual measurements are required. This work presents different
measurement approaches in a real scenario consisting of a power line with two independent
3-phase systems(150kV and 220kV) and their comparison to basic field simulations.
Materials
All field strength measurements were conducted using a personal ELF exposure meter [1], a three-
axis magnetometer for frequencies from DC up to 100kHz. This portable and lightweight
measurement device can also be used in conjunction with a Quadro copter to measure field
strengths at or above the conductor height of powerlines.
The platform used for the aerial measurements in this work consists of a Quadro copter frame [2]
with a Pixhawk flight controller [3]. Due to the additional weight of the ExpoM-ELF a maximum flight
time of about 8 minutes was possible with one fully charged battery. The flight controller is equipped
with GPS, Compass, IMU (Gyro and Accelerometer) and a barometric pressure sensor. These are
used for flight stabilization and the compensation of position drifts. As this flight controller supports
autonomous flight, a predefined path can be defined using waypoints consisting of GPS positions
and altitudes.
Methods
The following three types of measurements were carried out:
Stationary measurement - to investigate the change in the magnetic field over time due to the load
regulation of the substations, a long-term stationary measurement was conducted.
Ground measurements - data acquisition along a ground path perpendicular to the powerline. The
ExpoM-ELF was carried at waist height. To spatially compare the measurements to the simulation
results the built in GPS module was used.
Aerial measurement - data acquisition along measurement paths using predefined waypoints,
programmed to the Quadro copter. The paths covered a 300 m by 140 m area besides the
powerline and between two adjacent masts for different altitudes (15 m, 25 m, 35 m and 45 m above
ground). The holding time at each waypoint was set to 12 seconds to ensure the acquisition of at
least two ExpoM-ELF samples. For each different plane altitude, a new flight was carried out.
Fig. 1 shows the collected measurement data and describes the situation and paths of the different
methods. The measured data sets have been compared with an analytical solution in Matlab based
on actual operation current data provided by the grid operator.
Results
The comparison of the long term stationary measurement results with the power line current data
over the course of two days reflect the changes in the field strength due to load regulation (fig. 2,
left). The right plot of Fig. 2 contains a comparison of the simulated field 1 m above ground and the
measurements performed on waist height on a path perpendicular to the powerline. Fig. 3 shows a
comparison of the simulated and measured field strength values of a cross section in the middle of
the measurement area (aerial measurement).
Conclusion
At a distance of 30 meters from the conductors, the powerline does not interfere with the sensors of
the Pixhawk flight controller. Aerial measurements by Quadro copter are therefore a viable solution
for measurement locations away from obstacles and civil persons. The frequency selective
measurement principle of the ExpoM-ELF allows to isolate the 50 Hz component of the power grid
and reduces measurement errors due to magnetic fields generated from motor noise, and
movement artifacts of the aerial measurement platform in the collected measurement data. The
sensitivity of the ExpoM-ELF was sufficient to measure relatively small field strengths of 20 nT 130
meters away from the field source.

447
The differences between the measurement and simulation data cannot be directly attributed to
uncertainties in the different approaches. The simulation does not consider compensation currents
between the substations flowing in the ground wire of the power line, line sag and possible phase
differences between the two systems. As the field strengths are relatively weak for measurement
points far away from the source, uncertainties caused by motor noise and movement artifacts have
a higher influence on the measurement values.
References
[1] Fields at Work GmbH. "ExpoM ‐ ELF exposure meter, " http://fieldsatwork.ch/products/expom-elf/
[2] DJI - Da-Jiāng Innovations Science and Technology Ltd. "DJI F450, " https://www.dji.com/flame-
wheel-arf/
[3] PixhawkProject. "Pixhawk PX4, " https://pixhawk.org/
Figures

Figure 1. Collected Measurement points of the ExpoM-ELF shown in Google earth: 1)

stationary measurement location, 2a) ground path perpendicular to powerline direction, 2b)
environmental magnetic field measurement path from bus stop to measurement area and 3)
aerial measurement; (height of the red blocks correspond to the measured fieldstrength).

448
Figure 2. Long term stationary measurement results vs. power line current data over the
course of two days (left) and comparison of the simulated field 1 m above ground and the
measurements of the ExpoM-ELF carried on waist height, on a path perpendicular to the
powerline (right).

Figure 3. Comparison of simulated and measured field strength values of a cross section in
the middle of the measurement area (aerial measurement).

449
PA-57 [14:30]
Versatile RF electromagnetic fields exposure meter
Marco Zahner1, 2, Fabian Schneider1 & Jürg Fröhlich1
1Fields at Work GmbH, Zurich, Switzerland, 8006
2Institute of Electromagnetic Fields, ETH Zurich, Zurich, Switzerland, 8092
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Marco Zahner
To cope with the variety and complexity of today’s wireless communication systems and standards,
flexible exposure assessment instrumentation is necessary. The state of the art approach of
measuring the average field strength in a predefined set of frequency bands does not provide the
granularity required to discern the contributions from different mobile standards and devices
operating in the same frequency band, and it can also miss important contributions when new
frequency bands are allocated or modified. A measurement approach allowing to solve these
problems has been tested and implemented in a handheld prototype device.
Introduction
The diversity of modern wireless telecommunication systems and technologies poses high demands
on RF-EMF Exposure assessment instrumentation. Due to the demand for higher data rates, the
number of frequency bands used for telecommunication is steadily increasing. In addition,
nowadays several mobile standards are used on the same carrier frequency to maintain
compatibility to older user equipment. Due to this evolution, the traditional approach of identifying a
source of RF exposure based on its carrier frequency is not very accurate any more. Furthermore,
the various mobile standards use a variety of multiple access schemes and modulation formats
which exhibit very different timing patterns and peak-to-average power ratios. In certain areas,
exposure to radar transmissions must also be considered. Radars again rely on completely different
signal shapes than telecommunication systems. Current personal exposure meters measure only
the average field strength and are therefore not able to reflect this variety of exposure situations. An
advanced versatile exposure measurement device was therefore designed to provide the required
level of detail and flexibility to cope with the complexity todays RF-EMF exposure.
Methods
The current prototype of the versatile exposure meter fits into a handheld enclosure (Fig. 1) and
covers a frequency range from 500 MHz to 10 GHz. The lower frequency end is only limited by the
performance of the built-in antenna. It can be extended by using a larger measurement antenna.
A superheterodyne based hardware architecture allows to measure the field strength on arbitrary
frequencies within its measurement range, overcoming the limitations of exposure meter devices
with fixed frequency allocation. The detection principle of the device is designed to provide all
potentially relevant information about the measured signal while being suitable for implementation
on the limited energy and processing power of a portable device. A schematic of the detection
principle is shown in Fig. 2. In the RF front-end (not shown) the received signal is amplified, pre-
filtered, and mixed to a fixed intermediate frequency (IF). The output of the IF Filter (80 MHz
bandwidth) is then fed into an analog logarithmic detector, yielding a (compressed) envelope of the
measured signal. The envelope is then sampled and digitized in the time domain. All further signal
processing is performed in software.
This detection approach achieves an instantaneous dynamic range of more than 65 dB without the
need for a high-resolution ADC. By taking time-domain samples of the exposure, many parameters
like peak power, crest factor, patterns and burst duration etc. can be readily determined.
Furthermore, it has been shown that this approach allows to identify the used telecommunication
standard based on specific time-domain features [1].
Results
Fig. 3 shows the raw ADC data when measuring a WiFi signal with the device tuned to 5.3 GHz
450
(top), and a mobile base station (downlink) signal at 900 MHz (bottom). These recordings have been
taken in an office at ETH Zurich. These two measurements illustrate how different signal shapes of
various telecommunication standards can be. Whereas WiFi typically operates with short bursts, the
signal shown in the lower picture exhibits the typical features of a GSM downlink transmission. The
regular pattern reflects the 577 microseconds wide GSM TDMA time slots including a short guard
interval. Based on a series of similar measurements, a set of methods and classification rules has
been defined that allow to identify the mobile standard of an exposure meter measurement. The
developed algorithm has been successfully implemented on a Cortex-M4 based MCU.
Conclusion and Discussion
To cope with the variety and complexity of today’s wireless communication systems and standards,
flexible exposure assessment instrumentation is necessary. The state of the art approach of
measuring the average field strength in a predefined set of frequency bands [2,3] does not provide
the granularity required to discern the contributions from different mobile standards and devices
operating in the same frequency band, and it can miss important contributions if new frequency
bands are allocated or modified. The measurement approach presented in this work allows to solve
these problems. It has been successfully tested and implemented in a handheld prototype device.
References
[1] Yves Jundt, “Wireless Standard Classification on Low-Complexity Hardware”, Master Thesis,
ETH Zurich, 2017
[2] MVG – The Microwave Vision Group, Paris, France, "EME Spy 200", www.mvg-world.com
[3] Fields at Work GmbH, Zurich, Switzerland, "ExpoM ‐ RF exposure meter", www.fieldsatwork.ch
Figures

Figure 1. Prototype of the versatile RF exposure meter. The handheld device features an
integrated wideband isotropic antenna, battery, GPS, and WiFi connectivity.

451
Figure 2. Detection principle of the device. After down conversion to a fixed intermediate
frequency (IF) the envelope of the received signal is extracted by means of a logarithmic
detector and sampled in the time domain. All further processing is performed in software.

Figure 3. Top: measured envelope of a WiFi signal at 5.3 GHz; bottom: measured signal
envelope of a mobile phone base station at 900MHz. The repetitive TDMA pattern of a GSM
signal can be easily recognized.

452
PA-59 [14:30]
Multi-modal measurements of occupational exposure to extremely low frequency fields in a
110 kV substation
Lei Zhang1, 2, Peng Gao1, 2, Feizhou Zheng1, 2 & Zheng-Ping Yu1, 2
1Department of Occupational Health, Army Medical University (Third Military Medical University), Chongqing,
China, 400038
2Key Laboratory of Medical Protection for Electromagnetic Radiation, Ministry of Education, Army Medical
University (Third Military Medical University), Chong, China, 400038
Keywords: Dosimetry (measurements), ELF/LF, Completed (unpublished)
Presented by: Lei Zhang
This study described the characteristics of ELF-EMF in a typical 110 kV substation in China using a
multi-mode measurement protocol. The geometric means of electric field (E field) strengths in
different working areas ranged from 0.42 V/m to 2458.31 V/m. The geometric means of magnetic
field (H field) flux densities ranged from 0.12 μT to 16.29 μT. This study provides a comprehensive
description of occupational exposures to ELF-EMF in a typical Chinese 110 kV substation. In
addition, the multi-mode measurement protocol provides a scenario of EMF distribution both in time
and space-dependent manner.
A transformer substation is a workplace where workers may be exposed to high-level extremely low
frequency electromagnetic fields (ELF-EMF). However, occupational exposure to ELF-EMF in
substations in China is poorly documented. This study aims to comprehensively describe the
characteristics of ELF-EMF in a typical 110 kV substation in China using a multi-mode measurement
protocol. The geometric means of electric field (E field) strengths in different working areas ranged
from 0.42 V/m to 2458.31 V/m, with a maximal value of 4717 V/m. The geometric means of
magnetic field (H field) flux densities ranged from 0.12 μT to 16.29 μT, with a maximal value of 98.54
μT. We created contour and 3D (three dimensional) surface maps of ELF-EMF distribution in the
substation. Long-term field measurements taken beside the main transformer showed a variation in
the electric and magnetic fields over time. Safety distance mode measurement showed that
magnetic field decreased to below 0.4 μT at 30 m out of the substation. This study provides a
comprehensive description of occupational exposures to ELF-EMF in a typical Chinese 110 kV
substation. In addition, the multi-mode measurement protocol provides a scenario of EMF
distribution both in time and space-dependent manner.

PA-61 [14:30]
Novel and versatile instrumentation for electro-manipulation of cancer stem cells
Ilan W. Davis1, 2, Caterina Merla3, Alessandro Zambotti4, Arianna Casciati3, Mirella Tanori3, Jonh Bishop2,
Cristiano Palego1, Mariateresa Mancuso3 & Christofer P. Hancock1, 2
1School of Electronic Engineering, University of Bangor, Bangor, United Kingdom, LL57 1UT
2Creo Medical, Bath, United Kingdom, BA2 3DZ
3ENEA, Division of Health Protection Technologies, Rome, Italy, 00123
4ENEA, Division Resource Efficiency, Rome, Italy, 00123
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Caterina Merla
This paper presents novel instrumentation combined with an artificial 50 Ω buffer for possible fast
neutralization of cancer stem cells. The results from an initial bench study investigates the
performances of a developed electroporation generator capable of delivering non-thermal
treatments in combination with an original cuvette housing unit containing suitable solutions of
cancer stem cells. Initial investigation for electropermeabilization threshold of cancer stem cells was
conducted, indicating that the developed devices and protocols have a strong potential in achieving
electro-manipulation of this biological target.
453
Introduction
As part of the European Union’s Horizon 2020 research and innovation program, the SUMCASTEC
project explores a new approach for real time isolation and neutralization of Cancer Stem Cells
(CSCs), as Glioblastoma Multiforme (GMB) and Medulloblastoma (MB) via a novel micro-optofluidic
lab-on-chip platform [1]. An element of the deliverables is to develop off-chip electric pulse
generators in the rage of nanosecond and sub-microsecond, and off-chip applicators to target
CSCs. One of the goals is also attempting to understand the effect of applied fields for CSCs
neutralization [1].
GBM and MB are primary brain tumors, occurring in adults and children respectively. Brain cancer
results in more death per person than any other cancer with very poor outcomes; indeed, the 5-year
survival rate is around 4% [2]. The recurrence and relapse of these tumors seem related to the
presence of quiescent cancer stem cells that evade common therapeutic approaches (i.e.
chemotherapy and radiotherapy).
Cell electromanipulsation using very short (down to few hundreds of picosecond) and intense
electric pulses (up to tens of MV/m) named also electropulsation is a relatively novel physical
technique. Pulsed electric fields alter a cell’s transmembrane potential. This results in permeabilizing
the cell’s plasma membrane and disturbing intercellular homeostasis. The resultant permeabilization
of cell plasma membrane can be reversible or irreversible leading in this case to cell death [3]-[7].
Irreversible electroporation is used for cancer ablation, and has been successfully performed
intraoperatively, laparoscopically and percutaneously [3]-[7].
The electric field based neutralization of CSCs could be mediated by cell electropermeabilization,
but CSCs differentiation/modification in the absence of electropermeabilization seems another
interesting possibility. Usually, the electric pulse duration and repetition frequency are low enough to
ensure that the energy delivery into the biological system is non-thermal.
Non-thermal approach seems a preferred method for neutralizing CSCs. This method allows
targeted neutralization without heating the CSCs above body temperature. The non-thermal
approach predominantly spares normal tissues allowing tissue regrowth and preventing unwanted
damage.
In this abstract we present the combination of generators and applicators and their matching to
performed off-chip experiments on CSCs.
Instrumentation
One of SUMCSTEC’s technical milestones was to build off-chip generators in the sub-microsecond
and nanosecond electric pulse regimes. One of this system has to be capable of delivery pulse
amplitude in excess of 1 kV, with minimal pulse widths of 100 ns. Its modular design is based on a
push-pull switching of high voltage, fast switching MOSFETs that are directly driven by opto-isolators
with suitable switching times.
The ability of the modular design to generate the required pulses is dependent upon the switching
times and maximum drain-source voltage of the MOSFET. The MOSFET gate driver must also
provide enough current to charge the gate-source and gate-drain capacitance. The performances of
the developed generator named “SUMCASTEC generator 1 SPG1”, presented in Fig. 1, are
compared with a built LTSpice model. We demonstrate that our prototype exceeds the LTSpice
simulation results across a purely 50 Ω resistor load (data not shown).
In addition to producing 1 kV, 100 ns pulses, SPG1 control and programing allow for a wide range of
pulse profiles to be generated spanning from 100 ns up to 1 µs and amplitudes in excess of 1 kV.
We also verified that pulse amplitude is unaffected throughout its operating repetition frequencies
between 1 to 50 Hz.
Exposing the CSCs (suitably characterized D283 cell line) to high voltage, 100 ns-300 ns pulses
were delivered via a cuvette housing unit designed by ENEA. This unit allows commercially
available electroporation cuvettes with the CSCs suspended in a specific 50 Ω buffer (phosphate
buffer saline and distilled water in a suitable dilution [9]), to be exposed to pulsed electric fields. The

454
housing unit allows an easy connection with the generator (standard N connector) with the
possibility of pulse monitoring on an oscilloscope during pulse delivery. An optimized transition
between the coaxial connector and the planar electrode of the EP cuvette was also realized and
simulated. Fig. 1 shows the connection of the exposure unit with the SPG1 and its main features.
An artificial 50 Ω buffer, of 0.3 S/m conductivity, to suspend the CSCs was initially used to optimize
impedance matching between the SPG1 and the CSCs load. The buffer consists of 20 mL of
phosphate saline buffer, 80 mL of distilled water and 8.2 g of sucrose [9]. The sucrose counteracts
the occurrence of osmosis. Prior to the CSCs exposure to pulses from SPG1, YOPRO-1 dye (3 μM)
was added to the CSCs solution. The permeabilization of cells membranes allows the dye to enter
the CSCs cytoplasm and bind with nucleic acids (DNA RNA), thus the permeabilized CSCs will
become fluorescent. YOPRO-1 emission at 510 nm has been detected using florescence
microscopy. The percentage of cell population that becomes florescent indicates the success of the
CSCs electropermeabilization. The complete protocol for CSCs preparation can be followed here
[8].
Results
The previously mentioned artificial buffer was prepared to represent a 50 Ω load to aid impedance
matching between the load and the generator. We preventively verified, performing trypan blue
exclusion test, that CSCs remain alive in the buffer after more than three hours. Fig. 2 demonstrates
that the buffer load is 50 Ω, as the waveform measured with the buffer is comparable to the
waveform measured with a 50 Ω resistor. Fig. 2 depicts also the 100 ns, 200 ns and 300 ns pulse
waveforms that were measured across the electroporation cuvette containing CSCs suspended with
the 50 Ω, 0.3 S/m solution. These waveforms, shown in Fig. 2, illustrate the shape of pulses
delivered to the CSCs during the primary permeabilization rate investigation. Our preliminary results
indicate that exposing CSCs to 200 ns and 300 ns with 5 or more consecutive pulses results in
>70% of the CSCs becoming permeabilized (green fluorescence in Fig. 3). Fig.3 indicates that 90%
of the CSCs population is fluorescent and permeabilized when 20 consecutive 1.2 kV pulses, of 300
ns duration are delivered though the cuvette.
A principle of electropermeabilization is its non-thermal effect on the living organism. Therefore, we
demonstrated computationally that the heating effect of the applied pulses on the CSCs buffer load
with various SPG1 parameters is negligible. Following initial investigation that exposed CSCs to
electric pulses, other buffer solutions of various conductivities where investigated (0.1 S/m, and 0.4
S/m) with pulse widths of 100 ns, 200 ns and 300 ns. Our results show that the SPG1 performance
is insensitive of the load impedance, in the range between 10 Ω (0.4 S/m) to 60 Ω (0.1 S/m),
demonstrating broadband matching performance (for pulses down to 100 ns) as far as the pulse
waveform widths are considered. This effect shows good transition times preservation overall.
However, the delivered pulse amplitude does change with the buffer impedance (data not shown).
Discussion and Conclusions
In conclusion, the overall system developed: the SPG1, the cuvette housing unit, and the buffer
resulted in successful permeabilization of the CSCs with the possibility of real-time cell
neutralization via non-thermal electropermeabilization. Promising results were obtained in terms of
matching strategy and cell permeabilization to YOPRO-1 dye with various pulse durations.
The gathered results do not reflect whether the CSCs are dead or alive after electropulsation.
Therefore, it is unknown whether they underwent reversible or irreversible electropermeabilization.
Additionally, it is important to state that this does not mean that the CSCs exposed to 100 ns pulses,
which resulted in a lower fluorescence, are not affected even if only poorly permeabilized. Viability
studies are required to complement these preliminary experiments in the future.
The generator developed is well matched to the application and has low sensitivity to various types
of solution filling the cuvette. The low sensitivity and range of programmable pulse regimes in the
SPG1 is advantageous to the user which performs well in comparison to more generic, commercial
sub-microsecond and nanosecond generators.

455
Acknowledgement
This project has received funding from the European Union’s Horizon 2020 Research and
Innovation programme (FET-Open) under grant agreement No. 737164 SUMCASTEC
References
[1] "Home - SUMCASTEC", SUMCASTEC, 2018. [Online]. Available: http://www.sumcastec.eu/.
[2] "Tumor Types - National Brain Tumor Society", National Brain Tumor Society, 2018. [Online].
Available: http://braintumor.org/brain-tumor-information/understanding-brain-tumors/tumor-types/.
[3] R. Martin, K. McFarland, S. Ellis and V. Velanovich, "Irreversible Electroporation in Locally
Advanced Pancreatic Cancer: Potential Improved Overall Survival", Annals of Surgical Oncology,
vol. 20, no. 3, pp. 443-449, 2012.
[4] R. Martin, “Use of irreversible electroporation in unresectable pancreaticCancer”, HepatoBiliary
Surgery and Nutrition. Vol. 4, no. 3, pp. 221-215, 2015
[5] S. Bagla and D. Papadouris, "Percutaneous Irreversible Electroporation of Surgically
Unresectable Pancreatic Cancer: A Case Report", J. of Vascular and Interventional Radiology, vol.
23, no. 1, pp. 142-145, 2012.
[6] N. Jourabchi, K. Beroukhim, B. Tafti, S. Kee and E. Lee, "Irreversible electroporation (NanoKnife)
in cancer treatment", Gastrointestinal Intervention, vol. 3, no. 1, pp. 8-18, 2014.
[7] R. Sundararajan, Electroporation-based therapies for cancer. Waltham, MA: Woodhead Pub,
2014.
[8] C. Merla, A. Casciati, M. Tanori, B. Tanno and M. Mancuso,
"SUMCASTEC_180123_NA_protocolWP3_protocol_.pdf_Rome_C. Merla_Partners and
public_NA", Zenodo, 2018. [Online]. Available:https://zenodo.org/record/1157784#.Wm9N3a5l-po.
Figures

Figure 1. Instrumentation setup: the SPG1 generator, the cuvette housing unit filled with a 50
ohm buffer and CSCs.

456
Figure 2. 100 ns, 50 Hz, 20 burst waveform, with 50 Ω resistor as SPG1 load (blue). And 100
ns (red), 200ns (green) and 300 ns (purple) pulse, 50 Hz, 20 bursts waveform measured with
CSCs in the 50 Ω buffer load.

Figure 3. Example of permeabilized cells at 300 ns, ~1.2 MV/m, 20 pulses at 1Hz. Bright field
(visible light) superimposed to fluorescence microscope images.

457
PA-63 [14:30]
A stripline system for the exposure of ex-vivo samples to nanosecond pulsed electric fields
Alessandra Paffi1, Alessandro Banin1, Agnese Denzi1, Maura Casciola2, Micaela Liberti1 & Francesca
Apollonio1
1Department of Information Engineering, Electronics and Telecommunications (DIET), Sapienza University of
Rome, Rome, Italy, 00184
2Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA, 23508
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Alessandra Paffi
We present an exposure system, based on a slotted stripline, to stimulate frog nerves with the use
of nanosecond and sub-nanosecond pulsed electric fields (nsPEFs). The system guarantees the
electromagnetic compatibility with laboratory equipment and exhibits a very good impedance
matching up to 3 GHz, thus allowing the undistorted propagation of nsPEFs. Through a hole (3 mm
of diameter) on the outer enclosure, the system permits a localized exposure of the nerve to a
maximum |E| field of 0.8 kV/m for 1 W of input power. This is compatible with conventional electrical
stimulation of nervous systems using much longer pulses; however experimental studies are
needed to identify a stimulation threshold for such short pulsed signals.
Introduction
In the past decades, conventional electrical stimulation (ES) has been used to elicit action potentials
(APs) in excitable cells and tissues through the application of electric pulses of few Volts (1-5 V) and
durations from hundreds of microseconds to few milliseconds [1, 2].
Invasive and non-invasive tools for the electromagnetic (EM) stimulation were developed (e.g. deep
brain stimulation, functional electrostimulation, transcutaneous electrical nerve stimulation,
transcranial magnetic stimulation) both showing encouraging results as well as a series of
drawbacks [3, 4]. The invasive approaches are often associated with side effects and are not
riskless for the patients, the non-invasive approaches may be characterized by limited penetration
depth (order of a few centimeters) and limited spatial resolution (3-5 cm3).
The need to develop a technology without these limitations has increased the interest towards
nanosecond and sub-nanosecond pulsed electric fields (nsPEFs) as an alternative to conventional
stimulation [5]-[8]. However, little is known about the feasibility of nanosecond and sub-nanosecond
pulses to elicit APs in cells or biological tissues. Moreover, a question arises on whether this
stimulation may be achieved without electroporative damage to ensure a safety window for medical
applications [5].
In vitro and ex-vivo biological experiments are needed to investigate such issues.
This requires design and fabrication of exposure systems suitable to deliver this kind of short stimuli
to the biological sample and in particular to ex-vivo nerve fibers. Requirements for these exposure
systems are: impedance matching up to a few GHz to guarantee the waveform preservation, good
efficiency in terms of the electric field intensity induced in the sample per unitary input power, and
low radiated power to avoid interference with the surrounding instrumentation.
In this work, a system, based on a slotted stripline, suitably designed to expose frog nerve, is
presented.
Materials and Methods
Moving from a conventional stripline, the system was dimensioned on the basis of analytical
formulas [9] in order to obtain a 50 W input impedance up to 3 GHz. Then the system was modified
by completely enclosing the dielectric substrate (er=3.55) with a copper conductor except for a hole,
3 mm of diameter, on the upper wall, to expose the nerve placed upon it. The system model is
shown in Fig. 1A, with the main sizes; all the sizes are reported in Table I. This system configuration
was chosen to guarantee the EM compatibility due to the negligible power fraction that scatters out

458
of the hole.
Table I. Physical dimensions of the structure

DESCRIPTION SIZE [mm]

Physical length 200

Physical width 40.14

Slot diameter 3

Substrate height 3.188

Central strip thickness 0.14

Central strip width 1.55

External conductor thickness 0.07

The system was fabricated using the RO4003C laminate, shaped through a microforge, and
connectorized by using SMA-connectors. The resulting structure is shown in Fig. 1B.
The system was experimentally characterized by measuring the scattering parameters by using a
VNA Agilent Technologies PNA Network Analyzer E8363C 10 MHz-40 GHz in the frequency range
30 MHz-3 GHz, in the presence and in the absence of the nerve (about 2 mm of diameter), placed
on the hole along the structure length.
Dosimetric evaluation of the electric field induced in the nerve (permittivity 47.3; conductivity 0.338
S/m) in correspondence of the center of the frequency range (1.5 GHz) was achieved by means of
simulations performed with the Ansoft HFSS commercial software.
Results
Figure 2 shows a comparison between measured and simulated S11 parameters (panel A) and S21
parameters (panel B) in the absence and in the presence of the nerve.
Fig. 1A shows that measured and simulated S11 parameters exhibit similar trend with frequency.
The measured S11 values are always below -13 dB confirming a very good impedance matching
over the whole frequency band, and the presence of the nerve does not change this behavior. For
what concerns the S21, measured and simulated parameters exhibit a similar trend and comparable
values, always above -0.8 dB, showing that almost all the input power is transmitted to the output
port. The presence of the nerve is almost irrelevant.
The distribution of the E field strength inside the nerve, simulated at 1.5 GHz, is highly not-uniform
(see Fig. 3 showing this distribution on a horizontal plane). The E field is maximum just above the
hole, in correspondence of the edges of the copper conductor. The maximum value is about 800
V/m for 1 W of input power and seems strong enough to elicit the AP onset in the nerve without
inducing electroporation.
Conclusions
In this work, we propose the design, fabrication and characterization of an exposure system for ex-
vivo experiments on nerve samples exposed to nsPEFs. The system is based on a modified
stripline, surrounded by a metallic conductor, except for a small hole on top. This configuration is
459
chosen since it permits a localized exposure of a biological sample placed on the hole and
guarantees the electromagnetic compatibility with the laboratory instrumentation used to record the
sample electrical activity.
Numerical and experimental characterization shows that the system is well matched up to 3 GHz
(S11<-13 dB), thus allowing the undistorted propagation of nanosecond and sub-nanosecond
pulses.
Numerical dosimetry in the nerve reveals a localized exposure with a quite good efficiency: 0.8 kV/m
for 1 W input power. The E field values obtainable in the nerve with standard generators are
compatible with nervous system stimulation during conventional ES; however, experimental studies
are needed to identify a stimulation threshold for such short pulsed signals.
Further work will concern the system characterization even in time domain and the extension of the
application to other biological samples.
References
[1] Schlenstedt, C. et al. Effect of high-frequency subthalamic neurostimulation on gait and freezing
of gait in Parkinson’s disease: a systematic review and meta-analysis. Eur J Neurol 24, 18–26
(2017).
[2] Oldani, L., Altamura, A. C., Abdelghani, M. & Young, A. H. Brain stimulation treatments in bipolar
disorder: A review of the current literature. World J Biol Psychiatry 17, 482–94 (2016).
[3] Rossini et al., Non-invasive electrical and magnetic stimulation of the brain, spinal cord, roots
and peripheral nerves: Basic principles and procedures for routine clinical and research application.
An updated report from an I.F.C.N. Committee, Clinical Neurophysiology 126, 1071-1107, (2015).
[4] Di Lazzaro, V., Capone, F., Apollonio, F., Borea, P.A., Cadossi, R., Fassina, L., Grassi, C., Liberti,
M., Paffi, A., Parazzini, M., Varani, K., Ravazzani, P. A Consensus panel review of central nervous
system effects of the exposure to low-intensity extremely low-frequency magnetic fields, Brain
Stimulation, 6 (4), pp. 469-476, (2013).
[5] Casciola, M., Xiao, S., Pakhomov, A.G. Damage-free peripheral nerve stimulation by 12-ns
pulsed electric field, Scientific Reports, 7 (1), art. no. 10453, (2017).
[6] Rogers, W. R. et al. Strength-duration curve for an electrically excitable tissue extended down to
near 1 nanosecond. IEEE Transactions on Plasma Science 32, 1587–1599 (2004).
[7] Jiang, N. & Cooper, B. Y. Frequency dependent interaction of ultrashort E-fields with nociceptor
membranes and proteins. Bioelectromagnetics 32, 148–63 (2011).
[8]Wang, S. et al. Cardiac myocyte excitation by ultrashort high-field pulses. Biophysical journal 96,
1640–8 (2009).
[9] K.C.Gupta, R.Garg, I. Bahl and P. Bhartia, "Microstrip Lines and Slotlines", Second edition,
Artech House, (1996).
Figures

460
Figure 1. Model of the slotted stripline (A) and picture of the fabricated system (B).

Figure 2. Scattering parameters S11 (panel A) and S21 (panel B) measured (red and green
lines) and simulated (blue and light purple lines) with (dashed lines) and without (solid lines)
the nerve.

461
 Figure 3. |E| field distribution, calculated at 1.5 GHz, inside the nerve on a plane parallel to the
stripline surface at 1 mm of distance.

PA-65 [14:30]
STUDENT PAPER
Enhancement of in vitro gene electrotransfer using gold nanoparticles
Tjasa Potocnik1, Tina Batista Napotnik1, Tamara Pezić1, Matej Reberšek1 & Damijan Miklavčič1
1University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia, 1000
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Tjaša Potočnik
Gold nanoparticles when added to cell suspension prior to electroporation significantly increased
transfection rate using monopolar pulses in vitro. When bipolar pulses were used for gene
electrotransfer, the presence of gold nanoparticles had no effect on transfection rate. Our results
show the possibility of providing sufficient transfection rate with lower voltage applied to cells in
presence of gold nanoparticles.
Introduction
Gene electrotransfer is an established method for gene delivery where high-voltage pulses are used
to increase permeability of cell membrane and thus enabling transfer of genes.1 This method still
has some limitations regarding high applied electric voltage pulses and non-uniformly distributed
electric field. DNA molecules accumulate on cell membrane due to electrophoretic forces which
cause DNA presence only on the cell side facing cathode. Bipolar pulses are reported to increase
the interaction surface between DNA and the membrane.2 Gold nanoparticles (AuNPs) locally
enhance the electric pulse strength and reduce the solution resistance. AuNPs are randomly
dispersed through cell suspension and act like many virtual nanoelectrodes when electric pulses are
applied.3 We hypothesized that addition of AuNPs to cell suspension before performing gene

462
electrotransfer by applying mono- and bipolar pulses can improve transfection rate.
Methods
Chinese hamster ovary cells (CHO) were used for studying gene electrotransfer. Cells were
resuspended in NaPB buffer1 containing 50 µg/ml of 20 nm AuNPs (Merck, Darmstadt, Germany).
After 1 h incubation cell suspension was placed in a cuvette (1 mm) and pEGFP plasmid (40 µg/ml)
was added. Cells were exposed to 8 pulses, 100 µs, 1 Hz, and 800 V/cm, 1000 V/cm or 1400 V/cm
voltage to distance ratio. Bipolar pulses were applied by two different electroporation protocols. In
both protocols, 8 trains with 1 pair of a positive and a negative pulse, duration of pulses 50 µs, and
electric field of 1400 V/cm were applied with train repetition rate 1 Hz. The pauses between pulses
varied: in Protocol 1, the pause between pulses was 1 µs, and in Protocol 2, the pause between
pulses was 10 ms. For the application of pulses, a laboratory prototype pulse generator (University
of Ljubljana) based on H-bridge digital amplifier with 1 kV MOSFETs (DE275-102N06A, IXYS, USA)
was used.4 After electroporation, the cell suspension was transferred to 24-well plate containing 1
ml of cell culture media. Cells were incubated for 24 h, then three images per well on a distinct area
were recorded using a fluorescent microscope (excitation wavelength 488 nm, emitted fluorescence
through a bandpass filter 525/50 nm) and counted by ImageJ program for image analysis.
Transfection rate was determined as a number of GFP positive cells divided by all cells in each well,
expressed in %. Afterwards, cells were trypsinized and analyzed by flow cytometry (excitation
wavelength 488 nm, emitted fluorescence through a bandpass filter 530/30 nm). Statistical analysis
(Two-way ANOVA) was performed using SigmaPlot.
Results
Results obtained using fluorescence microscopy and flow cytometry are in agreement. It is clearly
seen that with increasing electric field, transfection rate (% of cells expressing GFP) is also
increasing. Transfection rate was always higher in cells with added AuNPs when monopolar pulses
were applied. However, statistically significant difference was observed only when cells were
exposed to 8 monopolar pulses, duration of 100 µs, frequency 1 Hz, and electric field of 1400 V/cm
where transfection rate has increased from 13.0±1.5% (without AuNPs) to 19.8±4.3% (with AuNPs)
as determined by flow cytometry.When varying a pause between bipolar pulses, the AuNPs only
slightly increased the transfection rate (Protocol 1: 9.5±4.2% without AuNPs vs. 10.0±6.6% with
AuNPs; Protocol 2: 9.7±2.6% without AuNPs vs. 11.2±5.3% with AuNPs) however, the AuNPs had
no effect on overall transfection rate (number of transfected cells in each sample divided by number
of cells in negative control).In all our experiments we observed decreased survival of cells with
added AuNPs compared to cells that were not exposed to AuNPs (number of cells per image).
Survival was the lowest when cells were exposed to monopolar pulses with electric field of 1400
V/cm and to bipolar pulses according to Protocol 2.
Discussion
Higher transfection rate in presence of AuNPs could be attributed to greater electric voltage
imposed on cells as a result of high conductivity of AuNPs.3 Increased transfection rate when
AuNPs are added is noticeable with all 3 sets of monopolar pulses. According to our results focusing
effect of AuNPs during electroporation could eliminate the need for high-voltage pulses used in gene
electrotransfer and still provide sufficient transfection rate. Electroporation with low-voltage pulses
should also have smaller effect on cell viability. Interestingly, the transfection rate of cells exposed to
8 monopolar pulses, duration of 100 µs, frequency 1 Hz, and electric field of 1400 V/cm with and
without addition of AuNPs was higher than transfection rates gained with bipolar pulses (with or
without AuNPs).When bipolar pulses were used for gene electrotransfer, the presence of AuNPs did
not considerably increase transfection rate. It appears that reversing the polarity of the electric
pulses might have an effect on AuNPs ability to locally enhance the electric pulse strength. It was
shown previously that decreased mobility of AuNPs in electroporation buffer diminishes their pulse
focusing effect which has a negative impact on transfection rate.3With electric pulses focused by
AuNPs it is possible that cells received higher than optimal electric pulses which decreased cell

463
survival. Concentration of AuNPs could also affect cell survival, however, Zu et al. (2014)3 reported
that AuNPs up to concentration 0.1 mg/ml had no major effect on cell survival.In accordance with
decreased survival of cells electroporated in presence of AuNPs we observed lower overall
transfection rate.
Conclusion
AuNPs were added to the cell suspension to enhance transfection rate using monopolar and bipolar
pulses in vitro. When applying monopolar pulses in presence of AuNPs, the transfection rate was
significantly increased, however when cells were exposed to bipolar pulses, the addition of AuNPs
has no effect on the transfection rate.
Acknowledgements
This study was supported by Slovene Research Agency (ARRS) and conducted in the scope of
Pulsed Electric Fields Applications in Biology and Medicine (LEA-EBAM). Experiments were
performed within Infrastructure Programme: Network of research infrastructure centers at University
of Ljubljana (MRIC UL IP-0510).
References
1. Kanduser M, Miklavcic D, Pavlin M, Mechanisms involved in gene electrotransfer using high- and
low-voltage pulses-an in vitro study. Bioelectrochemistry. 2009 Feb; 74(2):265-71. doi:
10.1016/j.bioelechem.2008.09.002. Epub 2008 Sep
2.Golzio M, Teissie J, Rols MP, Direct visualization at the single-cell level of electrically mediated
gene delivery. Proc Natl Acad Sci USA. 2002; 99:1292–1297
3. Zu Y, Huang S, Liao WC, Lu Y, Wang S, Gold nanoparticles enhanced electroporation for
mammalian cell transfection. J Biomed Nanotechnol. 2014 Jun; 10(6):982-92.
4. D. C. Sweeney, M. Reberšek, J. Dermol, L. Rems, D. Miklavčič, and R. V. Davalos,
“Quantification of cell membrane permeability induced by monopolar and high-frequency bipolar
bursts of electrical pulses,” Biochim. Biophys. Acta BBA - Biomembr., vol. 1858, no. 11, pp.
2689–2698, Nov. 2016.

PA-67 [14:30]
Ablate orthotopic pancreatic cancer in animal model with repetitive nanosecond pulsed
electric field
Shengyong Yin1, Guanlei Deng2, Zhen Liu2, Chao Zheng2, Keping Yan2 & Shusen Zheng1
1The First Affiliated Hospital,School of Medicine,Zhejiang University, Hangzhou, China, 310003
2Insitute of Industrial Ecology and Environment, Zhejiang University, Hangzhou, China, 310007
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Shengyong Yin
A homemade R-nsPEF can ablate tumor detectable tumor safely and effectively. It had been
investigated that the potential clinical application of this pulsed electric field energy supply by
ablating large volume tumor with the pulse duration less than 100 nanoseconds. It indicated that
more accurate targeted and minimally invasive focal ablation modality could be achieved by nsPEF
with less energy.
Permeabilization of cells induced by microsecond pulsed electric fields has been applied for tumor
ablation in clinical. Compared with traditional thermal ablation, the most significant advantage is of
non-thermal modality and it can ablate tumor cells but spare vital structures. By delivering pulse with
higher power but short duration, nanosecond pulsed electric field can induce cell death with less
energy, and thus could strengthen advantages of these novel technology. In this paper, a repetitive
nanosecond pulsed electric field (R-nsPEF) generator has been built up to ablate orthotopic xeno-
transplanted pancreatic cancer in nude mice model. The pulsed electric field is delivered by a pair of

464
paralleled electrodes through a transmission line. It is shown that a pair of electrodes can ablate
cylinder rat liver with a diameter over 1.0 cm and a height above 1.0cm, by delivering 1000 pulses
with a duration of 30ns duration, a peak voltage of 15kV, and a repetition rate of 20 Pulsed Per
Second (pps). The growth of orthotopic pancreatic cancer can be significantly inhibited by R-nsPEF
treatment (response ratio, 100%, P<0.05), and more than 40% complete remission ratio (P<0.05)
was found on two independent experiments. And we found that R-nsPEF can ablate tumor safely
and effectively. For the purpose of clinical application, experiments were conducted on rabbit livers,
and results indicate that the ablated rabbit liver with a cross-sectional area larger than 3.0×1.5 cm2
can be realized by using this pulse generator and a pair of paralleled electrodes. It indicated that
more accurate targeted and minimally invasive focal ablation modality could be developed based on
this technology.
Keyword: repetitive nanosecond pulsed electric field, ablation, electroporation, pancreatic cancer
Acknowledgement: This work is supported by National Science and Technology Foundation of
China (81572954), National S&T Major Project of China (2018ZX10301201-006).
Figures

Figure 1. Rat liver ablation by delivering repetitive nanosecond electric field of homemade
energy supply.

465
Figure 2. Ablate orthotopic xeno-transplanted pancreatic cancer in nude mice model with
repetitive nanosecond electric field.

466
 Figure 3. The clinical potential investigation of large volume soft tissue ablation with repetitive
nanosecond electric field.

PA-69 [14:30]
Maternal exposure to high levels of magnetic field non-ionizing radiation during pregnancy
and the risk of preeclampsia: A large prospective cohort study
De-Kun Li1, Hong Chen1, Jeannette Ferber1 & Roxana Odouli1
1Division of Research, Kaiser Foundation Research Institute, Kaiser Permanente, Oakland, California, USA,
94612
Keywords: Epidemiology, ELF/LF, Work in Progress
Presented by: De-Kun Li
In this prospective cohort study, maternal exposure to high levels of magnetic field (MF) non-ionizing
radiation during pregnancy was found to be associated with increased risk of preeclampsia. The
observed association showed a dose-response relationship with increasingly higher MF levels being
associated with a greater risk of preeclampsia. The presence of other risk factors for preeclampsia
appears to exacerbate the risk associated with MF exposure. The finding provides new evidence

467
that MF may impact immunological function.
Introduction
Exposure to high levels of magnetic field (MF) non-ionizing radiation has been associated with the
risk of adverse pregnancy outcomes (e.g., miscarriage) and autoimmune diseases. Preeclampsia is
a pregnancy complication with a strong underlying immunological etiology (e.g., the risk is higher
among nulliparous women and women carrying a male fetus). Thus, there is a strong rationale to
examine whether exposure to high MF levels during pregnancy is associated with increased risk of
preeclampsia.
Methods
A prospective cohort study was conducted among 1,063 pregnant members of the Kaiser
Permanente Northern California (KPNC) region. Participating pregnant women were recruited during
early pregnancy and asked to carry an EMDEX meter for 24 hours during their first or second
trimester to capture their MF exposure level (measured in milligauss) from all sources. Participants
with preeclampsia were identified through the KPNC electronic medical records (EMRs).
Preeclampsia was based on ICD-9 diagnostic codes 642.4x-642.7x. Potential confounders were
ascertained during in-person interviews during pregnancy. After excluding those without valid MF
measurements or whose pregnancy ended before 20 weeks of gestational age, 741 participants
were included in the analysis. The median MF level during the 24-hour period was used to
categorize participants’ level of MF exposure during pregnancy. To improve the accuracy of MF
measurement, MF measurements were further divided into two categories: those measured on a
typical day of pregnancy, which was more likely to reflect the true level of MF exposure during
pregnancy, versus those measured on a non-typical day, which might not reflect the true MF
exposure level during pregnancy. The Cox proportional hazard regression model was used to
estimate the risk of preeclampsia associated with MF exposure during pregnancy after controlling
for confounders.
Results
After controlling for multiple confounders including maternal age, education, parity, smoking during
pregnancy and infant gender using the Cox regression model, pregnant women with high median
MF level (in the top quartile) had 95% higher risk of preeclampsia compared to women whose
median MF level was in the lower three quartiles: adjusted hazard ratio (aHR)=1.95, 95%
confidence interval (CI): 1.03-3.71. The association was noticeably stronger among those whose MF
was measured on a typical day (aHR=2.35, 95% CI: 1.08-5.13) compared to those whose MF was
measured on a non-typical day (aHR=1.44, 95% CI: 0.41-5.03), reflecting that a more accurate MF
measurement allows for a greater likelihood of detecting an underlying association. There was also
a dose-response relationship among those whose MF exposure was in the top quartile with further
higher MF exposure being associated with a greater risk. Through stratification analyses, we further
observed that the association was stronger in the presence of commonly known risk factors for
preeclampsia: among nulliparous women (aHR=4.27, 95%CI: 1.60-11.38), and among women
carrying a male fetus (aHR=3.14, 95%CI: 1.22-8.08).
Conclusions
Maternal exposure to high level of MF non-ionizing radiation during pregnancy is associated with
increased risk of preeclampsia. While the results need to be replicated in future studies, this finding
provides additional evidence that MF may have a biological effect on the immune system.

PA-71 [14:30]
Radiofrequency electromagnetic field exposure and risk perception: A pilot study
Berihun Zeleke1, Christopher Brzozek1, Geza Benke1, Chhavi Bhatt1, Frederik Freudenstein2, Rodney
Croft2, Peter Wiedemann2 & Michael Abramson1

468
1Department of Epidemiology and Preventive Medicine, School of Public Health and Preventive Medicine,
Monash University, Melbourne, Australia, 3004
2School of Psychology, University of Wollongong, Northfields Ave,Wollongong, Australia, 2522
Keywords: Epidemiology, RF/Microwaves, Completed (unpublished)
Presented by: Berihun Zeleke
Despite exposure to far-field radiofrequency electromagnetic fields (RF-EMF) from mobile phone
base stations (MPBS) raising public concern in recent years, it remains unknown if people’s risk
perception is dependent on their knowledge of personal exposure. In this study, we compared risk
perception between three groups: The first group were provided with basic information, the second
with precautionary messages, and the third with measurements of their personal RF-EMF
exposures. No significant difference in the risk perception was observed between groups, but the
personal measurement group had higher mean values for confidence in protecting themselves from
RF-EMF.
Introduction
The relationship between exposure to far-field radiofrequency electromagnetic fields (from mobile
phone base stations, WiFi, FM radio, etc) and potential health effects has raised public concern in
recent years (1). The World Health Organization considered research into understanding the health
effects of radiofrequency electromagnetic fields (RF-EMFs) a priority and emphasized the need to
measure personal exposures in human epidemiological studies (2). Mobile phones are reported to
be the main contributors to personal RF-EMF exposure (3). Far-field RF-EMF sources, such as
mobile phone base stations (MPBS), result in a much lower but constant level of RF-EMR exposure
than near field sources, such as mobile phones, although people often perceive risk from MPBS
exposure to be higher than that from their personal phone use (4). However, it is not well known if
risk perception is lowered by providing personal RF-EMF exposure information.
Objectives
This pilot study aimed to demonstrate the feasibility of objectively measuring personal exposures to
RF-EMF and comparing the level of risk perception between groups of participants.
Design
A comparative experimental study was conducted between June and November 2017 amongst 383
adult participants, recruited from Melbourne, Australia. Participants were randomized to one of the
three study groups: 1) basic information group (n=162) who were provided with basic information
about RF-EMF to read prior to completing a risk perception assessment questionnaire; 2) the
precautionary group (n=158) who were provided with an information pack containing a precautionary
message; and 3) the dosimetry group (n=63) who were provided with a summary of the magnitude
of RF-EMF exposure from MPBS following personal measurements with a dosimeter (ExpoM3). The
same basic information about RF-EMF was also given to the precautionary and dosimetry groups.
For analyses, four outcome variables were considered: (1) risk perception to RF-EMF in general, (2)
exposure perception to MPBS, (3) Risk perception from MPBS, and 4) confidence in protecting from
exposure to RF-EMF. Associations were investigated using independent t-tests, ANOVA or multiple
linear regression models. All analyses were performed using STATA version 13.0 (StataCorp) and
p-values<0.05 were considered statistically significant.
Results
Of the 383 participants (mean age: 34.3 ±12.2 years, and 58.0% female), 44.1% had noticed a
MPBS in their neighbourhood, while 11.5% were unsure. For the personal exposure measurement
group, the mean duration of measurement with the ExpoM3 across participants was 27.4 (SD=4.5)
hours, including night-time and time spent outside the home. The median of the average far-field
RF-EMF from MPBS was 22 μW/m2 (SD=23) of the total 89 μW/m2 for the 63 participants in the
personal RF-EMF measurement group. On average, downlink exposure from MPBS contributed

469
27.7% of the total far-field RF-EMF exposure (range: 1.6%-70.2%). There was no statistically
significant difference in demographic profile between the three groups, nor in their risk perceptions.
The personal measurement group were more confident to be able to protect themselves from RF-
EMF than those in the precautionary group (p=0.019) or the basic information group (P=0.045)
(Figure 1).
Discussion
This study examined people’s perceptions of the risks of non-ionizing radiation, specifically that from
MPBS. The strength of the study was the measurement of personal exposure to RF-EMF and the
comparison of risk perceptions of RF-EMF associated with MPBS in a group of people randomized
into three distinct groups. Previous studies have assessed risk perception of people towards RF-
EMF in general, as well as that from MPBS (4, 5), and the findings were mainly depending on
information obtained from a questionnaire. This study is unique in that participants were provided
with multiple modalities of information (basic text information, precautionary measures, or personal
exposure measurement) and subsequently their risk perception was assessed. No significant
difference in the risk perception was observed between groups, but the personal measurement
group had higher mean values for confidence in protecting themselves from RF-EMF. However, the
findings are limited by the relatively small sample size, particularly in the personal measurement
group.
Conclusion
Providing people with personal measurement data did not reduce their risk RF-EMF risk perception
in general or that relating to MPBS, but it increased their confidence in being able to protect
themselves.
Keywords – Dosimetry, Mobile phones base stations, Risk Perception, Radiofrequency
electromagnetic fields
References
1. Siegrist M, Earle TC, Gutscher H, Keller C. Perception of mobile phone and base station risks.
Risk Anal. 2005;25(5):1253-64.
2. van Deventer E, van Rongen E, Saunders R. WHO research agenda for radiofrequency fields.
Bioelectromagnetics. 2011;32(5):417-21.
3. Roser K, Schoeni A, Struchen B, Zahner M, Eeftens M, Frohlich J, et al. Personal radiofrequency
electromagnetic field exposure measurements in Swiss adolescents. Environ Int. 2017;99:303-14.
4. Freudenstein F, Wiedemann PM, Brown TW. Exposure Perception as a Key Indicator of Risk
Perception and Acceptance of Sources of Radio Frequency Electromagnetic Fields. J Environ
Public Health. 2015;2015:198272.
5. Boehmert C, Wiedemann P, Croft R. Improving Precautionary Communication in the EMF Field?
Effects of Making Messages Consistent and Explaining the Effectiveness of Precautions. Int J
Environ Res Public Health. 2016;13(10).
Figures

470
 Figure 1. Risk Perception (mean values out of maximum of 7) by study group

PA-73 [14:30]
STUDENT PAPER
Repeated long term exposure of Mesenchymal Stem Cells to electric fields using a new pulse
generator
Shirmone Botha1, Borja Lopez2, Óscar Lucía2, Héctor Sarnago2, Alejandro Naval2, José-Miguel Burdio2,
Adeline Muscat1, Tomás García-Sánchez1, Franck Andre1 & Lluis M. Mir1
1Laboratory of Vectorology and Anticancerous Therapeutics UMR8203, Institut Gustave Roussy, Paris,
France, 94805
2Department of Electronic Engineering and Communications, University of Zaragoza, Zaragoza, Spain,
50018
Keywords: Human, Pulsed, Work in Progress
Presented by: Shirmone Botha
In mesenchymal stem cells (MSCs), numerous studies describe the spontaneous oscillations of
cytosolic calcium at various concentrations. We demonstrate that the application of one high voltage
micro second pulse disrupts natural calcium oscillations making it possible to insert an additional
synthetic calcium oscillation pattern over several hours. This study strengthens the perception that
pulsed electric fields can be used as a tool to manipulate calcium oscillations in human adipose
derived MSCs using a novel pulse generator as the one specifically developed. These
manipulations might be used as a tool to improve MSC engineering.
Background
In mammalian somatic cells, the pivotal role of calcium intracellular signaling has been well
established (Pinto et al., 2015). Cytosolic calcium oscillations were shown to influence stem cell

471
differentiation and proliferation, although these mechanisms remain unknown (Sun et al., 2007,
Forostyak et al., 2016). In mesenchymal stem cells (MSCs) recent studies describe the
spontaneous and asynchronous oscillation of calcium (Kawano et al., 2002, Forostyak et al., 2016).
These observations have garnered much attention owing to the critical nature of the stem cell state
and influences that minute changes such as microenvironment, stem cell source and cellular age
may have on the differentiation of MSCs.
Direct electrical fields of low amplitude (~1V/cm) delivered to human MSCs (hMSCs) have shown to
reduce calcium oscillations and impact hMSC osteodifferentiation (Sun et al., 2007). Using nano
second and micro second electric pulses at amplitudes between 200V/cm and 900V/cm, previous
investigations by our group showed that it was possible to generate additional calcium spikes and/or
inhibit spontaneous calcium oscillations in hMSCs over a period of at least 15 minutes (Hanna et al.,
2017). Although these studies adequately demonstrate the manipulation of calcium oscillations over
a short time period, there is still a lack of knowledge regarding the application of pulsed electric
fields (PEFs) on hMSCs over extensive periods of time.
Materials and Methods
In the current study, our intent was to characterize the calcium oscillation profiles and cellular
survival of adipose derived hMSCs (haMSCs) after the repeated application of 100 micro second
PEFsof various electric field amplitudes (200V/cm – 2000V/cm) over several hours. To this end, it
was critical to develop a versatile pulse generator with the capability of delivering both bipolar pulses
to limit the electrolytic effect and pulses at diverse amplitudes over long exposures. HaMSCs were
loaded with Fluo-4 AM to validate modifications in calcium oscillation patterns after pulse delivery.
Results
In this study we demonstrate that natural calcium oscillations can be temporarily discontinued from a
few minutes to scores of minutes after the application of one high voltage pulse at various field
amplitudes. This allows the insertion of a synthetic calcium oscillation pattern to induce an altered
calcium profile over several hours. In addition, we report the continuous high voltage pulsation of
haMSCs over several hours without affecting cellular viability and recovery.
Conclusion
Although the inhibition of calcium oscillations through physical stimulation has been previously
described (Sun et al., 2007, Hanna et al., 2017), this study amplifies the notion that PEFs can be
used as a tool to manipulate calcium oscillations in haMSCs over several hours using a novel pulse
generator as the one specifically developed for this study. Notably, we showed that it is possible to
combine PEFs of various amplitudes and frequencies to ultimately control the calcium oscillation
signature of haMSCs. This application might be used to improve stem cell differentiation and deliver
haMSCs of an improved quality.
References
Hanna, H., Andre, F. M. and Mir. L.M. (2017) Electrical control of calcium oscillations in
mesenchymal stem cells using microsecond pulsed electric fields. Stem cell research and therapy,
8: 91
Forostyak, O., Serhiy Forostyak, S, Kortus, S., Sykova, E., Verkhratsky, A. and Dayanithi, G. (2016)
Physiology of Ca2+ signalling in stem cells of different origins and differentiation stages, Cell
Calcium, 59: 57-66
Kawano, S., Shoji, S., Ichinose, S., Yamagata, K., Tagami, M., and Hiraoka, M. (2002)
Characterization of Ca2signaling pathways in human mesenchymal stem cells. Cell Calcium 32:
165–174
Pinto, M. C. X., Kihara, A. H., Goulart, V. A. M., Tonelli, F. M. P., Gomes, K.N., Ulrich, H., Resende,
R. R. (2015) Calcium Signaling and cell proliferation, Cellular Signalling, 27: 2139-2149
Sun, S., Liu, Y., Lipsky, S., Cho, M. Physical manipulation of calcium oscillations facilitates

472
osteodifferentiation of human mesenchymal stem cells. FASEB J. 21: 1472–1480

PA-75 [14:30]
STUDENT PAPER
A novel method for estimation of brain tissue electrical conductivity: from in silico to in
clinico results
Andres Carvallo1, Julien Modolo1, Pascal Benquet1 & Fabrice Wendling1
1Univ Rennes, INSERM, LTSI - U1099, Rennes, France, 35000
Keywords: Human, ELF/LF, Completed (unpublished)
Presented by: Andres Carvallo
OBJECTIVE: Estimating brain tissue electrical conductivity using low-intensity pulse stimulation.
METHODS: We derived an analytical model of the electric potential generated by SEEG electrodes.
We coupled the electric potential with an electrode-electrolyte interface model to i) derive an
analytical expression of brain tissue response to biphasic pulses and ii) estimate conductivity.
RESULTS: We validated our biophysical model using i) saline solutions calibrated for electrical
conductivity, ii) rat brain tissue, and iii) intracerebral electrophysiological data recorded in epileptic
patients during pre-surgical evaluation. CONCLUSIONS: Rapid and reliable brain tissue electrical
conductivity estimation is achieved using the proposed method.
Introduction
Conductivity is an electrical property that characterizes biological tissue and has been studied
recently in medical fields such as oncology [1, 2] for diagnosis purposes. In brain research, the
measurement of brain tissue conductivity for biomedical applications has been studied since the
1960s [3], different methods for estimating conductivity include non-invasive techniques such as
diffusion tensor imaging -DTI- [4], magnetic resonance electrical impedance tomography -MREIT-
[5], EIT based on scalp electrodes [6].
Each of these methods has its limitations: for example, MREIT suffers from estimation errors
inherent to inverse/forward problems [5], motivating for further efforts in new approaches for
measuring conductivity.
Significant interest also exists in the field of neurology in general, and in epilepsy research in
particular, for potential diagnostic applications that could be developed based on measures of
electrical conductivity.
Indeed, one third of epileptic patients are drug-refractory, and might be candidates for resective
surgery if the epileptogenic zones are well identified and the functional consequences of resection
acceptable. The pre-surgical evaluation aiming at localizing the epileptogenic zone, termed SEEG
(stereoencephalography) is invasive since it involves the implantation, for a few days, of ~20
intracranial electrodes, and extensive clinical monitoring. From the recorded electrophysiological
signals, it can still be challenging to identify without ambiguity epileptogenic areas. Therefore, novel
biomarkers that could be used to identify epileptogenic areas are needed. In this study, we show
that electrical conductivity is a potential candidate.
We developed and validated a novel method for accurate estimation of electrical conductivity in vivo.
After evaluation in saline solutions and rat brain, we made some tests in patients which indicate that
this method can discriminate epileptogenic brain regions from healthy ones.
Methods
We derived an analytical expression of brain electrical conductivity as a function of the applied
electric field generated by pulse stimulation. This involved solving Maxwell equations, taking into
account boundary conditions and initial values, which are typically not available. Therefore,
analytical solutions, when they exist, are highly complex to solve. Consequently, we used
reasonable assumptions that led to a semi-analytical formula for electrical conductivity as a function

473
of the electric field.
To derive our model, termed “cylindrical electric field model” (as a reference to the electrode’s
geometry, which constrains the symmetry of the induced electric field), we used a differential ring,
and integrated its contribution along the electrode’s height, as shown in Figure 1(a) and (b). Using
this approximation, we derived an analytical expression of the medium resistance Rm that depends
only of the electrode’s geometry and conductivity.
The biophysical phenomena taking place at the interface between the electrode and tissue also
have to be examined. We assumed that electric charge is injected from the electrode to the medium
through faradaic processes of reduction/oxidation, where electrons are transferred between the two
phases. The proposed electrical model for the electrode-tissue interface is shown in Figure 1(c). In
this model, Zf is the Faradaic impedance representing Faradaic processes, whereas Cdl models the
double layer capacitance, i.e. the ability of the electrode to cause charge flow in the medium without
electron transfer.
The medium resistance Rm was estimated using the stimulation artifact itself. With the knowledge of
this estimation and the analytical expression of the medium resistance for cylindrical electrodes,
electrical conductivity can be estimated.
Results
The electric potential computed using the cylindrical model and Comsol Multiphysics (AC/DC
module) are shown in Figure 2(a) and (b). Comsol was used as a benchmark to validate our
cylindrical electrical model. As illustrated, both solutions are in agreement, with an error between the
two estimations remaining on average less than 9%, while being considerably faster to compute.
Next, in order to validate the electrode-tissue model, we used a conductivity calibrated saline
solution. The electrode-medium double layer model reproduced the time course of the recorded
stimulation artifact, and enabled us to estimate the resistance, and hence the conductivity of the
calibrated solutions as shown in Table I. We used four solutions with values of conductivity of
0.1033, 0.2027, 0.3943 and 0.5786 S/m, respectively. Estimated conductivity values were close to
actual conductivity values, with a standard deviation smaller than 10%.

Double layer Faradaic impedance

Resistance (Ω) Conductivity (S/m)
capacitance (µF) (Ω)

2079.5 ± 20.9 1.209 ± 0.040 1652.6 ± 92.1 0.1130 ± 0.0012

1161.2 ± 15.3 1.344 ± 0.033 1676.1 ± 83.6 0.2177 ± 0.0034

720.2 ± 10.1 1.449 ± 0.032 1623.4 ± 66.1 0.3922 ± 0.0073

TABLE I. Saline solutions parameter estimation. Estimations were performed from the experimental
recordings carried out in each of the four calibrated saline solutions.
In the next step, we performed post-mortem conductivity measurements in the brain of N=2 adults
Sprague-Dawley rats euthanized using a CO2 gradient in accordance with the European
Communities Council Directive of 24 November 1986 (86/609/EEC). Stimulation was performed
within few minutes post-mortem within the right hemisphere. Estimated values of double layer
parameters and conductivity are shown in Table II.

Double layer Faradaic impedance

Resistance (Ω) Conductivity (S/m)
capacitance (µF) (Ω)

474
 First Rat - Right Hemisphere

1255.1 ± 61.5 0.546 ± 0.046 1631.2 ± 145.2 0.1194 ± 0.0064

Second Rat - Right Hemisphere

1904.5 ± 14.8 0.975 ± 0.014 2804.7 ± 118.6 0.0785 ± 0.0006

TABLE II. Rat brain parameter estimation. Estimations were performed from the experimental
recordings carried out in the ex-vivo rat brain N=2.
Finally, in order to provide a first test for this novel conductivity estimation method, we acquired
electrophysiological data from N=2 epileptic patients undergoing SEEG in the context of pre-surgical
evaluation for epilepsy. Conductivity estimations were performed by stimulating/recording several
regions based on the electrophysiological signals for clinical evaluation.
An example of patient data is shown in Figure 3, where seven electrode contacts (labeled A4-5,
A9-10, CR4-5, B’3-4, TP1-2, B1-2 and TP3-4, referring to each pair of contacts) from different
regions are highlighted. Estimated values of conductivity are presented for each electrode along
with corresponding electrophysiological recorded signals. For A4-5 and A9-10, normal
electrophysiological activity characteristic of grey matter was observed, and estimated conductivity
values were 0.32 S/m. The signal recorded from the CR4-5 electrode had a low amplitude
electrophysiological activity, typically seen in white matter, with an estimated conductivity value of
0.17 S/m. Electrode contacts B’3-4, TP1-2, B1-2 and TP3-4 displayed epileptic activity, with spikes
and high frequency oscillations. Interestingly, for these electrodes, lower conductivity was estimated
as compared to A4-5 and A9-10, which were electrodes located in healthy brain tissue.
Discussion
In a typical clinical setting, electrical conductivity cannot be measured during pre-surgical
intracranial recordings since functional (electrical) stimulation aims at identifying the epileptogenic
zones by using electrophysiological markers such as after-discharges and fast ripples. In this study,
we proposed to measure conductivity as a complementary observation to electrophysiological
markers, to provide further information about the biophysical properties of tissue. Potential
correlation between the two would have diagnostic value.
One originality of our method is that the stimulation artifact is typically discarded during stimulation
sessions due to amplifier saturation. Indeed, in a clinical setting, SEEG acquisition systems are
typically conceived for signals up to ~100 µV. Stimulation artifacts (depending greatly on the current)
are on the order of ~1 V, preventing artifact recorded using standard SEEG hardware. Importantly,
our method does not required higher stimulation currents than used clinically, and even provides
accurate estimations for significantly lower currents. Furthermore, the biphasic pulses used are
typically between 0.1 ms and 2 ms, , motivating the need for a much higher sampling frequency
(100 kHz here) than those used in clinical SEEG acquisition systems (in general, 1024 or 2048 Hz)
to capture the stimulation artifact time course accurately. Also, we presented a simplified analytical
electric field/potential model, which enables to adapt to cylindrical electrodes of different geometries
with a drastically lower computational cost than numerical methods such as Comsol.
Brain tissue is composed primarily of cells and extracellular medium. There are several ways of
studying the tissue impedance from an electric circuit perspective [7]. In this model, we considered
that the tissue was only resistive, involving that other properties of tissues such as cell membrane
capacitance [8] were not taken into account. There might be an interest in evaluating other
approaches; for example investigating the encapsulation tissue forming around the electrode after
surgery [9].

475
Conclusion
We proposed a novel method to estimate biophysical properties of brain tissue, typically the
electrical conductivity. We computed analytically the electric potential and field induced by the
stimulation electrode, and also for the electrode-electrolyte interface characteristics, to describe the
brain tissue response to pulse stimulation. From fitting simulated responses with actually- recorded
responses, the model provided access to biophysically relevant parameters. This method was
validated using in vitro and ex vivo settings, and also in patients. This method of electrical
conductivity evaluation can be potentially used n a broad range of applications, including biophysics
and medicine.
References
[1] G. Daglar et al., “Effectiveness of breast electrical impedance imaging for clinically suspicious
breast lesions,” Bratisl. Med. J., vol. 117, no. 09, pp. 505–510, 2016.
[2] A. Peyman et al., “Variation in dielectric properties due to pathological changes in human liver:
Dielectric Properties of Liver Tumors,” Bioelectromagnetics, vol. 36, no. 8, pp. 603–612, Dec. 2015.
[3] B. W. Robinson, “Localization of intracerebral electrodes,” Exp. Neurol., vol. 6, no. 3, pp.
201–223, Sep. 1962.
[4] M. Akhtari, D. Emin, B. M. Ellingson, D. Woodworth, A. Frew, and G. W. Mathern, “Measuring the
local electrical conductivity of human brain tissue,” J. Appl. Phys., vol. 119, no. 6, p. 064701, Feb.
2016.
[5] H. J. Kim et al., “In vivo electrical conductivity imaging of a canine brain using a 3 T MREIT
system,” Physiol. Meas., vol. 29, no. 10, pp. 1145–1155, Oct. 2008.
[6] L. Fabrizi et al., “Factors limiting the application of electrical impedance tomography for
identification of regional conductivity changes using scalp electrodes during epileptic seizures in
humans,” Physiol. Meas., vol. 27, no. 5, pp. S163–S174, May 2006.
[7] E. T. McAdams and J. Jossinet, “Tissue impedance: a historical overview,” Physiol. Meas., vol.
16, no. 3A, p. A1, 1995.
[8] E. T. McAdams and J. Jossinet, “Problems in equivalent circuit modelling of the electrical
properties of biological tissues,” Bioelectrochem. Bioenerg., vol. 40, no. 2, pp. 147–152, Aug. 1996.
[9] S. F. Lempka, S. Miocinovic, M. D. Johnson, J. L. Vitek, and C. C. McIntyre, “In vivo impedance
spectroscopy of deep brain stimulation electrodes,” J. Neural Eng., vol. 6, no. 4, p. 046001, Aug.
2009.
Figures

476
Figure 1. (a) Intracranial electrode, accounting of a pair of contacts of radius R, height hand
separated by a distance l. (b) Differential ring contribution, constant current density normal to
the ring. (c) Two-electrode-medium circuit model, where Cdl represents the double layer
capacitance, Zf the faradaic impedance and Rm models the medium resistance. In a clinical
situation, two contacts are located in brain tissue (the medium); one of the electrodes is
typically termed “working electrode” and the other “counter electrode”.

Figure 2. (a) Electric potential distribution of an SEEG electrode, computed with the cylindrical
model and (b) computed using Comsol Multiphysics. The electric potential distribution appears
qualitatively and quantitatively in agreement with the cylindrical model.

477
 Figure 3. Electrophysiological activity of an epileptic patient recorded in healthy and epileptic
regions, with comparison of the estimated conductivity.

PA-77 [14:30]
Risk perceptions of mobile communication in Japan: A 2017 survey
Ayumi Masuchi1
1Hokkai-Gakuen University, Sapporo, Japan, 0628605
Keywords: Human, RF/Microwaves, Completed (unpublished)
Presented by: Ayumi Masuchi
This study was conducted to reveal average person’s knowledge and risk perceptions related to
mobile communication in Japan. A Web survey was completed by 300 teenagers and 2,000 adults.
The results were compared to our 2013 survey results. While most used mobile phones more
frequently, respondents’ information and knowledge level about the risks of electromagnetic fields
radiating from mobile phones and base stations was about the same or lower than in 2013. Their
concern level was slightly lower. Most respondents did not consider the risk as serious and did not
have a clear opinion. Young people use mobile phones longer, feel more physical discomfort and
worry more about the health effects when using mobile phones than older adults.
Introduction
The International Agency for Research on Cancer (IARC) reported its findings about the potential
effects of exposure to electromagnetic fields (EMFs) radiating from mobile phones on human health
in May 2011.
After the report by IARC, we conducted a survey in Japan in 2013 to reveal the average Japanese
person’s knowledge and perceptions about the health risks related to mobile communication

478
(Masuchi, 2014). The results of the survey indicated that most of the respondents showed slight
concern about the risks of exposure to EMFs and they did not consider the risks as serious or might
not have formed opinions on this topic yet. However, about 10% of the respondents expressed
strong concern regarding the risk of using mobile phones. Those 10% of respondents were relatively
sensitive to risk information about the potential health threats and were unwilling to take this risk.
In recent years, mobile communication technology has been developing rapidly, with more and more
users having upgraded their mobile phones to smartphones. Using mobile phones seems to be
increasingly indispensable for not only adults but also teenagers.
This study was conducted to analyze the average Japanese person’s knowledge and risk
perceptions related to mobile communications in Japan five years after our previous study in 2013.
In addition to adults, teenagers aged between 15–19 years old were also the targets of this survey.
Responses from the adult sample were compared with responses from the 2013 survey.
Methods
Respondents
Our survey was conducted in December 2017. Two representative samples of the Japanese
population were invited to participate in the study: i.e., teenagers aged between 15–19 years and
adults aged between 20–79 years. These samples were collected from an online panel. In one
week, 2,300 respondents (300 teenagers and 2,000 adults) completed the questionnaire. Adult
respondents were stratified on demographic features (age, sex, and area of residence) to ensure
that the sample was an accurate representation of the population.
Measures
The questionnaire contained a wide range of questions concerning the issue of mobile phones and
base stations.The variables measured the frequency of using a mobile phone per day, knowledge
about EMFs, familiarity with EMFs key words, beliefs about the health effects caused by exposure to
EMFs radiating from mobile phones and base stations, perceived risks and benefits of using mobile
phones, reaction to information about the potential health risks of EMFs as reported by the IARC in
2011, and concerns about a formal risk assessment that will be conducted by the World Health
Organization (WHO) in the future.These questions were almost identical to the 2013 survey. In this
study, some questions were added: i.e., the frequency of feeling discomfort when using a mobile
phone, beliefs about the health effects caused by exposure to EMFs radiating from some systems or
devices (e.g., WiFi, microwaves, metal detectors, magnetic resonance imaging [MRI], and other
medical instruments).
Results
The main results of this study are as follows.
Almost all respondents used a mobile phone in the 2017 sample (teenagers: 96.0%; adults: 97.5%).
The usage rate has increased slightly compared with the 2013 sample (94.7%). The frequency of
daily mobile phone use has also increased in the 2017 adult sample. 14% of the 2017 adults use
their mobile phone for 1 hour or more a day (2013: 10.8%). The percentages are different across
age groups in the 2017 sample (teenagers: 40.7%, 20s: 31.9%, 30s: 26.2%, 40s: 14.9%, 50s:
13.0%, 60s: 8.4%, 70s:4.1%). A majority of respondents considered using a mobile phone useful,
especially amongst teenagers (teenagers: 90.3%, 20-30s: 85.6%, 40-50s: 78.3%, 60-70s: 69.9%).
The percentage of the 2017 adults (76.9%) decreased slightly compared with the 2013 adults
(78.9%). While the number decreased slightly among the 40-70s, it increased among the 20-30s.
Overall, respondents’ information and knowledge level about EMFs in the 2017 sample was about
the same or lower than that in the 2013 sample. Only half of the respondents (2017: 51.8%, 2013:
50%) knew that mobile communications operate using radiation similarly to radio and television
broadcasts. The percentage of the respondents who were aware of the existence of base stations in
the neighborhood decreased from 50% in 2013 to 40% in 2017. Out of the total respondents, 83.1%
did not know the term SAR (specific absorption rate). This ratio has increased by about 20%
compared to the 2013 sample (59.0%). In 2017, about 40% of the respondents reported interest in
479
the issue of the potential risk of EMFs in 2017. This percentage has decreased compared with 2013
(2017: 41.1%, 2013: 50.0%). Less than 30% (2017: 27.4%, 2013: 26%) reported a fear of EMFs.
Table 1 shows the subjective evaluations of the health threats caused by the exposure to EMFs from
mobile phones in the 2017 and 2013 samples. In both of the 2017 and 2013 samples, about 40% of
the respondents believed that exposure to EMFs might cause a headache or a feeling of discomfort
(2017: 42.6%, 2013: 39.0%). About 35% believed that EMFs caused negative effects on pregnant
women and infants (2017: 31.8%, 2013: 35.3%). This percentage has decreased somewhat in 2017.
Almost half of respondents indicated no knowledge of exposure to EMFs might increase cancer or
allergy risk (2017: 50.0% and 52.5%, 2013: 45.5% and 49.5%, respectively). Over 30% of the
respondents did not believe that exposure to EMFs increase cancer or allergy risk (2017: 31.9% and
35.5%, 2013: 34.8% and 41.4%, respectively). Perceived risk was relatively low across all
respondents. Only about 25% (2017: 25.9%, 2013: 23%) considered using a mobile phone as a risk.
Table 2 indicates comparative estimates for risk and benefit perception of using mobile phones and
base stations in the 2017 and 2013 samples. The benefits of mobile phones outweigh the risks for
most of the respondents (2017: 49.2%, 2013: 60.7%). While this ratio decreased by about 10% in
2017, the number of respondents who had no idea increased by 10%. These results are similar for
base stations. More than half of participations (2017: 51.1%, 2013: 70%) indicated that they
received information on this issue through mass media, such as television and newspapers. Only a
minority pursued an active strategy for gaining more information. This tendency was more salient in
2017 than in 2013.
In this study as in the 2013 study, the respondents were asked to read an article explaining the
potential risk of using a mobile phone reported by the IARC in 2011. After reading the article,
respondents answered several questions. In 2017, 79.2 % of the respondents reported that they had
not heard about or did not remember the report, while 18.0% reported that they were not sure. In
2013, the respective numbers were 88.8% and 8.8%. However, about half of respondents were
interested and had a feeling of anxiety in both 2013 (53.9%) and 2017 (41.9%).
Only 7%–8% of the respondents felt sufficiently informed about this issue by the government or
companies that had a responsibility for this issue in 2013 and in 2017. Although about half of the
respondents (50.2%) in 2017 indicated that they were concerned about the formal risk assessment
to be reported by the WHO in the near future, this percentage has decreased by nearly 30%
compared with 2013 (77.0%).
Table 3 indicates how often respondents felt afraid of base stations and their acceptance of base
stations in their neighborhood in the 2017 and 2013 samples. A majority of the respondents did not
report clear opinions, while only about 13% feared to base stations in both samples. The percentage
of respondents who did not accept base stations decreased slightly in 2017 (2017: 21.1%, 2013:
26.0%).
Conversely, several respondents (2017: 11%, 2013: 13%) were more concerned about the risks of
using mobile phones. These respondents were relatively sensitive to risk information on potential
health threats and were unwilling to take this risk even if using mobile phones was relatively more
convenient for them than those who expressed less concern. This tendency was the same as in
2013. In the 2017 survey, 54.8% of respondents worried about any effects of EMFs radiating from
some devices or systems on human health. The highest percentage was for microwaves (30.0%),
followed by mobile phones (28.3%) and MRI (20.3%). Younger age groups worried more about
mobile phones than did other age groups (teenagers: 33.7%, 20-30s: 36.7%, 40-50s: 27.8%,
60-70s: 21.1%). However, older age groups worried more about microwaves and MRIs than did
younger groups (microwaves: teenagers: 20.3%, 20-30s: 26.9%, 40-50s: 31.8%, 60-70s: 34.3%,
MRI: teenagers: 14.7%, 20-30s: 16.2%, 40-50s: 21.0%, 60-70s: 24.5%).
A majority of 63.9% reported one or more physical discomforts out of six (e.g., eye strain, stiff
shoulders, bad posture) when using a mobile phone, especially among younger age groups in 2017
(teenagers: 85.0%, 20s: 82.2%, 30s: 79.4%, 40s: 70.0%, 50s: 63.4%, 60s 47.7%, 70s: 38.3%). In
total, “eye strain” took the top position (49.0%), followed by “always looking down” (31.8%) and “stiff

480
shoulders” (30.8%). The average number of reported discomforts were different by age groups
(teenagers: 2.3, 20s: 2.5, 30s: 2.2, 40s: 1.9, 50s: 1.5, 60s 1.0, 70s: 0.8).
Discussion
Overall, the respondents in this study use mobile phones more frequently compared with 2013.
Teenagers and 20–30s use mobile phones longer and consider them more useful than do over-40s.
However, respondents’ information and knowledge level about the risks of EMFs radiating from
mobile phones and base stations is about the same as or lower than in 2013. Also, their concern
level is slightly lower compared with 2013. Most do not consider the risks as serious and do not
have clear opinions. These tendencies are more salient in 2017 compared with 2013.
However, as shown in the 2013 study, about 10% expressed stronger concern about the risks and
estimated the risks higher compared with others. Most of the young respondents feel physical
discomfort when using their mobile phones. They are more concerned about the potential health
risks of EMFs from mobile phones than are elderly respondents. These results suggest that it is
important to inform people who use mobile phones frequently about the scientific evidence related to
the health effects of EMFs.
Acknowledgment
This study was funded by the Association of Radio Industries and Businesses (ARIB) in Japan as a
research program in 2017. The authors would like to express their gratitude to all persons who
participated in and supported this study.
Figures

Figure 1. Beliefs about potential health effects.

481
 Figure 2. Comparative estimates for risks and benefits of using mobile phones.

Figure 3. Feeling fear towards and acceptance of base stations in the neighborhood.

PA-79 [14:30]
Effect of Radio Frequency Radiation (RFR) emitted from cell phone on human health
Radhey Shyam Sharma1, Ravindra Nath1 & Rajeev Singh2
1Reproductive biology, Maternal and Child Health, Indian Council of Medical Research, Ministry of Health and
Family Welfare, New Delhi, India, 110029
2Department of Environment Science, Delhi University, New Delhi, India, 110052
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Dr. Radhey Shyam Sharma

482
No conclusive data available so far on the adverce effect of RFR emitted from cell phone. Even
epidemiological evidences generated so far are inadequate for a comprehensive evaluation of risk
caused by RFR. To address the above issue, the ICMR is undertaking a multi-disciplinary
prospective cohort study in Delhi to find out adverse effects of RFR, if any, emitted from cell phone
on adult Indian population. The details of the protocol, the current status of the study and other
ongoing research activates in this area of public importance going on in the country will be
presented during the conference.
Cell phone technology has revolutionized the telecommunication scenario worldwide including India.
Mobile phone entered in India in 1994 and now there are more than 1003.49 millions cell phone
users with around 7,36,654 Cell Phone Towers in the country. The cell phone and cell phone tower
transmit and receives the radio frequency radiation (RFR) which falls in the microwave part of the
electromagnetic spectrum covering frequency ranged from 100KHz to 300GHz. The widespread use
of mobile phone has resulted in increased human exposure to radiofrequency radiation. Following
the enormous increase in the use of cell phone around the world, the health concerns are being
raised from time to time both in print and electronic media.
RFR emitted from cell phone is of a short-term, repeated nature (coherent) at a relatively high
intensity, whereas RFR emitted from cell phone tower is of long duration but a very low intensity.
Biological effects of low frequency (<100Hz) electric and magnetic fields are well established and
experts committees of various countries have reviewed these effects. There are two distinct
possibilities by which health could be affected as a result of RFR exposure. There are thermal
effects caused by holding mobile phones close to the body. Secondly, there could be possible non-
thermal effects from both phones and base stations. On review of existing literature, it is evident that
exposure to radiation from cell phones causes adverse health effects but there is no conclusive data
available so far on this issue. Even epidemiological evidences generated so far are inadequate for a
comprehensive evaluation of risk caused by RFR. However, the growing body of scientific evidences
indicates some bio-effects and possible adverse health effects of RFR which merits further
investigations.
The numbers of studies conducted and published from India have indicated that the RFR emitted
from cell phone and cell phone towers are adversely influencing the biological system by altering
levels and expression of various biomarkers. The results of these studies were presented during the
last year BioEM conference held at Asilomar, California, USA. The Indian data is in compliance with
the other international studies reporting the biological effects of RFR emitted from cell phone and
cell phone towers. But it is not yet clear that how much time these biological effects will take to result
in to health hazard.
To address the above issue, the ICMR has already initiated a multi-disciplinary prospective cohort
study in Delhi & National Capital Region (NCR) to find out adverse effects of RFR, if any, emitted
from cell phone on adult Indian population. Under this study efforts are going on to examine whether
use of cell phone is associated with reproductive dysfunctions, male infertility, neurological disorders
(cognitive behavior, sleep related disorders, depression etc.), cardiovascular disorders,
Otorhinolaryngology (ENT) disorders and promote cancer if any, in Human Volunteers. Under this
study efforts are also going on to survey the health status of the people residing near the cell phone
tower in Delhi. The details of the protocol, the current status of the study and other ongoing research
activates in this area of public importance going on in the country will be presented during the
conference.

PA-81 [14:30]
Effect of the 1800 MHz-RF exposure on in vitro neuronal networks activity: role of the
modulation frequency
Corinne El Khoueiry1, Thinhinane Yebda1, Audrey Giremus1, Rémy Renom1, Florence Poulletier De
Gannes1, Emmanuelle Poque-Haro1, Isabelle Lagroye1, Bernard Veyret1 & Noëlle Lewis1

483
1IMS Laboratory, Bordeaux University, Talence, France, 33405
Keywords: In vitro, Pulsed, Work in Progress
Presented by: Isabelle Lagroye
In a previous study, we reported a clear decrease in bursting rate of neuronal cultures under
RF-1800 MHz exposure. The amplitude of the effect was greater with the GSM-1800 MHz signal
compared to the CW signal. The present work aimed at assessing the contribution of the modulation
frequency of the pulsed RF signal in the elicitation of the effect. We tested different multiples of the
GSM modulation frequency (108.5; 217; 434 and 868 Hz) at the same averaged SAR level (2.4
W/kg). The preliminary results show that the effect was present for the GSM frequency (217 Hz) but
not for other frequencies. Further work is ongoing to elucidate the mechanism of this effect.
Introduction
The rapid development of wireless communications has raised questions about their potential health
risks. Up to now, the only identified biological effects of radiofrequency fields (RF) are caused by
heating. However, several human electroencephalography (EEG) studies have reported variations in
the EEG power spectrum during and/or after RF exposure, in resting EEG and during sleep (Van
Rongen et al. 2009; SCENIHR 2015), suggesting that RF exposure may directly influence brain
dynamics. Thus, a question remains open about potential nonthermal biological effects, especially
on the central nervous system (CNS). In our previous work, we studied the effect of a 15-min RF
exposure at SAR (Specific Absorption Rate) ranging from 0.01 to 9.2 W/kg on in vitro neuronal
networks for both pulsed (GSM-1800 MHz) and non-pulsed continuous wave (CW-1800 MHz)
signals. Both GSM and CW signals elicited a clear decrease in bursting rate during the RF exposure
phase. This effect grew larger with increasing SAR, and its amplitude was greater with the GSM
signal (El Khoueiry et al., 2018). However, we have been unable so far to elucidate the mechanism
of the observed RF effect on bursting activity of neuronal networks.
In view of the differential results detected between GSM and CW, we further characterized this effect
by testing combinations of the pulsed signal parameters (amplitude, repetition rate, and pulse
width), in order to reveal their respective roles in the induction of the effect. In this study, we
examined the role of the pulse modulation frequency, using pseudo-GSM signals pulsed at half
(108.5 Hz), double (434 Hz) and quadruple (868 Hz) of the original frequency of a GSM-1800 MHz
signal (217 Hz). These signals were used to expose neuronal cultures for 15 min at a same
averaged SAR (2.4 W/kg), in order to assess the contribution of the modulation frequency of the
pulsed RF signal in the elicitation of the effect.
Materials and Methods:
Cell Culture
We used cultures of cortical neurons from rat embryos at E18. Using standard methods, 105 cells
were plated on MEAs (multi-electrode arrays) coated with Laminin and either PEI
(polyethyleneimine) or Poly-D-lysine. The MEAs (Qwane, Switzerland) are fitted with 60 platinum
electrodes (200 μm spaced; 40 μm diameter tips). Each MEA was placed on a mounting support
with integrated 60 channels filtering and amplifying, (x1200 gain; Multi Channel Systems, Reutlingen
MCS, Germany). Raw data were sampled and recorded at 10 kHz/channel using the commercial
software MC Rack (MCS).
RF signals and Exposure System
The GSM-1800 MHz is a pulsed-modulated signal at 217 Hz with a 1/8 duty-cycle and hence a
pulse width of 579 µs. In this study, we changed the modulation frequency while keeping the original
pulse width of 579 µs. We thus generated “pseudo-GSM” signals, with a modulation frequency at
half (F1 = 108.5 Hz), double (F2 = 434 Hz) and quadruple (F3 = 868 Hz) of the original frequency of
the GSM signal (217 Hz). Comparing the effects of the GSM and pseudo-GSM signals was made at
the same time-averaged SAR (2.4 W/kg), which guaranties the same thermal conditions.
The MEAs hosting cultured neuronal networks were placed inside a TEM (transverse

484
electromagnetic) cell. In total, we used 24 cultures to compare exposures at different modulation
frequencies (n = 6 per group). The exposure protocol consisted of a 15-min exposure phase (E) to
the pulsed RF signal, preceded by two “pre” phases (sham-exposed; S1 and S2) of 15 min each,
and followed by two “post” phases (sham-exposed; P1 and P2) phases of 15 min each. Spikes were
detected using an algorithm implemented in the SPYCODE analysis software [Maccione et al.,
2009]. To quantify neuronal electrical activity, we used mainly the mean bursting rate (MBR) metric,
defined in terms of bursts/minute.
Acquisition and signal Processing
RF exposures were carried out inside a dry incubator (37 °C, 5% CO2), which contained the MEA,
pre-amplifier and exposure system. A membrane was used to prevent evaporation from the MEA
culture chamber during exposure. The neuronal networks were exposed at ages between 17 and 28
DIV (days in vitro), a range within which the neuronal activity is optimal in terms of the balance
between random spikes and bursts [Van Pelt 2005; Chiappalone et al., 2005].
As expected in the dosimetry study (Merla et al. 2011), the electromagnetic field created
interference on the electrodes, which resulted in a recorded artifact. This artifact has the time profile
of the modulation signal, and was successfully eliminated for the GSM frequency using a stop-band
filter (Moretti et al. 2013), in our previous experiments. The new “pseudo-GSM” signals induce a
new time profile of the artifact and the previous filter reveals less efficient; an adaptation is being
implemented.
Results
The 15-min exposure of the neuronal cultures to the GSM signal at 2.4 W/kg resulted in a decrease
of ca. 70% in MBR. Moreover, this decrease continued during the P1 phase (ca. 82%) and P2 (ca.
78%) phases (Figure 1 A).
However, when using the pseudo GSM signals at F1, F2 or F3, we observed a clear increase in
MBR for the P1 phase (ca. 81%, 92% and 63%, respectively). The MBR of the P2 phase was also
increased (for F1 and F2), or slightly decreased (for F3) (ca. 121%, 141% and – 3%, respectively)
(Figure 1 B, C, D). We have been so far unable to exploit the MBR during the E phase because of
the poor efficacy of the current artifact filter. However, the results of P1 and P2 suggest that the
exposure to the pseudo-GSM signal at F1, F2 or F3 have no effect on the bursting activity of the
neuronal cultures.
Discussion
Our results suggest a contribution of the modulation frequency of a pulsed RF signal in the observed
effect on neuronal networks activity. An ongoing work is aiming at developing the artifact filter in
order to better investigate the effects during the RF exposure. More studies are needed in future
investigations to understand the role of the modulation frequency in the observed effect, as well as
to identify the contribution of other parameters (amplitude and pulse width).
This work was funded by the French ANSES agency (2015- EST-15RF-19 MOTUS project).
References
Chiappalone M, Novellino A, Vajda I, Vato A, Martinoia S, Van Pelt J. 2005. Burst detection
algorithms for the analysis of spatio-temporal patterns in cortical networks of neurons.
Neurocomputing 65:653–662.
El Khoueiry C, Moretti D, Renom R, Camera F, Orlacchio R, Garenne A, Poulletier De Gannes F,
Poque-Haro E, Lagroye I, Veyret B, Lewis N. 2018. Decreased spontaneous electrical activity in
neuronal networks exposed to radiofrequency 1800 MHz signals. J. Neurophysiol (under revision)
Maccione A, Gandolfo M, Massobrio P, Novellino A, Martinoia S, Chiappalone M. 2009. A novel
algorithm for precise identification of spikes in extracellularly recorded neuronal signals. J Neurosci
Methods 177:241–249.
Merla C, Ticaud N, Arnaud-Cormos D, Veyret B, Lévêque P. 2011. Real-time RF exposure setup

485
based on a multiple electrode array (MEA) for electrophysiological recording of neuronal networks.
IEEE Trans Microw Theory Tech 59 : 755–762.
Moretti D, Garenne A, Haro E, Poulletier de Gannes F, Lagroye I, Lévêque P, Veyret B, Lewis N.
2013. In-vitro exposure of neuronal networks to the GSM-1800 signal. Bioelectromagnetics.
SCENIHR. Health effects of EMF – Scientific Committee on Emerging and Newly Identified Health
Risks, 2015.
Van Pelt J, Vajda I, Wolters PS, Corner MA, Ramakers GJ. Dynamics and plasticity in developing
neuronal networks in vitro. Prog Brain Res 147:173–188, 2005.
Van Rongen E, Croft R, Juutilainen J, Lagroye I, Miyakoshi J, Saunders R, de Seze R, Tenforde T,
Verschaeve L, Veyret B, Xu Z. Effects of radiofrequency electromagnetic fields on the human
nervous system. J Toxicol Environ Health B Crit Rev 12:572–597, 2009.
Figures

Figure 1. Column bar graph of normalized Mean Bursting Rate (MBR) for the RF exposed
cultures, along the five phases (S1, S2, E, P1, P2) of exposure protocol, for each of the 4 used
RF signals (n = 6 for each group).

486
PA-83 [14:30]
The BBEMG research program - A multidisciplinary approach in studying long-term effects
of ELF-EMF
Maryse Ledent1, Maurice Hinsenkamp2, Jean-Francois Collard2, Catherine Bouland3, Christophe Geuzaine4,
Veronique Beauvois4 & Luc Verschaeve1
1Risk Assessment, Scientific Institute of Public Health (WIV-ISP), Brussels, Belgium, 1050
2Laboratoire de recherche en orthopédie traumatologie, Brussels, Belgium, 1070
3Ecole de Santé Publique, Brussels, Belgium, 1070
4Applied and Computational Electromagnetics (ACE), Liège, Belgium, 4000
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: Maryse Ledent
Since 1995, the Belgian BioElectroMagnetics Group (BBEMG) investigates health effects of ELF-
EMF. In the current 2017-2021 project, six research teams from three universities (Liège, Gent,
Brussels) and one research institute (Brussels) work together in a multidisciplinary way. Three main
topics are considered: (1) Epidemiological reviews and meta-analysis of childhood leukaemia; (2)
EMF exposure modelling from the HV network; (3) Long-term occupational and residential ELF-EMF
exposure. This paper is focusing on the latter topic. A collaborative research combining a
cytogenetic biomonitoring study of EMF-exposed workers, an analysis of gene expression profiles
and a survey on cumulative environmental nuisances is being set up.
Introduction
Despite numerous studies conducted to date, the potential effects of extremely low frequency
electromagnetic fields (ELF-EMF) are still uncertain. On the one hand, results of epidemiological
studies are controversial because of their limitations (bias, too small populations, divergent
results...), and on the other hand in vitro and in vivo studies often involve exposures limited in time
or animal models providing results that cannot easily be extrapolated to humans (SCENIHR, 2015).
The purpose of this collaborative work is to have a slightly different approach in studying long-term
exposures to ELF-EMF, to state whether biological effects are possible and plausible. It consists in
studying the human being in her/his everyday life, in her/his residential and occupational
environments, as well as conducting ex-vivo/in vitro studies. Herewith different approaches will be
linked.
The research is supported by three teams specialized in cytogenetic studies (WIV-ISP, Brussels), in
gene expression (Université Libre de Bruxelles) and in environmental health (Université Libre de
Bruxelles). These teams are accompanied by ACE team (ULiège) for technical support and ELF-
EMF in-situ measurements and modelling.
Collaborative work
A cytogenetic biomonitoring study will be performed on workers from ELIA (national Belgian
Transmission System Operator) who are regularly exposed to 50 Hz electric and magnetic fields
(line inspectors, maintenance personnel, switchyard workers, ...) and on workers who are not
particularly exposed to 50 Hz EM fields (e.g., administrative workers) and will be used as ‘non
exposed’ controls. An increased frequency of cytogenetic changes means that 50 Hz fields or other
environmental agents in the workplace are able to damage the hereditary material and that
measures should be taken to lower the exposure(s). Similar studies were already published but
contained too much shortcomings to be decisive (Maes & Verschaeve, 2016; Verschaeve, 2009;
Vijayalaxmi & Obe, 2005).
A close collaboration with the company doctor(s) and other relevant persons will help in matching
exposed and control workers.
Micronucleus and comet assays will be used as they cover different kinds of genetic effects, i.e.,

487
structural and numerical chromosome aberrations (micronucleus test) and DNA damage (comet
assay). Additional other endpoints, as for example gene amplifications and apoptosis can also be
investigated (cytome test).
Apart from our investigations on ‘electricity workers’ our project will also consist in studying the
contribution of cumulative exposure to perceived and/or objective health. This part of the study will
be carried out among the general population in the Brussels Region. Healthy participants will be
recruited according to their distance to transformers and buried cables. Measurements of exposure
to EMF and other environmental parameters such as noise, fine particles, volatile organic
compounds, indoor and outdoor pollutants... will be carried out in order to draw up a multi-exposure
map. People with environmental sensitivities will be included. Statistical analyses will be conducted
on the entire group of participants and others excluding hypersensitive people to account for their
particularities.
The collaboration between BBEMG members will enable a multidisciplinary approach to long-term
exposure to the environmental exposures studied.
Funding
The BBEMG is funded by donations from national Belgian Transmission System Operator ELIA. The
research program is controlled by a Scientific Advisory Committee. The group adheres to the
Belgian Ethical code of scientific research.
References
Maes, A., & Verschaeve, L. (2016). Genetic damage in humans exposed to extremely low-frequency
electromagnetic fields. Arch Toxicol. 90(10):2337-48.
SCENIHR (2015). Opinion on Potential health effects of exposure to electromagnetic fields (EMF).
Scientific Committee on Emerging and Newly Identified Health Risks, SCENIHR, European
Commission, DG Health and Food Safety.
Verschaeve, L. (2009). Genetic damage in subjects exposed to radiofrequency radiation. Mutat Res.
681(2-3):259-70.
Vijayalaxmi, & Obe, G. (2005). Controversial cytogenetic observations in mammalian cells exposed
to extremely low frequency electromagnetic radiation. A review and future research.
Bioelectromagnetics 26:412–430.

PA-85 [14:30]
The effect of radiofrequency electromagnetic field at 1.7 GHz LTE on human adipose-derived
stem cells and liver cancer stem cells
Kyeongrae Min1, Nam Kim2, Hyung-Do Choi3, Yun-Sil Lee4 & Kiwon Song1
1Department of Biochemistry, Yonsei University, Seoul, Korea, 03722
2School of Electrical and Computer Engineering, Chungbuk National University, Cheongju, Korea
3Department of EMF Research Team, Electronics and Telecommunications Research Institute , Daejon,
Korea
4Graduate School of Pharmaceutical Sciences, Ewha Woman’s University, Seoul, Korea
Keywords: In vitro, Static, Work in Progress
Presented by: Kiwon Song
Due to the technology of mobile phone, we are daily exposed to LTE RF-EMF, but its biological
effects have not been clarified. In particular, in spite of the physiological importance of adult stem
cells and cancer stem cells, their response to LTE RF-EMF is not much studied. In this study, we
investigated the cellular effect of 1.7 GHz RF-EMF (LTE) on adult and cancer stem cells. When
continuously exposed to 1.7 GHz RF-EMF at 1 W/kg for 72 h, the proliferation of both adipose stem
cells (ASC) and cancer stem cells of Huh7 was decreased: 8% in ASC and 30% in cancer stem

488
cells of Huh7 compared to unexposed controls. These results suggest that the cancer stem cells are
highly more sensitive to 1.7 GHz RF-EMF than normal adult stem cells.
Introduction
With improving technologies and affordability of cell phone, we are continuously exposed to LTE RF-
EMF and its potential effects at low-intensity radiation levels on biological systems are of great
interest. However, not much data have been accumulated to reach the consensus of its effects. In
particular, various adult stem cells play key functions for physiology, but only a few studies reported
the effect of LTE range RF-EMF on adult stem cells. Glaser et al. showed that a short (4 h) and a
long (20 h) exposure of UMTS 1950 MHz and LTE 2535 MHz to human hematopoietic stem did not
induce noticeable effect [1]. Exposure to 1.8 GHz RF-EMF was reported to impair neurite outgrowth
of embryonic neural stem cells [2, 3]. There has been no study for the effect of LTE range RF-EMF
on cancer stem cells, which are regarded as a cause for cancer development, survival, and
recurrence [4]. In this study, we examined the effect of 1.7 GHz LTE RF-EMF on the adult stem cell
ASC and the CSC of hepatocellular carcinoma Huh7.
Methods
ASC was purchased from Thermo Fisher Scientific and a stem cell population of Huh 7 was
provided by Prof. Y. N. Park (Dept. of Pathology at Yonsei University College of Medicine). These
cells were verified with the expression of stem cell markers by reverse transcription-polymerase
chain reaction (RT-PCR): positive cell markers (CD44, CD105) and a negative marker (CD45) for
ASC and CD133, EpCAM, and ALDH1A for Huh7 were used. ASC of 300,000 and Huh7 CSC of
200,000 were seeded, incubated for 12 h, and exposed to 1.7 GHz at 1 W/kg for 72 h. During the
exposure to RF-EMF, the temperature was maintained within a range of 35.5 ± 0.5 °C by circulating
water within the cavity. Before and after the exposure to RF-EMF, cells were counted by Cellometer
Auto T4 (Nexcelom).
Results
We observed that cell number was decreased on average 8% in the exposed ASCs and about 30%
in the exposed Huh7 CSCs, compared with that of each unexposed control.
These results demonstrate the anti-proliferative effect of a LTE 1.7 GHz on the ASCs and Huh7
CSCs by the continuous exposure with 1 SAR. These observations also suggest the different
physiological effects of 1.7 GHz RF-EMF on adult stem cells and cancer stem cells.
Discussion and Conclusion
In this study, we observed the anti-proliferative effect of a LTE range 1.7 GHz RF-EMF on the adult
stem cell ASCs and the cancer stem cell population of hepatocellular carcinoma Huh7. We also
observed that the cell number was decreased about 30% in the RF-EMF exposed CSCs of Huh7
compared with the unexposed control, while that was decreased about 8% in the RF-EMF exposed
ASCs. These observations strongly suggest that CSCs are more sensitive to RF-EMF than ASCs. In
order to confirm the different effects of 1.7 GHz RF-EMF on adult stem cells and cancer stem cells,
we are examining the effect of 1.7 GHz RF-EMF on normal fibroblast IMR-90 cells, another
hepatocellular carcinoma CSCs of Hep3B, and various cancer cell lines.
References
1. K. Gläser. et al., Effect of Radiofrequency Radiation on Human Hematopoietic Stem Cells.
Radiation Research 2016 186(5):455-465.
2. Chen, C. H. et al., Exposure to 1800 MHz radiofrequency radiation impairs neurite outgrowth of
embryonic neural stem cells. Sci Rep. 2014 4:5103.
3. Xu, F. et al., Age-dependent acute interference with stem and progenitor cell proliferation in the
hippocampus after exposure to 1800 MHz electromagnetic radiation. Electromagn Biol Med.
2017 36(2):158-166
4. Lerchl A. et al., Tumor promotion by exposure to radiofrequency electromagnetic fields below
exposure limits for humans. Biochem Biophys Res Commun.2015 459(4):585-90.

489
Figures

Figure 1. Expression of cell markers for stemness was verified in Huh7 CSCs and
ASCs.Before ASCs and Huh7 CSCs were exposed to 1.7 GHz RF-EMF, the expression of cell
markers for stemness was verified by RT-PCR. ASCs expressed positive cell markers (CD44,
CD105) and a negative marker (CD45). FABP4 used as an ASC differentiation marker was not
expressed. Huh7 expressed cancer stemness markers (CD133, EpCAM) and a hepatocyte
marker (ALDH1A), when human actin was used as a positive control.

Figure 2. A continuous exposure of Huh7 CSC and ASC to 1.7 GHz RF-EMF decreases their
proliferations.The same number of seeded ASCs and Huh7 CSCs was incubated for 12 h and
exposed to 1.7 GHz at 1 W/kg for 72 h. Before and after the exposure, cell numbers were
counted and plotted with each unexposed control. Same experiments were performed 9 times
and the average was plotted with standard deviations. The results were represented as mean 
± SEM; N = 3. P<0.001 (***) P > 0.05(n.s).

PA-87 [14:30]
Development of intermediate frequency band exposure apparatus and its effect on cultured
cells
Eijiro Narita1, Shin Koyama1, Yoko Shimizu1, Naoki Shinohara1 & Junji Miyakoshi1

490
1Laboratoryof Applied Radio Engineering for Humanosphere, Kyoto University, Uji, Japan, 611-0011
Keywords: In vitro, IF, Work in Progress
Presented by: Eijiro Narita
We developed an exposure apparatus that generates an intermediate frequency at 85 kHz. We
carry out various cell experiments in pursuit of the possibility that the intermediate frequency exerts
a non-thermal effect. We examined the cell proliferation test, the colony forming ability test, and the
cell cycle distribution test using HCE-T cells, which is derived from human cornea. We confirmed
that this device retains a normal culture environment without any artefact. Exhaustive gene
expression analysis of human cells will be carried out by standard cell experiment method widely
used as safety evaluation in future.
Introduction
No clear data has been presented on its effect on the human body by the intermediate frequency
EMF waves due to wireless power transmission etc., which is expected to spread globally in the
near future. Many safety tests have been carried out on very low frequencies in the commercial
frequency band and on high frequencies used in mobile phones. Some studies have reported that
very low and high frequencies EMF exposure had a significant effect on human cells, although this
view has not been confirmed by other studies. There are extremely small amount of data on the
intermediate frequency band, which is expected to be rapidly spreading in countries across the
world, and urgent response to research on safety to the human body is required. In particular,
various possibilities have been pointed out for non-thermal effect on human health. Some research
reports show positive effects, but there is not a single report recognized as a reliable scientific basis.
We therefore developed an exposure apparatus that generates an intermediate frequency band,
and we will carry out various cell experiments in pursuit of the possibility that the intermediate
frequency exerts a non-thermal effect. We first developed a device that can be exposed to cells, and
confirmed that this device retains a normal culture environment without any artefact.
Materials and Methods
We used specially designed exposure apparatus (Figure 1) that employs a wireless power supply
system (D-Broad Slim, DAIHEN, Osaka, Japan) for Automatic Guided Vehicles (AGVs). A
uniform magnetic field of an intermediate frequency at 85 kHz was generated from the power
transmission unit of the wireless power supply system and a magnetic field of 65.1 A/m (81.8 μT)
was observed on the power receiving side when the input power was set at 360 W. The power
transmitting and receiving unit were placed inside an incubator, and arranged so that the culture
dishes could be fixed between them.
We examined the cell proliferation test, the colony forming ability test, and the cell cycle distribution
test using HCE-T cells, which is derived from human cornea, to confirm that this exposure apparatus
holds a normal culture environment without generating any unwanted artefact. HCE-T cells were
seeded using DMEM and HamF12 medium (1:1) supplemented with 5% fetal bovine serum, 5 μg/ml
insulin, 10 ng/ml hEGF, and cultured in an exposure apparatus for 24 hours. For the cell proliferation
test, 500 to 1000 cells were seeded in one well after cell collection, and the number of cells was
counted automatically at fixed time intervals using a cell analyser (RTCA DP, Roche, Basel,
Switzerland). 100 to 500 cells were seeded in one dish for the colony forming ability test, and
ethanol fixation and Giemsa staining were performed after 7 to 10 days to count the number of
colonies. For cell cycle distribution test, cells were collected 48 hours later and the cell cycle was
measured using a flow cytometer (FACS Calibur, BECTON DICKINSON, NJ, USA).
Results and Discussion
Results of cell proliferation test, colony forming ability test, and cell cycle distribution using HCE-T
cells are shown in Figure 2. In these three cell kinetic tests, no significant difference was found
between this newly developed exposure apparatus and a conventional incubator. In order to verify
the influence on cells by magnetic field, the coil of the power transmission unit will be modified so as

491
to generate a uniform magnetic field to a wider range. We then plan to analyse exhaustive gene
expression of human cells by standard cell experiment method widely used as safety evaluation.
Conclusion
As a result of examining cell proliferation test, colony forming ability test, and cell cycle distribution
using HCE-T cells, it was confirmed that our newly developed intermediate frequency exposure
apparatus of 85 kHz retained the normal culture environment.
Acknowledgement
This work has supported in part by a Grant from the Ministry of Internal Affairs and Communications,
Japan, and the Research Institute for Sustainable Humanosphere, Kyoto University (Mission 5-1)
References
[1] Zhang MB, He JL, Jin LF, et al. Study of low-intensity 2450-MHz microwave exposure enhancing
the genotoxic effects of mitomycin C using micronucleus test and comet assay in vitro. Biomed
Environ Sci. 2002;15:283–90.
[2] Tice RR, Hook GG, Donner M, et al. Genotoxicity of radiofrequency signals. I. Investigation of
DNA damage and micronuclei induction in cultured human blood cells. Bioelectromagnetics.
2002;23:113–26.
[3] Karkabounas S., Havelas K., Kostoula O.K., Vezyraki P., Avdikos A., Binolis J., Hatziavazis G.,
Metsios A., Verginadis I., Evangelou A., et al. Effects of low intensity static electromagnetic
radiofrequency fields on leiomyosarcoma and smooth muscle cell lines. Hell. J. Nucl. Med.
2006;9:167–172.
Figures

Figure 1. Newly developed exposure system that generates an intermediate frequency at 85

kHz.

492
 Figure 2. Results of cellular kinetics test using HCE-T cells (a) cell proliferation test, (b) colony
forming ability test, (c) cell cycle distribution.

PA-89 [14:30]
Effects of ultraviolet radiation on full thickness human skin model in vitro (SKIN-RF project)
Zsuzsanna Nemeth1, Györgyi Kubinyi1, Jozsef Bakos1, Brahim Selmaoui2 & György Thuroczy1
1Department of Non-Ionizing Radiation, National Public Health Institute, Budapest, Hungary, 1221
2Unité de Toxicologie Expérimentale, L'Institut National de l'Environnement Industriel et des Risques
(INERIS), Verneuil en Halate, France, 60550
Keywords: In vitro, Optical, Work in Progress
Presented by: Zsuzsanna Nemeth
The aim of this study was to examine whether the solar ultraviolet (UV) radiation has any effect on
cytokine (IL-1α, IL-6, IL-8) and MMP-1 enzyme secretion on human skin model in vitro. The full
thickness skin models were exposed by a sun simulator lamp to UV radiation (1.5 SED, 3 SED). The
IL-1α, IL-6, IL-8 and MMP-1 concentrations were measured 24 hours after UV exposure from the
culture medium by the ELISA method. In this study we managed to show that UV radiation doses of

493
1.5 SED, 3 SED did not alter the level of these factors. This experiment was done as a preliminary
study for the project named “Cellular response to co-exposure of radiofrequency (RF) and solar
ultraviolet (UV) radiation in human in vitro skin model (SKIN-RF)”.
Introduction
Ultraviolet (UV) radiation is responsible for a wide variety of different acute and chronic effects on
the skin. Acute responses of human skin to UV radiation include photodamage, erythema, mutation,
immunosuppression, synthesis of vitamin D and tanning. Chronic UV radiation effects include
photoaging and photocarcinogenesis (Ullrich, 2002).
UVB (280-315 nm) and UVA (315-400 nm) exposure from the sun and artificial sources (sun beds)
are the most important etiological factors for development of skin cancer. UVB and UVA radiation
has been classified as Class I carcinogen by the International Agency for the Research on Cancer
(IARC, 2012).
Interleukin IL-1α is a primary cytokine in keratinocytes capable of initiating cutaneous inflammation
via secondary cytokines like IL-6 and IL-8. The UV-induced matrix metalloproteinase-1 (MMP-1) is a
crucial biomarker of photoaging. Interleukin IL-1α and IL-6 from UVB-irradiated keratinocytes can
regulate MMP-1 expression in fibroblasts through paracrine effects.
The aim of this experiment was to evaluate the effects of exposure to UV radiation on reconstructed
human skin models regarding inflammation and photoaging. We used the ELISA method to
determine the concentration of the cytokines (IL-1α, IL-6 and IL-8) and the MMP-1 enzyme from the
cell culture supernatant.
This experiment was done as a preliminary study for the project named “Cellular response to co-
exposure of radiofrequency (RF) and solar ultraviolet (UV) radiation in human in vitro skin model
(SKIN-RF)” funded by ANSES.
Methods
In this experiment MatTek EpiDerm Full Thickness (EPI-300) reconstructed skin models were used.
After the shipment skin models set on inserts were removed from the agarose matrix, placed in a
6-well plate containing 1 ml medium and equilibrated at 37 °C, 5% CO2 for one hour. After that
tissues were maintained in culture until experiments with UV radiation. Three tissues were prepared
for each test condition: sham exposed as negative control, 5 % SDS (sodium dodecyl sulphate) as
positive control and 2 doses of UV exposure.
For UV radiation skin tissues were transferred to Petri dishes filled with 1 ml DPBS and exposed to
doses of 1.5 SED and 3 SED by a solar simulator lamp (SOL 500, Hönle, Germany). Standard
Erythema Dose (SED) is equivalent to an effective erythemal exposure of 100 J/m2. The distance of
the lamp from the tissues was 61 cm, the duration of the exposure 30 and 60 minutes for 1.5 and 3
SED, respectively. The UV doses were checked by the International Light ILT-900 NIST calibrated
spectroradiometer. The temperature of the DPBS in the Petri dishes was monitored through the
whole duration of the UV exposure and it never exceeded 37 °C.
After the UV radiation tissues were incubated at 37 °C and 5 % concentration of CO2 for 24 hours to
investigate the post-exposure response. Following the incubation the cell culture medium was
collected and frozen until analysis and the viability of the full thickness skin tissues were evaluated
with the MTT test. The concentration of IL-1α, IL-6 and IL-8 and MMP-1 was measured from the
collected cell cultured medium with ELISA method.
Results
The viability of skin tissues was determined 24 hours after exposures to UV radiation or to 5 % SDS.
We found that the 5 % SDS significantly decreased the viability of the skin models tissues to 41.7 %
compared to negative control. The viability of the skin tissues that were exposed to UV radiation
maintained at control level.
The concentration of the IL-1α in the positive control group significantly increased. The exposure to

494
both UV doses had not changed the concentration of the studied markers significantly.
Conclusions
In this experiment the influence of the UV exposure on Mattek EPI-300 full thickness skin model was
studied. The concentration of the cytokines and the MMP-1 enzyme from the cell culture medium
were measured with the ELISA method 24 hours after UV exposure. In this preliminary study we
managed to show that the UV dose of 1.5 SED and 3 SED did not have significant effects on
MatTek EPI-300 reconstructed full thickness skin models compared to negative control.
References
Ullrich SE (2002) Photoimmune suppression and photocarcinogenesis. Front Biosci.
2002;7:d684–703.
IARC (2012) Monographs on the Evaluation of Carcinogenic Risks to Humans. Solar and ultraviolet
radiation. IARC Monogr Eval Carcinog Risks Hum, 100D: 35–101

PA-91 [14:30]
STUDENT PAPER
Local thermal dosimetry applied to in vitro studies at millimetre waves
Rosa Orlacchio1, Maxim Zhadobov1, Stanislav Alekseev2, Ronan Sauleau1, Yann Le Page3 & Yves Le
Dréan3
1Institut d'Électronique et de Télécommunications de Rennes (IETR), University of Rennes 1, Rennes,
France, 35042
2Institute of Cell Biophysics, Russian Academy of Sciences, Pushchino, Russian Federation, 142290
3Institute for Research on Environmental and Occupational Health (IRSET), University of Rennes 1, Rennes,
France, 35042
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Rosa Orlacchio
This study investigates the role of thermal convection arising from high SAR gradients during in vitro
exposure to millimetre waves (MMW). The impact of the liquid volume and exposure duration on
initiation of convective currents is investigated. Water and a convection-free medium with water-
equivalent electromagnetic properties have been exposed to continuous wave (CW) and amplitude
modulated MMW. Continuous and pulsed heating were recorded using a microthermocouple. Our
results suggest that convection plays an important role in MMW-induced heating in in vitro
experiments, and its initiation depends both on the volume of the exposed liquid and the duration of
exposure.
Introduction
The possibility to exploit the 20-100 GHz range to selectively focus heat within skin was
demonstrated in [1], suggesting that it can be used for non-invasive thermal treatment of skin
cancers such as melanoma. In in vitro experiments, MMW can be efficiently used to heat cell
monolayers with a typical thickness of 3-10 µm. Heating can be delivered in a continuous way or as
thermal MMW-induced pulses. Accurate control of the local temperature distribution inside the
irradiated liquid sample is of outmost importance and directly impacts the thermal dose [2], which
increases exponentially with temperature. Methodology for local temperature measurements using
microthermocouples was provided in [3], although no information was given about the effect of
convection. The latter plays an important role at MMW [4]: shallow penetration creates great
temperature gradients inside the liquid, which generate complex convection currents that influence
the temperature rise and therefore the response of cells attached to the bottom of the cell culture
plate. The aim of this study is to extend the knowledge about in vitro thermal dosimetry at MMW by
analysing the role of thermal convection effect during continuous or pulsed (PW) MMW-induced
heating.

495
Methods
The exposure system used in this study consists of two main units: a) a signal generator sub-unit
and b) an exposure chamber described in detail in [3]. Deionized water and an agar phantom (water
mixed with 0.5% agar) were exposed from the bottom. Agar phantom is used in this study to prevent
convection. Except viscosity, both electromagnetic [5] and thermophysical parameters [6] of the agar
phantom closely match those of pure water. Samples were placed into one well (Fig. 1) of a 12-well
culture plate (353072, Microtest 96, Becton Dickinson, Franklin Lakes, NJ); volumes used range
from 0.5 to 4 ml corresponding to 1.3 to 10.43 mm height, respectively. The finite integration
technique (FIT) solver of CST Microwave Studio was used for the numerical analysis. The numerical
thermal model detailed in [3] was used to retrieve SAR from local thermal measurements. Thermal
pulses were induced by exposing the sample to amplitude modulated CW MMW with a pulse
repetition rate of 0.05 pps at the peak generator power (up to 4.2 W), while the pulse duration was
changed between 1.5 and 6 seconds through a programmable power supply HMP (Hameg
Instruments, Hampshire, UK). Local temperature monitoring was performed using a K type
thermocouple (TC) probe with the lead diameter of 75 µm.
Results
Numerical results
Numerical simulations show high SAR gradients inside the exposed sample. SAR value rapidly
decreases by moving from the antenna beam axis (maximum value) toward the well edges; for
instance at 4 mm away from the antenna axis, along the E plane, SAR is about 50% of its maximum
value. The thermal model described in [3] was used to retrieve SAR values in water and agar
phantom. Heating recorded in the initial exposure phase when the effect of convection is negligible
was fitted with the above-mentioned thermal model. Peak SAR value of water obtained with this
model is 18% less than value obtain in simulations. Taking into account 20% underestimation of
temperature measurements with 75 µm TC [3] our experimental data are in excellent agreement
with computations (only 2% difference). Note that there is not a statistically significant difference
between SAR values for water and agar phantom (P=0.74).
Temperature measurements
Fig. 2 shows the impact of the liquid volume on convection initiation during 30 s exposure to CW at
195 mW. Convection is triggered when water volume is above 1.5 ml. In 2 ml of liquid, convection
appears about 24 s after the beginning of exposure, while at higher volumes (3 and 4 ml) it starts
earlier after about 14 s. The role of the pulse duration was analysed as well. Initiation of convection
was investigated for 10 pulses with a duration ranging from 1.5 and 6 s with the same period (20 s).
Convection currents appear when the duration exceeds 3 s affecting also the shape of the pulse,
which becomes less sharp with increasing pulse duration (data not shown; complete results will be
presented during the conference).
Conclusion
In this study we investigated the role of convection generated during continuous and pulsed heating
induced by MMW at 58.4 GHz. Our results demonstrate that initiation of convection depends on (1)
the liquid volume inside the exposed well and (2) the duration of the thermal pulse.
References
[1] M. Zhadobov, S. I. Alekseev, Y. Le Dréan, R. Sauleau and E. Fesenko, “Millimeter waves as a
source of selective heating of skin,” Bioelectromagnetics, vol. 36, n°16, pp. 464-475, Sep.2015.
[2] S. A. Sapareto and W. C. Dewey, “Thermal dose determination in cancer therapy,” Int J Radiat
Oncol BiolPhys, vol. 10, pp. 787-800, 1984.
[3] M. Zhadobov, S. I. Alekseev, R. Sauleau, Y. Le Dréan, Y. Le Page and E. E. Fesenko,
”Microscale temperature and SAR measurements in cell monolayer models exposed to millimeter
waves,” Bioelectromagnetics, vol. 38, p. 11–21, 2017.
[4] E. P. Khizhnyak et M. C. Ziskin, “Temperature oscillation in liquid media caused by continuous
496
(nonmodulated) millimiter wavelenght electromagnetic irradiation,” Bioelectromagnetics, vol. 17, pp.
223-229, 1996.
[5] N. Chahat, M. Zhadobov and R. Sauleau, "Broadband tissue-equivalent phantom for BAN
applications at millimeter waves," in IEEE Transactions on Microwave Theory and Techniques, vol.
60, n°7, pp. 2259-2266, May 16 2012.
[6] F. Erdogdu, “Optimaization in food engineering,”CRC Press, Taylor & Francis Group, 2008
Figures

Figure 1. Exposure scenario: one-well of a 12-well tissue culture plate, with a TC attached to
the bottom, set at 5 mm from a rectangular WG.

497
 Figure 2. Impact of the sample volume on convection initiation during 30 s of exposure (58.4
GHz, 195 mW) in water (black line) and agar phantom (gray line).

PA-93 [14:30]
Radiofrequency radiation (RFR) induces Bad, p-Bad activation leading to cell apoptosis on
DLD1 human colon carcinoma cell lines
Elcin Ozgur1, Fatih Senturk1, Gorkem Kismali2, Tevhide Sel2 & Goknur Guler1
1Biophysics Department, Gazi University, Ankara, Turkey, 06500
2Biochemistry Department, Ankara University, Ankara, Turkey
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Fatih Senturk
In this study, radiofrequency radiation (RFR) effects investigated on Bad and p-Bad the pro-
apoptotic member of the Bcl-2 family inhuman colon cancer cell, DLD1.We evaluated the effects of
intermittent RFR exposure (digitally modulated 3GPP signals at 900, 1800, 2100 MHz) for 1 hr and
4 hrs (15 min on, 15 min off). Bad, p-Bad expressions determined by ELISA increased due to 900
and 2100 MHz RFR exposure for 1h. However, 4 hours of RFR exposure lead to decrease in all
parameters (p<0,05 and less). Overexpression of Bad has been shown to induce apoptosis, so it is
concluded that intermittent exposure of RFR might induce apoptosis by Bad and p-Bad expression
on DLD-1 cancer cells.
Introduction
Colon cancer is one of the most common malignancies in developed countries due to environmental
factors. In this study, DLD1, one of the human colon cancer cell line was used to investigate
radiofrequency radiation effects on Bad and p(phospho)-Bad which is the pro-apoptotic member of
the Bcl-2 family.
Method

498
We evaluated the effects of intermittent mobile phone like signal (digitally modulated 3GPP signals
at 900, 1800, 2100 MHz)during 1 h and 4 hrs(15 min on, 15 min off)on DLD 1 cells. RFR exposure
system is composed of vector signal generator and a horn antenna in a temperature-controlled
shielded room. Elisa assay was used to determine Bad, p-Bad expression levels. Experiments were
performed in triplicates. Statistical analysis was performed by using GraphPad Prism 7 software.
P<0,05 and less were considered statistically significant.
Results
Bad, p-Bad expression increased due to 900 and 2100 MHz exposure of radiofrequency radiation
for 1h. However, 4 hours of RFR exposure lead to decrease in all parameters(p<0,05 and less).
Conclusion
Overexpression of Bad which is Bcl-2 family proteins growing family of apoptosis regulators has
been shown to induce apoptosis. Since exposed level of RFR in this experiment caused to increase
Bad levels in DLD1 cells, it might be concluded that intermittent exposure of RFR might induce
apoptosis by Bad and p-Bad expression on DLD-1 cancer cells.
Acknowledgements
The author(s) would like to acknowledge the networking support by COST Action BM1309 (COST
EMF-MED).

PA-95 [14:30]
Evaluation of biological effects induced by electromagnetic radiation waves
Flavien Pillet1, Jelena Kolosnjaj-Tabi1, Rene Vezinet2 & Marie-Pierre Rols1
1IPBS-CNRS, Toulouse, France, 31077
2CEA-Gramat, CEA, Gramat, France, 46500
Keywords: In vitro, Pulsed, Completed (unpublished)
Presented by: Marie-Pierre Rols
We developed a set of micro applicators to study the potential effect of electromagnetic fields on
biological samples. The applicators allow a real-time observation.Different signals were used:
electromagnetic waves and square-wave monopolar electric field pulses causing electroporation.
The biological effects were assessed on giant unilamellar vesicles and mammalian cells, and
compared to results obtained at full scale with signals generated by antennas.
The effects of electromagnetic radiation waves on health is one of the major public concern. These
waves are mainly produced at a large scale but it is important to evaluate these effects on biological
samples at the laboratory scale. Here we developed a set of micro applicators, which allow
evaluating the effect of electromagnetic fields on biological samples with volumes in the microliter
range. The applicators can be coupled to an optical microscope and allow a real-time observation of
potential structural and functional alterations of the tested sample induced by different waveforms.
New design approaches are suggested to simultaneously achieve maximized electric field coupling
effect and optimized electric field homogeneity in the tested sample, while minimizing the return loss
when the applicators are loaded with the biological samples. These applicators allow studying the
biological effect of a variety of different signals, due to their wide frequency bandwidth (beyond 1.5
GHz) and their high permissible power. In addition, different electromagnetic parameters such as the
electromagnetic field magnitude, pulse repetitive factor, number of bursts or delay between bursts
may be set. The efficacy of the applicators was addressed for three different signals: two types of
electromagnetic waves - a damped sinusoid centered at 200 MHz (wideband signal), a radar-like
signal at 1.5 GHz (the ultra-narrow band signal) and a train of millisecond square-wave monopolar
electric field pulses (causing electroporation). The biological effects were thus assessed (at the
microscopic scale) on two different biological models, the giant unilamellar vesicles, and tumor and
normal human cells, as well as compared to results obtained (at full scale) with signals generated by

499
antennas.

PA-97 [14:30]
50 Hz MF exposure has no significant DNA damage effect on ATM deficient MEFs
Yumin Jin1, Chuan Sun1, Xiaoxia Wei1, Aziguli Yimaer1, Zhengping Xu1 & Guangdi Chen1
1Bioelectromagnetics lab , Zhejiang University School of Medicine, Hangzhou, China, 310058
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: Yumin Jin
The genotoxicity of ELF-ME exposure has been widely studied, however, the results are
inconsistent. In our study, we investigated the effect of 2.0 mT 50 Hz MF exposure on a DNA repair
factor (ATM) proficient and deficient MEFs. The results showed that, in either wide type or ATM
deficient MEFs, 50 Hz MF exposure did not significantly induce DNA damage, and could not affect
cell apoptosis and cell viability, suggested that the 50 Hz MF exposure enhance DNA damage in
ATM deficient MEFs.
With the rapid increasing application of electric power, concerns about potential health hazards
caused by extremely low frequency magnetic fields (ELF-MF) have been raised. Based on limited
epidemiological studies, IARC classifies ELF-MF are possible carcinogens to humans, however, the
experimental evidence is lacking. The basic experimental evidence, the genotoxicity of ELF-MF, has
been investigated in hundreds of studies, but the results are inconsistent. For further exploration on
the genotoxicity of EMF-MF exposure, we used a DNA repair deficient cellular model, Ataxia
telangiectasia mutated (ATM) gene deficient mouse embryonic fibroblasts (MEFs), to investigate
whether 50 Hz MF exposure is genotoxic to cells when DNA repair is disabled.
Methods
Sub-cultured Atm+/+ and Atm−/− MEF cells were exposed or sham exposed to 2.0 mT 50 Hz MF for
up to 24 h. Immediately after exposure, gamaH2AX foci formation assay and alkaline comet assay
were used to detect DNA damage, and flow cytometry and CCK8 analysis were used to detect cell
cycle and viability.
Results
The effect of 50 Hz exposure on DNA damage
Exposure to 50 Hz MF for 15 min, 1 or 24 h had no significant effect on γH2AX foci formation in
either Atm+/+ or Atm−/− MEFs (Fig. A), and exposure to 50 Hz MF for 1 or 24 h had no significant
effect on DNA fragmentation in both MEFs (Fig. B). The results suggested that 50 Hz MF exposure
do not induce DNA damage in either Atm+/+ or Atm−/− MEFs.
The effect of 50 Hz exposure on cell behavior
Exposure to 50 Hz MF for 1 or 24 h did not affect cell viability in either Atm+/+ or Atm−/− MEFs (Fig.
C), and exposure to 50 Hz MF for 24 h had no significant effect on cell cycle progression (Fig. D).
The results suggested that 50 Hz MF exposure has no significant effect on cell behavior in both
MEFs.
Conclusion
Under current experimental conditions, we did not find the genotoxicity of 50 Hz MF exposure, and
the deficiency of Atm did not result in cellular sensitivity to 50 Hz MF exposure.
Figures

500
Figure 1. (A) DNA damage detected by γH2AX foci formation assay, the average number of
γH2AX foci per cell were showed; (B) DNA fragmentation detected by alkaline comet assay,
the olive tail moment and percentage of tail DNA were showed; (C) cell viability detected by
CCK8 analysis, the relative cell viability were showed; (D) cell cycle progression detected by
follow cytometry, the percentage of cell population were showed. MEFs treated with 0.1 mM
4NQO were positive control.

501
PA-99 [14:30]
STUDENT PAPER
Effect of high frequency electric field on growth factor of cultured osteoblasts
Shunsuke Yamaguchi1, Masatake Akutagawa1, Hiromichi Yumoto2, Kouji Hirao2, Takahiro Emoto1, Hiroo
Tarao3, Toshihiko Tominaga4, Toshitaka Ikehara5 & Yohsuke Kinouchi1
1Department of Electrical and Electronic Engineering, Tokushima University , Tokushima, Japan, 770-8506
2Department of Periodontology and Endodontology, Tokushima University , Tokushima, Japan, 770-0042
3Department of Electrical & Computer Eng., National institute of technology Kagawa colllege, Takamatsu,
Japan, 761-8058
4Tominaga Dental Clinic, Naruto, Japan, 771-0360
5Institute for Health Sciences, Tokushima Bunri University, Tokushima, Japan, 770-8514
Keywords: In vitro, IF, Work in Progress
Presented by: Shunsuke Yamaguchi
In this experiment, high frequency AC voltage was applied to cultured osteoblasts. The voltage
waveform used is a sinusoidal wave of 500 kHz 20 V. Temperature, voltage, osteoblastic VEGF,
VEGFmRNA and ALP were measured. VEGF and VEGFmRNA were measured 1 day after voltage
application and ALP was measured 7 days after. Durations of voltage application were 30, 60, 90
and 120 s. As results of them, it is obvious that it affects the osteogenesis of osteoblasts by
electrical factors. Although it depends on the initial state (cycle) of the cells, VEGF was significantly
increased at 90 seconds and VEGFmRNA at 120 seconds. Change of ALP by application of the
elecrical field was not observed in our experiment.
Introduction
As caries progress, the roots of the teeth will melt and cause apical periodontitis in severe cases.
As a conventional treatment method, a root canal cleaning using a disinfectant is performed.
However, the tip of the root canal is very thin, so bacteria may remain (refractory apical
periodontitis). EMAT (Electro-Magnetic Apical Treatment) is newly proposed therapy for the apical
periodontitis. In this therapy, the electric fields is applied by electrodes in the root canal and corner
of the mouth. Sterilization and alveolar bone regeneration by applied voltage are obtained by clinical
findings. It is thought that alveolar bone regeneration occurs by activating osteoblasts around the
apical lesion.
The authors investigate the changes growth factor of osteoblasts due to voltage application.
Materials and Method
Osteoblasts used in the experiment
In this experiment, we focus on cultured osteoblasts. Osteoblast (MC3T3-E1) cells are seeded on
the cell insert of the 6-well plate and experimented as soon as it becomes subconfluent. Regarding
the amount of culture solution is 3.0 ml (between the cell insert and 6-well plate) and 1.5 ml (inside
the cell insert). After the end of the applied voltage, culture under the same conditions. For the
medium of osteoblast, two kinds of medium are used. One is a medium containing α-MEM + 10%
FBS, and thereafter a normal medium. The other is added to a normal medium and is a medium
containing β-glycerophosphate (10 mM) + ascorbic acid (50 μg / mL), hereinafter referred to as
calcification induction medium.
Experiment procedure
In this experiment a sinusoidal AC voltage of 500 kHz using an oscillator (WF1946A/NF
Corporation) is applied and the change of the object to be measured is observed. Four experiments
are done. In each experiment, the voltage application time is different. The experimental circuit is as
follows. The AC power supply composed of the oscillator and the amplifier (HSA4011/NF

502
Corporation) is connected to the electrode plate via the cement resistance (10Ω). Measure the
power supply voltage and the electrode plate voltage with an oscilloscope. Replace the culture
medium with PBS, when voltage is applyed. The liquid volume was placed in order of 3ml, 2.5ml.
Finally apply a voltage after inserting the thermocouple into the PBS on the 6-well plate side. During
the experiment, temperature and voltage measurement are performed. After completion of voltage
application, cultivate osteoblast in an incubator under the same conditions as before the experiment.
Measurement item
VEGF(Vascular Endothelial Growth Factor), VEGFmRNA
During bone formation, angiogenesis is occurring. Based on the VEGF that plays a role in
angiogenesis, it was considered as promoting osteogenesis of osteoblasts and listed as a
measurement item.
ALP(Alkaline Phosphatase)
Osteoblasts differentiate into bone cells. According to the literature, although the subject is a rat
osteoblast, there are three periods until osteoblasts differentiate into bone cells. When the
proliferation phase is over and calcification occurs, ALP is one of the most secreted enzymes.
Measuring method
VEGF
The osteoblasts (MC3T3-E1) were seeded at a cell insert concentration of 5.0 × 104 cells / ml
placed in a 6-well plate and cultured until subconfluent, followed by applied voltage. After 24 hours
of applied voltage , culture supernatants in the cell insert and outside the cell insert were recovered,
respectively. Mouse VEGF concentration in the culture supernatant was measured with a microplate
reader using a commercially available ELISA kit (Duo Set ELISA Development Systems, R & D
SYSTEMS) at a wavelength of 450 nm.
VEGFmRNA
Osteoblasts (MC3T3-E1) were seeded at a cell insert concentration of 5.0× 104 cells / ml placed in a
6-well plate, cultured until subconfluent, and then voltage was applied 24 hours later NucleoSpin
RNA (MACHEREY- NAGEL), and cDNA was synthesized by reverse transcription using 20 ng of
PrimeScrip RT Master Mix (TaKaRa) specific primers of VEGF,↑ -actin were obtained from
RocheUniversal Probe Library Assay Design Center Designed using ProbeFiner software, specific
primers were added to each synthesized cDNA, and using Fast SYBRGreen Mas-ter Mix (Thermo
Fisher Scientific) and StepOne Plus Real-Time PCR System (Thermo Fisher Scientific) Real-time
Reverse Transcription-Polymerase Chain Reaction (RT-PCR) was performed at an annealing
temperature of 60 ° C, and the mRNA expression level of the housekeeping gene ↑-actin was used
as a reference , And the mRNA expression level of VEGF was quantified by the ΔΔCt method.
ALP
Osteoblasts (MC3T3-E1) are seeded at a concentration of 5.0× 104 cells / ml placed in a 6-well
plate, cultured until subconfluent, then changed to calcification induction medium, One week later,
the cells were collected with 0.05% TriotnX-100 and the intracellular alkaline phosphatase was
measured for absorbance at 405 nm using a laboratory assay ALP kit (WAKO) with a microplate
reader (iMark microplate reader, Bio-Rad ).
Method of calculating electric field etc using voltage measurement value
Measure the power supply and inter-electrode voltage with an oscilloscope. Since the voltage of the
cement resistance can be obtained, the current value is calculated. The current flows across the
membrameof the cell insert diameter of 23.1mm, calculate current density. Also an electric field is
obtained by using the distance between electrode plates (000mm). Also the energy given to the
PBS is calculated from the measured value.
Results
The measurement results of VEGF, VEGF mRNA, ALP in each experiment are shown in Fig. 2 and

503
3.
Discussion
Temperature measurement result
When 120 seconds was applied, the initial temperature was 20 degrees and the final temperature
was about 35 degrees. The optimum temperature of osteoblasts is near body temperature (37
degrees). Since temperature rise has not reached 37 degrees, activation of osteoblasts by
temperature is not considered to have occurred.
VEGF, VEGF mRNA measurement result
Experiment 1
The VEGF in the cell insert does not have a significant difference, but it seems that it decreases as
the voltage application time increases. However outside the cell insert, it increased at 60s (p <0.07).
Inside and outside the cell insert, the inside showed high numerical values. However, this tendency
was not obtained. VEGFmRNA significantly increased at 120s.
Experiment 2
It increased significantly at 90s inside and outside the cell insert. Unlike Experiment 1, it showed the
same increase trend both inside and outside. Inside and outside the cell insert, the inside showed
high numerical values. For VEGFmRNA, both 30s and 90s did not change.
When osteoblasts were observed immediately after 90s voltage application, it was found that the
osteoblasts on the end side of the membrane were peeled off.
Since VEGF is increasing, I think that it can be said that osteoblast proliferation and activation are
occurring due to electrical factors.
ALP activity measurement result
In experiment 1 it decreased significantly. Because factors leading to an increase are not obvious, it
is necessary to change electrical conditions.
Calculation results
Results are shown only at 90s and 120s. VEGF, VEGFmRNA change was seen. The electric field
[V/mm] changed to the initial 1.58±0.020, the final 1.21±0.064, the current value [A] changed
0.78±0.009 to 0.943±0.03 at 90s. On the other hand it changed like 1.711±0.145 to 1.25±0.014,
0.748±0.066 to 0.952±0.006 at 120s. As time goes by, current flows more easily. The reproducibility
of the experiment was obtained by calculating the given energy.
Conclusion
From this experiment, VEGFmRNA significantly increased at 120s in experiment 1, and VEGF
significantly increased at 90s application in experiment 2. Throughout the experiments, VEGF and
VEGF mRNA did not show the same trend, and it was found that the numerical value inside the cell
insert is higher in the inside. The same change was not observed in Experiment 1 and 2 because it
is considered that the cell cycle and the like are related to the initial VEGF amount of control.
However, it is clear that there is a change in osteoblasts by voltage application, which will also affect
bone formation. In the future, we would like to investigate more optimal electrical conditions and
mechanisms by measuring changes in electrical parameters of each measurement item,
measurement items that most affect bone formation, and changes in other enzymes.
References
1. BANDO Naoki, TOMINAGA Toshihiko, YUMOTO Hiromichi, SUMITOMO Takashi, HIRAO Saki,
HIRAO Kouji and MATSUO Takashi: “Application of electro-magnetic wave irradiation for enddontic
treatment –EMAT (Electro-Magnetic Apical Treatment)-”, The Journal of Japan Endodontic
Association Vol.32(3), 184-200, 2011-9-30 (in Japanese)
2. H.Yumoto, T.Tominaga, K.Hirao, T.Kimura, K.Takahashi, T.Sumitomo, N.Bando and T.Matsuo:
"Bactericidal activity and oral pathogen inactivation by electromagnetic wave ittadiation", Journal of

504
Applied Microbiology 113, 181-191, 2012
3. THOMASA OWEN, MICHAEL ARONOW, VICTORIA SHALHOUB, LEESAM BARONE,
LAURENS WILMING, MELISSA S. TASSINARI, MARY BETHKENNEDY, ShlRWlN POCKWINSE,
JANE B. LIAN, AND GARY S.STEIN "Progressive Development of the Rat Osteoblast Phenotype In
Vitro: Reciprocal Relationships in Expression of Genes Associated With Osteoblast Proliferation and
Differentiation During Formation of the Bone Extracellular Matrix", JOURNAL OF CELLULAR
PHYSIOLOGY 143:420-430(1990)
Figures

Figure 1. Scenery around the electrode plate Place the plate and osteoblast seeded insert
in a 6-well plate. Osteoblasts are immersed in PBS.

505
Figure 2. VEGF, VEGFmRNA measurement result It is the measurement result of
Experiment 1 and 2 (a, b). It is the measurement result of VEGFmRNA in experiment 1 and 2
(c).

506
 Figure 3. ALP measurement result It becomes ALP measurement result of Experiment 1.

PA-101 [14:30]
Impact of microwave irradiation on enzyme activities, cell membrane properties and bacterial
growth at constant temperatures
Tony Yan1
1Department of Chemistry and Centre for Biotechnology, Brock University, St. Catharines, Ontario, Canada,
L2S 3A1
Keywords: In vitro, RF/Microwaves, Other
Presented by: Tony Yan
Exposure of trypsin, but not amylase or phosphatase, to microwave at constant temperature leads
to a significant increase in its apparent activity. The membrane properties of PC3 prostate cancer
cells were found perturbed when the cell culture was subjected to 10 W microwave at 37°C. Lastly,
bacterium E. coli growth was slowed down under non-lethal microwave irradiation, again, at
constant 37°C. Proteomic analysis revealed subtle changes to E. coli under these conditions.
Introduction
In our attempts to study the potential interactions between microwave and biomolecules and
biological systems, it became of particular interest to maintain constant temperatures in the
experimental system in order to probe "microwave-specific effects". In this respect, experiments
were carried out to probe chemical reactions, enzymatic kinetics, cancer cell membrane integrity,
and bacterial growth in the presence of up to 10 W microwave while temperatures were maintained
through simutaneous cooling.
Methods
A CEM Coolmate microwave reactor operating at 2.45 GHz was used in the study, where the
temperature within the reaction vessel is maintained by simutaneous cooling through a jacket. The
temperature of the system is monitored by a fibre optic temperature probe in real time. With this
setup, temperatures of the system were maintained within 2°C from set values.
In order to examine the influence of exposure to microwave on enzymatic activities, hydrolysis
reactions catalyzed by trypsin, amylase and phosphatase were carried out while exposed to 10 W
microwave at 37°C. In an attempt to study the impact of microwave on cell membrane properties,
PC3 prostate cancer cells were cultured while subjected to non-lethal microwave irradiation. MTT
(3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium) and appoptotic assays were carried out to

507
compare uptake of MTT by cell membrane and scrambling of phosphatidylserine. Further, labelling
of cell membranes by a BODIPY fatty acid was carried out while the cell culture was exposed to
non-lethal microwave. Finally, bacterium Escherichia coli growth under exposure to microwave at
constant temperature was studied. Aspects of the bacterial growth, such as dissolvent oxygen
levels, cell viability, protein expression and transcription were investigated.
Results
Impact of microwave irradiation on enzymatic activities [1,2]
As Figure 1 shows, when enzymatic reactions were carried while exposed to microwave irradiation
(10 W at 37°C), trypsin activity was significantly increased, while those of amylase and phosphatase
remain virtually unchanged. It was also found that the changes in trypsin activity by microwave is
transient, and hydrolysis of casein catalyzed by trypsin pre-treated with microwave irradiation
showed the same reaction rate as that catalyzed by untreated trypsin.
Impact of microwave irradiation on cell membrane properties [3]
When PC3 prostate cancer cells were cultured in the presence of 10 W microwave at 37°C, there
was no indication of apoptosis as suggested by the Annexin V-assay (Figure 2a), and cells remain
viable as supported by propidium iodide assay (Figure 2b) [3]. MTT assay, however, showed that
reduction of MTT by PC3 cells irradiated with microwave is much more effective (Figure 2c),
presumably due to the more effecient uptake of MTT by cells exposed to microwave. In an
experiment where PC3 cells were incubated with a BODIPY fluorescent dye, incorporation of this
probe into cell membranes was consistantly increased in cells treated with microwave (Figure 2d).
Impact of microwave irradiation on E. coli growth [4]
When E. coli cultures were subjected to microwave irradiation at 10 W and 37°C, growth rate was
slowed down (Figure 3a), however, only transiently. Once microwave irradiation is stopped, bacterial
growth was restored, and there was virtually no difference in bacterial viability (Figure 3b). While the
dissolved oxygen level was slightly lower in the culture treated with microwave, work is currently
ongoing to investigate protein expression by 2D-gel based proteomics approach.
Discussion
The experiments described in this work strongly suggest that differences, in some cases quite
subtle, in the properties and functions of biomolecules and biological systems can result from
exposure to microwave, and of importance to note, while temperatures of the experimental systems
are maintained virtually constant through simutaneous cooling. The consideration of constant
temperature in these experiments are important, as the interpretation of results will otherwise be
complicated by temperature rises due to microwave heating.
Enzyme actities influenced by microwave exposure
Explanation to the observation that trypsin, but not amylase and phosphstase, activity is significantly
increased while exposed to microwave irradiation will require further investigation. It is possible that
the amino acids at the active sites of enzymes might be rendered better pre-organized due to
rotations as a result of microwave excitation. It will also be useful to carry out these experiments
while the reaction mixture is irradiated with microwave of different frequencies.
Cell membrane properties influenced by microwave irradiation
Cell membranes are fluidic. Input of microwave energy provides possibilities for membranes to
become more permeable, as suggested by the MTT assay and BODIPY labelling results. A more
direct approach will be necessary in order to probe membrane permeability or lipid flip-flop enabled
by microwave irradiation. Current work in this lab focuses on the scrambling of lipids in a microwave
field.
Cell growth and protein expression in the presence of microwave irradiation
Our current work demonstrated a delay in bacterial growth by low power microwave at constant
temperature. While we are tempted to suggest that electromagnetic waves such as microwave can

508
be perceived by living organisms as a stress signal, triggering changes in cascades of pathways in
order to adapt to this condition, we are currently pursuing a proteomic approach to identify possible
differential protein expressions in cells that are exposed to microwave irradiation.
Conclusion
While the influence of non-lethal microwave on the properties of biomolecules and biological
systems remains to be explored further, our investigation revealed subtle differences in microwave
on enzymatic activities, cell membrane properties, and bacterial cell growth. While it might be
argued that accumulation of heat and formation of "hot spots" contribute toward these differences,
future studies are required to elucidate these possible "microwave-specific effects" on biomolecules
and biosystems.
References
[1]. Mazinani, S. A.; DeLong B.; Yan, H. Microwave radiation accelerates trypsin-catalyzed peptide
hydrolysis. Tetrahedron Lett. 2015, 56, 5804.
[2]. Mazinani, S. A.; Yan, H. Impact of microwave irradiation on enzymatic activity at constant bulk
temperature is enzyme-dependent, Tetrahedron Lett. 2016, 57, 1589.
[3]. Mazinani, S. A.; Moradi, F.; Stuart, J. A.; Yan, H. Microwave Irradiation of PC3 Cells at Constant
Culture Temperature Alters the Incorporation of BODIPY into Cells and the Reduction of MTT,
ChemistrySelect, 2017, 2, 7983.
[4]. Mazinani, S. A.; Noaman, N.; Coorssen, J.; Yan, H. unpublished results.
Figures

Figure 1. Activities of trypsin, amylase and phosphatase in the presence of microwave (10 W
at 37°C).

509
Figure 2. PC3 prostate cancer cell membrane properties inflluenced by non-lethal microwave
exposure. (a) Annexin V assay; (b) propidium iodide assay; (c) MTT assay; and (d) labelling of
membrane by a BODIPY dye.

Figure 3. E. coli growth in the presence of microwave irradiation at 37°C. (a). Growth curve;
and (b). cell viability after microwave irradiation is removed.

510
PA-103 [14:30]
Localized intracellular temperature increase and ROS formation in immune relevant cells
under ELF-MF exposure: exposure set-up and procedures for real-time and post-exposure
measurements
Olga Zeni1, Anna Sannino1, Maria Rosaria Scarfi1, Stefania Romeo1, Mats-Olof Mattsson2, 3 & Myrtill Simko3
1Institute for Electromagnetic Sensing of the Environment, National Research Council of Italy, Napoli, Italy,
80124
2Center for Energy – ERT, Austrian Institute of Technology, Tulln, Austria, A-3430
3SciProof International AB, Östersund, Sweden, S-83132
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: Olga Zeni
In this study we address the question if there is a possible correlation between reactive oxygen
species formation and intracellular localized temperature increase (“hot spots”) in immune relevant
cells under ELF magnetic field exposure. A specific in vitro experimental procedure has been
established for real-time and post-exposure measurements of ROS formation and intracellular
temperature change.
Introduction
In in vitro studies, a multitude of biological effects have been reported following exposure to weak
extremely low frequency electromagnetic fields (ELF-MF), but they do not provide any conclusion
regarding the first point of interaction between MF and cells and the coupling mechanisms. As a
matter of fact, the studies have been carried out on different cell models, and different exposure
conditions have been applied to analyse different biological endpoints, making it difficult to perform
any comparisons.
The modulation of oxidative responses after ELF MF exposure has been described, and the release
of reactive oxygen species (ROS) has been more frequently reported than a decrease in ROS or an
absence of effects, but it has not been addressed systematically and in relation to mechanisms [1].
Simkó and Mattsson have found that ELF-MF induces ROS production in different immune relevant
cell types [2-3] and it was summarized in a recent review that ELF-MF exposure in vitro causes
changes in oxidative status as an early response [4].
The present study aims to address the question if there is a possible correlation between reactive
oxygen species formation and intracellular localized temperature increase (“hot spots”) in immune
relevant cells under ELF magnetic field exposure. For that purpose, immune relevant cells have
been exposed to ELF magnetic fields under different conditions of magnetic flux densities and
exposure durations to find out which ones that are eliciting ROS formation. An appropriate and
sensitive procedure has been established to analyse the ROS formation by flow cytometry as a bulk
measurement. Taking advantage of an in-house-manufactured temperature-controlled ELF
exposure system, we will measure local intracellular temperature changes in real time during ELF-
MF exposure in ROS producing cells under a confocal laser scanner microscope by using the
temperature sensitive fluorescence dye Rhodamine B. The hypothesis is that the interaction
between cells and ELF-MF can be induced either via heat development seen in the form of
intracellular hot spots, by induced electric fields that change the conformations and properties of
molecules such as receptors and ion channels, by radical pair mechanisms, or by a combination of
these or other mechanisms.
Materials and Methods
Cell cultures and experimental procedures
The macrophage cell lines J774.2, RAW 264 and MonoMac 6 were cultured in DMEM medium,
supplemented with 10% foetal bovine serum (FBS), 2 mM L-glutamine and 100U/ml penicillin, 100
µg/ml streptomycin, and maintained at 37°C in a humidified atmosphere containing 5% CO2. For
consistency and reproducibility, cell cultures were routinely maintained as monolayer by sub-

511
culturing twice per week by trypsinisation.
For the experiments, 48 h before exposure/treatment, 3 ml cell cultures (8 x104 cells/ml) were
established in 35-mm-diameter Petri dishes (Corning, NY) and FluoroDishes (WPI, Inc) for bulk
measurements and confocal microscopy, respectively.
Exposure system set up for bulk measurements by flow cytometry
The exposure system (Figure 1) consisted of two identical apparatuses each consisting of four
coaxial circular coils, placed horizontally, and thus parallel with respect to the surface of the Petri
dishes. The geometry of the coil system was numerically calculated in order to optimize the
extension of the magnetic field uniformity [5]. Coils were wound by a pair of parallel wires, so that,
according to different connections, the current could either flow in the same direction ‘‘wound
configuration’’ or in the opposite direction ‘‘counter-wound configuration’’. In the latter case the
magnetic fields produced by counter-wound coils cancel each other, allowing a ‘‘true’’ sham system,
in which the current and the power dissipation are the same as in the ‘‘wound configuration,’’ but the
magnetic flux density is theoretically zero. The four coils of the two systems were connected in
series and powered by an in-house-made power amplifier connected to a function generator (TTi
TG1010A). The magnetic flux density was monitored by a Hall probe gaussmeter (F.W. Bell, model
4048, accuracy ± 2% of reading). The frequency of the sinusoidal field was maintained at 50 Hz,
while different magnetic flux densities were used in the range 0.1 - 1.0 mT rms. The exposure and
sham-exposure systems were placed in the same incubator at 37°C and 5% CO2 in a humidified
atmosphere.
Exposure system set-up for real-time measurements by confocal microscopy
A double-coil system, with 12 cm inner diameter, was designed and realized to produce the 50 Hz
ELF magnetic field which allowed the real-time monitoring of cellular and molecular changes during
magnetic field exposure by confocal microscopy (Figure 2). In order to obtain temperature-controlled
exposure conditions, an Indium Tin Oxide (ITO)-based heating system, that is transparent and
electrically conductive, was realized and served as support for the Fluorodish to be exposed under
the microscope. In particular, an ITO-coated glass was connected, by means of a pair of carbon
electrodes, to a temperature controller, which, in turn, was controlled by a thermometer equipped
with a non-perturbing thermistor probe. According to the temperature measured in the sample, the
temperature controller provides a DC current to the ITO-based system, which heats the sample until
reaching a pre-set temperature value.
The ITO based system allowed us to calibrate the changes in fluorescence of the Rhodamine B
(RhB, λex = 550 nm), a temperature sensitive dye, with a quantum yield that decreases linearly as
temperature increases [6]. Loading of cells in DMEM containing 50 µM RhB for 2 h in incubator at
37°C was carried out prior of each measurement by confocal microscopy.
Measurement of intracellular ROS levels
For both bulk and real time measurements, the method developed by Le Bel et al. was used [7]. It
employs the fluorescent probe 2′,7′-dich lorofluorescin diacetate (DCFH-DA), which is a non-polar
compound that easily passes the cell membrane and is hydrolysed by intracellular esterases to the
non-fluorescent polar derivative, DCFH. In the presence of ROS, DCFH is oxidized to fluorescent
dichlorofluorescein (DCF). The assay was carried out as follows: concurrently with
exposure/treatments, cell monolayers were loaded in absolute DMEM medium w/o serum containing
10 μM final concentration DCFH-DA. After washing twice in cold PBS, cell monolayers were treated
in accordance with the different methods of analysis and results presented accordingly. In both
cases, menadione (2-methyl-1,4-naphtoquinone, MD, 1 mg/ml in DMSO) was given to cell cultures
at 0, 10, 20 and 40 µM final concentrations to trigger ROS formation.
Results and Discussion
Immune relevant cells can be activated either by the classical or by the alternative pathway to
release ROS. The classical pathway is active during an immune defence action against e.g.
invading pathogens leading to a high level of ROS release, whereas the alternative pathway is

512
active during e.g. healing processes and is accompanied by a low level of ROS production.
Specifically, very sensitive, in vitro experimental procedures have been established for real-time
measurement of intracellular temperature change and post-exposure measurements of ROS
formation. The decrease in RhB mean-intensity upon increasing temperature is shown in Figure 3
for a single cell region of interests (ROI) during confocal microscopy.
From the first exploratory experiments it appears that different cell lines react similarly to specific
chemicals such as menadione. However, the responses to ELF-MF seem to be different among the
used cell lines. This suggests that the induced/activated molecular mechanism is different in the
used cell lines. These findings regarding different cell responses were also confirmed by real-time
microscopy measurements in individual cells. Responses to both chemical treatment (menadione)
and physical stimulation (external temperature increase or ELF-MF exposure) were seen only in
subpopulations (responding and non-responding cells) of the cultured cells. This indicates that e.g.
temperature change is either tolerated by specific/individual cells or that other biophysical processes
are important which have to be elucidated in further studies.
Experiments are in progress to increase the number of measurements to provide robustness and
reliability to these preliminary observations.
References
[1] Wang H and Zhang X. Int J Mol Sci (2017) 18, 2175.
[2] Simkó M. Curr Med Chem (2007) 14:1141–1152.
[3] Simkó M, Mattsson M-O. J Cell Biochem (2004) 93:83–92.
[4] Mattsson M-O, Simkó M. Front Public Health (2014) 2:132.
[5] Gottardi G, Mesirca P, Agostini C, Remondini D, Bersani F. Bioelectromagnetics (2003)
24(2):125–133.
[6] Chen Y and Wood AW. Bioelectromagnetics (2009) 30:583-590
[7] Le Bel CP, Ali SF, Mc Kee M & Bondy SC. Tox Appl Pharmacol (1990) 104:17–24.
Figures

513
Figure 1. The ELF-MF exposure set-up. The field generation system is shown on the right,
while the incubator with the two-coil system for exposure and sham exposure is shown on the
left.

Figure 2. The ELF-MF exposure set-up and ITO-based heating system for real-time
measurements by confocal microscopy

514
 Figure 3. Decrease in RhB mean intensity upon increasing temperature. Data are presented
from a ROI placed on a representative single cell.

PA-105 [14:30]
Mobile phone radiation induces mode-dependent DNA damage in a mouse spermatocyte-
derived cell line: a protective role of melatonin
Lei Zhang1, 2, Chuan Liu1, 2, Qin-Long Ma1, 2, Zhou Zhou2, 3 & Zheng-Ping Yu1, 2
1Department of Occupational Health, Army Medical University (Third Military Medical University), Chongqing,
China, 400038
2Key Laboratory of Medical Protection for Electromagnetic Radiation, Ministry of Education, Army Medical
University (Third Military Medical University), Chongqing, China, 400038
3Department of Occupational and Environmental Health, Zhejiang University, Hangzhou, China, 310058
Keywords: In vitro, RF/Microwaves, Completed (published)
Presented by: Zheng-Ping Yu
This study evaluated whether exposure to mobile phone radiation (MPR) can induce DNA damage
in male germ cells. The levels of DNA damage were significantly increased following exposure to
MPR in the listen, dialed and dialing modes. Moreover, there were significantly higher increases in
the dialed and dialing modes than in the listen mode. However, the DNA damage effects of MPR in
dialing mode were efficiently alleviated by melatonin pretreatment. These results regarding mode-
dependent DNA damage have important implications for the safety of inappropriate mobile phone
use by males of reproductive age and the protective role of melatonin.
Purpose: To evaluate whether exposure to mobile phone radiation (MPR) can induce DNA damage
in male germ cells.
Materials and methods: A mouse spermatocyte-derived GC-2 cell line was exposed to a commercial
mobile phone handset once every 20 minutes in standby, listen, dialed or dialing modes for 24 h.
DNA damage was determined using an alkaline comet assay.
Results: The levels of DNA damage were significantly increased following exposure to MPR in the
listen, dialed and dialing modes. Moreover, there were significantly higher increases in the dialed
and dialing modes than in the listen mode. Interestingly, these results were consistent with the
radiation intensities of these modes. However, the DNA damage effects of MPR in dialing mode
were efficiently alleviated by melatonin pretreatment.

515
Conclusions: These results regarding mode-dependent DNA damage have important implications
for the safety of inappropriate mobile phone use by males of reproductive age and also suggest a
simple preventive measure, which is keeping our body as far away from mobile phones as possible,
not only during conversations but during “dialed” and “dialing” operation modes as well. Since the
“dialed” mode is actually part of the standby mode, mobile phones should be kept at distance from
our bodies even during standby operation. Furthermore, the protective role of melatonin suggests
that it may be a promising pharmacological candidate for preventing mobile phone use-related
reproductive impairments.

PA-107 [14:30]
Characterization of the suppressive effects of extremely-low-frequency electric fields on a
stress-induced increase in the plasma glucocorticoid level in mice
Takuya Hori1, 2, Takaki Nedachi2, Hiroshi Suzuki1 & Shinji Harakawa1, 2
1Bio-Self-Regulating Science Laboratory, Obihiro University Agriculture and Veterinary Medicine, Obihiro,
Japan, 0808555
2Research and Development, Hakuju Institute for Health Science, Shibuya-ku, Japan, 1510063
Keywords: In vivo, ELF/LF, Completed (unpublished)
Presented by: Takuya Hori
Recently, we found that in male BALB/c mice, the immobilization-induced increase in serum
glucocorticoid (GC) was reduced by exposure to an electric flied (EF) of 50 Hz, when the voltage
was applied via the upper electrode of a parallel plate electrode system. This effect was dependent
on both intensity (kV/m) and exposure time. The present study aimed to assess the anti-stress effect
of the EF in 3 scenarios with the experimental system: the use of either 50 or 60 Hz, which are the
standard power frequencies in most regions; varying levels of environmental brightness; complete or
partial shielding of the mouse from the EF. We compared the GC levels among control, EF-alone,
immobilization-alone, and co-treatment groups.
Introduction
Studies on extremely-low-frequency (ELF) electromagnetic fields generally have two perspectives:
health risk and clinical application. While experimentally evaluating either the efficacy or hazards of
certain substances or physical phenomena, adoption of a well-evaluated model is an extremely
important factor; furthermore, elucidating the precise biological effects of power line-frequency
electric fields (EFs) also requires experimental methods for quantitative and qualitative assessment.
We previously established an experimental system to evaluate the efficacy of power line-frequency.
With the experimental system, we found that in male BALB/c mice, the immobilization-induced
increase in serum glucocorticoid (GC) was reduced by exposure to an EF of 50 Hz. This effect was
dependent on both intensity (kV/m) and exposure time (1).
The present study aimed to investigate the anti-stress effect of EF in 3 scenarios: the use of either
50 or 60 Hz, which are the standard power frequency in most regions; different environmental
brightness levels, because the anti-stress effect was attenuated when a heavy-weight electrode was
used to prevent noise due to vibration of the electrode (data not shown); complete or partial
shielding of the mouse from the EF.
Materials and Methods
Animals
Eight-week-old male BALB/c mice were maintained in a specific pathogen-free environment at 24 ±
1 °C, with 50 ± 10% humidity and daily artificial illumination (12:12-h light/dark cycle with lights on
from 7:00 to 19:00). The animals hadad libitumaccess to standard laboratory chow and water,
except during EF exposure and immobilization.
EF exposure system

516
The EF exposure system (Fig. 1) consisted of a high-voltage transformer unit, a constant-voltage
unit to prevent unexpected interference from electrical noise originating from the commercial power
supply, and a parallel plate electrode system. The parallel plate electrode system comprised a
cylindrical plastic cage and two stainless-steel electrodes (1,000 × 600 mm) placed over and under
the cylindrical cage. To measure the field intensity and verify the operativeness of the system, an
optical fiber voltmeter was used. The EF intensity of 10 kV/m applied to the cage in which the
mouse was kept had ± 1% margin of error.
Immobilization
Stress was induced by immobilizing each mouse separately within a 50-mL centrifuge
(polypropylene) tube.
Test 1: Effect of EF frequency on the suppressive effect of EF on the stress-induced increase
in plasma GC in mice
The mice were divided into six groups (n = 6): a control group, two EF-alone groups (50 Hz or 60
Hz), immobilization-alone group, and two co-treatment groups [Stress(+)/EF(50 Hz or 60 Hz, 10
kV/m)] (Fig. 2).
Test 2: Effect of environmental brightness on EF effects
Male mice were divided into the following four groups (n = 6): a control group, EF-alone group,
immobilization-alone group, and co-treatment group (Fig. 2). In addition, three levels of illuminance,
i.e., 0, 200, and 490 lux, were established for each of the 4 aforementioned groups.
Test 3: Effect of complete or partial shielding from the EF
Male mice were divided into eight groups (n = 8): control, EF-only, immobilization-only, co-treatment,
and co-treatment with shielding [Stress(+)/EF(+)/shielded (sheet width: 5, 20, 80 mm for partial
shielding, or 200 mm for complete shielding)] (Fig. 2). In the co-treated but shielded group, animals
were wrapped with an EF shield consisting of a polytetrafluoroethylene sheet (0.1 mm thick, 5-200
mm width) containing carbon 0.05% (w/w).
Plasma GC and blood parameter measurements
Immediately after EF treatment, blood was collected from each mouse under 3% isoflurane
anesthesia. Immediately afterward, 10 µL of each blood sample was used to analyze blood
properties using a Cell-Tak system (Nihon Kohden, Japan). The remainder of the sample was
centrifuged, and the plasma was collected and stored at −80 °C until use. GC level was measured
by derivatization and Fluorimetric Assay.
Statistical analysis
To assess the effects of EF exposure with immobilization-induced changes in plasma GC levels,
two-way analysis of variance (ANOVA) was used, followed by Bonferroni’s multiple comparison test
for post-hoc analysis. Correlations were analyzed using Pearson’s correlation analysis and slope
regression analysis was conducted. Significance was defined as P < 0.05.
Results
Test 1
There was no significant difference in the plasma GC level between mice in the EF-only groups and
the control group.
However, plasma GC levels in mice treated with stress alone were significantly higher than those in
the control group mice (P < 0.0001). GC levels in the co-treatment groups (50 Hz or 60 Hz) were
also significantly higher than those in the control group mice and lower (P < 0.05, respectively) than
those in the immobilization-alone group.
Test 2
For every treatment, there was no correlation between plasma GC level and illuminance level. In all
illuminance conditions, there were no significant differences in plasma GC between the EF-alone
group and the control group, and GC levels in the stress-alone group were significantly higher than

517
those in the control group (P < 0.0001). Only at 200 lux, the GC level in the co-treatment group was
lower than that in the stress-alone group (P = 0.07 at 0 lux, P < 0.005 at 200 lux, P = 0.71 at 490
lux).
Test 3
GC levels in the co-treatment groups were also significantly higher than those in the control group;
however, they were lower than those in the stress-alone group (P < 0.01). GC levels in the co-
treated but shielded animals [Stress(+)/EF(+)/shielded (5, 20, 80, or 200 mm)] were higher than
those in the control mice (P < 0.0001). In the [Stress(+)/EF(+)/shielded (5 or 20 mm)] groups, the
levels were lower than those in the stress-alone group (P < 0.01 for 5 mm, P < 0.05 for 20 mm),
while the 80-mm- or 200-mm-shield groups had GC levels that were similar to those in the
immobilization-alone group.
On analyzing reciprocal plots of GC levels (Fig. 3), a significant negative correlation was observed
by Pearson’s correlation analysis (P < 0.005) and the slope was significantly non-zero as indicated
by linear regression analysis (P < 0.005, Y = −0.0016*X + 1.4).
Discussion
The following three experiments were conducted to characterize the biological effects of ELF EF: 1)
comparison of the effect of either 50 Hz or 60 Hz, which is the standard power frequency in most
regions, 2) evaluation of the effect of environmental brightness on the EF effect, and 3) evaluation of
the effect of complete or partial shielding of the mouse from the EF and the effect of the EF.
In all tests, plasma GC levels in immobilized mice were approximately 2 to 4.5 times those in the
control group, suggesting that the immobilization procedure affected the hypothalamic-pituitary-
adrenal axis of the endocrine system and the sympathetic-adrenomedullary system, which
manifested as stress (2).
However, because there was no difference in plasma GC between the EF-alone group and the
control group, we concluded that EF exposure of at least 50 Hz (or 60 Hz in test 1), 10 kV/m for 60
min did not activate the pituitary-adrenocortical axis or the sympathetic-adrenomedullary system in
mice.
In test 1, the finding that not only an EF of 50 Hz, but also an EF of 60 Hz, suppressed the increase
in the GC level in immobilized mice is consistent with our previous findings (3, 4). Therefore, one
conclusion of this study is that the anti-stress effect of a 60-Hz 10 kV/m EF should be similar to that
of a 50-Hz EF. Moreover, our findings provided fundamental knowledge about the biological effects
of exposure to EFs at the frequencies commonly used as power frequency.
The anti-stress efficacy of an EF was confirmed for illuminance at 200 lux, which is a condition
similar to that used in our previous study (1) and in tests 1 and 3 in this study; however, at 0 or 490
lux, no EF effect was observed. Therefore, we considered that lack of differences in the GC level
between the immobilization and the co-treatment groups was dependent on illuminance. The
suppressive effect of the EF on stress response, as observed in this study, promises to be a
potential candidate for preventing or treating stress-related disorders, and thus, has great social
significance.
Next, we aimed to characterize the relationship between the effect of EF exposure and the animal’s
body surface area exposed to the EF. GC levels in the co-treatment but shielded groups
[Stress(+)/EF(+)/shield (5 or 20 mm)] were lower than those in the immobilization-alone group, but
not lower than those in the groups [Stress(+)/EF(+)/shielding (80 or 200 mm)]. Therefore, the effect
of the EF on plasma GC was observed when the body surface area exposed to the EF was
relatively small, whereas the EF effect was diminished when the body surface area exposed to the
EF was relatively large. In addition, a reciprocal plot of plasma GC (as an index of the effect of the
EF) against body surface area shielded from the EF revealed a negative correlation. We conclude
that shielding from the EF inhibits its suppressive effect on stress-induced increases in GC in
immobilized mice; furthermore, the suppressive effect of EF exposure depends on the body surface
area exposed to the EF in mice. Considering either the clinical applications or health risks of EF, the

518
insight regarding the relationship between the exposed area to EF and EF-induced biological effects
might have widespread implications.
Potential mechanisms underlying the decrease in GC level induced by an EF include the
suppression of GC release from the adrenal cortex or enhanced degradation of GC in the liver. Our
latest study suggested that the effect can be attributed to the suppression of GC secretion (3). In
this scenario, GC release from the adrenal cortex may be suppressed directly or indirectly via
suppression of the upstream stress response. Because ELF EFs seem to have a certain effect on
ACTH secretion in the pituitary (5), the hypothalamic-pituitary-adrenal axis would participate in the
ELF EF-induced suppressive effects on GC levels in immobilized mice. Furthermore, GC receptor
gene transcript levels after EF, stress, or co-treatmfent should be investigated as a downstream
biological effect of EF exposure.
Conclusion
Our findings corroborated that exposure to an ELF EF at both 50 and 60 Hz suppresses
immobilization-induced changes in the endocrine system, at least with respect to acute
immobilization stress in mice. Our findings further suggested that the suppressive effect is
influenced by environmental illuminance. Furthermore, the suppressive effect of EF exposure
depends on the body surface area exposed to the EF in mice. Our findings could potentially be
extended to other species, including humans, upon incorporating further developments in study
methodology.
References
1. Hori, T., Inoue, N., Suzuki, H. & Harakawa, S. (2015). Bioelectromagnetics, 36, 302-308.
2. Selye, H. (1946). J Clin Endocrinol Metab, 6, 117-230.
3. Harakawa, S., Hori, T., Inoue, N. & Suzuki, H. (2017). Bioelectromagnetics, 38, 272-279.
4. Hori, T., Inoue, N., Suzuki, H. & Harakawa, S. (2017). Bioelectromagnetics, 38, 265-271.
5. Harakawa, S., Takahashi, I., Doge, F. & Martin, D.E. (2004). Bioelectromagnetics, 25, 346-351.
Figures

Figure 1. Electric field (EF) exposure system. A: Voltage generator and electrodes. B:
Summary of EF exposure space. C: EF exposure cage. D: Mouse restrained in a 50-mL
centrifuge tube.

519
Figure 2. Experimental design for tests 1, 2, and 3 to investigate the suppression of
immobilization-induced increases in the plasma GC level on EF exposure in mice.

520
 Figure 3. Effect of immobilization and EF exposure on the immobilization-induced increase in
the plasma GC level. Correlation between shield width and the reciprocal of plasma GC levels.

PA-109 [14:30]
STUDENT PAPER
Effect of long-term RF-EMF exposure on microglia activation in 5xFAD mice
Ye Ji Jeong1, 2, Hyung-Do Choi3, Jeong-Ki Pack4, Nam Kim5, Yun-Sil Lee6 & Hae-June Lee1
1Division of Basic Radiation Bioscience, Korea Institute of Radiological and Medical Sciences, Seoul, Korea
2Department of Life Sciences, Korea University, Seoul, Korea
3Department of EMF Research Team, ETRI, Daejeon, Korea
4College of Engineering, Chungnam National University, Daejeon, Korea
5School of Electrical and Computer Engineering, Chungbuk National University, Cheongju, Korea
6Graduate School of Pharmaceutical Sciences, Ewha Womans University, Seoul, Korea
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Ye Ji Jeong
To explore the long-term RF-EMF effect on Alzheimer’s disease, we performed 6 months RF-EMF
exposure to 5xFAD mice (SAR 5W/kg, 2 h/day, 5 days/week) and investigated behavioral changes
and neuroinflammatory response. 5xFAD mice exposed to long-term RF-EMF showed mild
decrease in Aβ deposition, memory improvement, and decrease in expressions of Iba-1 and
microglia regulator genes compared to sham exposed group. Our finding suggests possible
mechanism of beneficial effect of long-term RF-EMF on AD mice.
Introduction
The increasing use of mobile phones has generated concern regarding the impact of radiofrequency
electromagnetic fields (RF-EMF) on public health [1]. In addition to amyloid-β plaque and tau
neurofibrillary tangle deposition, neuroinflammation is considered a key feature of Alzheimer's
disease pathology. Inflammation in Alzheimer's disease is characterized by the presence of reactive

521
astrocytes and activated microglia surrounding amyloid plaques, implicating their role in disease
pathogenesis [2]. Previously we reported beneficial effect of chronic RF-EMF on 5xFAD mouse [3]
but biological effects of RF-EMF on the AD brain should be further investigated. In the present study,
we investigated whether long-term RF-EMF affects microglia activation and memory impairment in
5xFAD mice, a widely used Alzheimer’s animal model.
Methods
At 1.5 months of age, 5xFAD mice were assigned to one of 2 groups (the RF-EMF- and sham-
exposed groups, 11 mice per group). The RF-EMF group was exposed to a 1950 MHz EMF field
within a reverberation chamber for 6 months (5 W/kg SAR, 2 h/day, 5 days/week). The Y-maze and
novel object recognition memory test were used to evaluate memory function following the 6-month
RF-EMF exposure. And Aβ deposition and microglia activation were evaluated in the hippocampus
and cortex of the 5xFAD mice using Immunoblotting, immunohistochemistry or qPCR.
Results and conclusion
In the Alzheimer's disease model, there was no significant decrease in brain inflammation cytokine
by RF-EMF, however, significant decrease in microglial cell marker Iba-1 and genes involved in
microglia activation. Also, RF-EMF exposure for 6 months revealed induced memory improvements
in behavioral testing. Collectively, these results demonstrate that long-term RF-EMF contributes to
microglia inactivation, as well as memory improvement in 5xFAD mice.
References
[1] Hardell L, Sage C. 2008. Biological effects from electromagnetic field exposure and public
exposure standards. Biomed Pharmacother. 62(2):104–109.
[2] Farfara D, Lifshitz V, Frenkel D. 2009. Neuroprotective and neurotoxic properties of glial cells in
the pathogenesis of Alzheimer's disease. J Cell Mol Med. 12(3):762–780.
[3] Jeong YJ, Kang GY, Kwon JH, Choi HD, Pack JK, Kim N, Lee YS, Lee HJ. 2015. 1950 MHz
Electromagnetic Fields Ameliorate Aβ Pathology in Alzheimer's Disease Mice. Curr Alzheimer Res.
12(5):481–492.
Acknowledgement
This work was supported by Institute for Information & communications Technology Promotion (IITP)
grant funded by the Korea government (MSIT) (2017-0-00961, Study on the EMF Exposure Control
in Smart Society)

PA-111 [14:30]
The effect of exposure to radiofrequency-electromagnetic field on serotonin metabolism in
rat
Hye Sun Kim1, Man-Jeong Paik3, Sangbong Jeon4, Hyung Do Choi4, Jeong-Ki Pack5, Nam Kim6 & Young
Hwan Ahn1, 2
1Department of Neurosurgery, Ajou University , Suwon , Korea, 16499
2Neuroscience Graduate Program, Ajou University, Suwon, Korea, 16499
3College of Pharmacy, Sunchon National University, Sunchon , Korea, 57938
4Radio Technology Research Department, Electronics and Telecommunications Research Institute, Dea Jeon
, Korea, 34129
5Department of Radio Sciences and Engineering, Chungnam National University, DaeJeon , Korea, 34134
6School of Electrical and Computer Engineering, Chungbuk National University, Cheongju, Korea, 28356
Keywords: In vivo, RF/Microwaves, Completed (unpublished)
Presented by: Young Hwan Ahn
The study investigated whether exposure to the radiofrequency identification signal could affect on

522
serotonin metabolism in rat. The preliminary results showed that RF-EMF exposure for 8 hours daily
for 2 weeks at whole-body SAR of 2 W/kg to the adult Sprague-Dawely rats caused a significant
reduction of urinary level of 5-methoxyindole-3-acetic acid, although urinary levels of serotonin and
5-hydroxyindoleacetic acid were not change significantly.
Introduction
Rapid expansion of the mobile technology raises issues regarding the health effect of
radiofrequency electromagnetic field (RF-EMF) signal. Serotonin (5-hydroxytryptamine) derived from
tryptophan and relays signals between nerves, cells, or neurons, and play a important role in central
nervous system and is one of the brain chemicals responsible for a variety of psychological and
other body functions. Serotonin are mostly produced from gastrointestinal enterochromaffin cells.
Metabolism of serotonin is carried out primarily by monoamine oxidase (MAO) that convert
serotonin, up to 80% to 5-hydroxyindole acetaldehyde, which in turn is readily metabolized to
produce 5-hydroxyindole acetic acid as the major excreted metabolite of serotonin. The aim of this
study is to evaluate whether exposure to the radiofrequency identification signal could affect on
serotonin metabolism in rat.
Materials and methods
A reverberation chamber was used as a whole-body exposure system. Adult Sprague-Dawely rats
were divided into two groups; sham-exposed and RF-exposed group. Rats were exposed to the RF-
EMF signal from a 915 MHz RFID exposure system for a eight hours daily for 2 weeks at whole-
body SAR of 2 W/kg and were sacrificed immediate after completion of exposure experiment. Using
the gas chromatography mass spectrometry machine, urinary levels of serotonin,
5-hydroxyindoleacetic acid (5-HIAA) and 5-methoxyindole-3-acetic acid (5-MIAA) were measured
and compared them between the groups. Data were analyzed using the statistical package SPSS
ver. 12.00(SPSS, Chicago, IL, USA). Data are presented as means ± standard deviations. P values
0.05 indicated statistical significance.
Results
The preliminary results showed that a significant reduction of urinary level of 5-MIAA was observed
in RF-exposed group compared with those in sham-exposed group (p = 0.03) (Fig 1, will be
presented).
Urinary levels of serotonin, and 5-HIAA were not changed significantly in RF-exposed group
compared with those in sham-exposed group (p < 0.05 in each value) (Fig 2, will be presented).
Conclusion
The result suggests that RF-EMF exposure at the whole body SAR 2W/kg may cause changes in
serotonin metabolism in adult male rat. Further studies including repeated experiment are needed.
Grant sponsor: IT R&D program of MSIP/IITP; grant number: 2017-0-00961, Study on EMF
exposure control in smart society.

PA-113 [14:30]
Brain sensitivity to electromagnetic fields (4G) at various ages in the rat: memory
persistence and associated genetic programs
Anne Pereira de Vasconcelos1, Aurélie Bonelli-Salvadori1, Marie-Muguet Klein1, Brigitte Cosquer1, Niels
Kuster2, Myles Capstick2 & Laurette Boutillier1
1Laboratoire de Neurosciences Cognitives et Adaptatives, UMR 7364 CNRS-University of Strasbourg,
Strasbourg, France, F-67000
2IT'IS Foundation, Zurich, Switzerland, CH-8004
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Anne Pereira de Vasconcelos

523
We studied, in young, adult and aged rats, the effects of a 3-month exposure to a 4G signal (LTE,
900 MHz, SARwhole body 0.33 W/kg, ~ 61 V/m) in a reverberation chamber (4 h/day, 5 d/week).
Chronic RF exposure did not affect acquisition or long-term retention of a spatial memory in the
Morris water maze, whatever the age. Anxiety and locomotion were also not changed by RF. In adult
rats and in resting conditions, RF exposure induced a significant number of differentially expressed
genes (RNA Seq) in the prefrontal cortex. The 47 up-regulated genes were functionally associated
with myelination, while the 33 down-regulated genes were associated with different isoforms of
collagen. These gene expression data are currently under validation.
Aims
The exponential use of cellular mobile communication (phone, relay stations) causes populations to
be actively or passively exposed to electromagnetic fields (EMFs). Clinical studies showed that
acute EMFs (GSM or UMTS) exposure may or may not induce subtle cognitive changes with
possible age-dependent effects (Regel & Acherman, 2011). Experimental studies in rodent showed
either deleterious, beneficial or no effect on cognition, including memory, after acute or chronic
EMFs exposure. No study explored yet the impact of EMFs on memory persistence, i.e.
consolidated at the systems level (Winocur et al, 2010). Our aim was to study in the Rat at various
ages, the effects of a chronic Radiofrequency (RF, LTE, 4G) exposure on the formation and
persistence of a spatial memory and gene expression in one memory-associated brain region.
Methods
RF exposure: Male Long Evans rats at 3 ages (Young: 1 month-old, Adult: 6 month-old and Aged:
18 month-old) were used. In each age-group, 3 sub-groups of rats were either exposed for 3 months
to RF (4G, LTE, 900 MHz, SARwhole body 0.33 W/kg, ~ 61 V/m) in a reverberation chamber (4 h/day,
5 d/week) (RF) or not exposed to RF in the same chamber and same conditions (Sham), or were
placed in the same room but outside of the chamber (Controls).
Behavioral experiments: After the end of the 3-month RF/Sham exposure, all rats were subjected to
behavioral tests consisting of: i) home-cage locomotor activity over 24h, ii) anxiety in a plus-maze,
and iii) spatial memory in the Morris water maze (8 consecutive acquisition days, 4 trials/day with a
fixed immersed platform location) with a probe test at recent (1 day)- or remote (8 or 25 days)-post-
acquisition delay.
Gene expression analyses: In a separate experiment, the Adult rat group (6 months of age) was
used to assess gene expression by RNA sequencing in the medial prefrontal cortex in the same 3
experimental conditions (RF, Sham, Controls). In addition, both resting condition (no memory test;
Home Cage) or Morris water maze acquisition (after 3 days of acquisition; Learning) were
considered. RNA-Seq libraries were generated from 300 ng of total RNA (minimum RIN>8;
n=3/group). Sequencing was performed by the GenomEast Platform (Illkirch, France), a member of
the ‘France Génomique’ consortium (ANR-10-INBS-0009), on the Illumina Genome Hiseq2500 as
single-end 50 base reads. Reads were mapped onto the rn6 assembly of the Rat genome.
Statistical analysis was performed with the method proposed by Love et al. (2014) implemented in
the DESeq2 Bioconductor library (v1.0.19). Gene ontology (GO) analyses for functional enrichments
were performed on deregulated genes considering a |log2 Fold Change| > 0.5 and FDR<0.01, using
the tools DAVID (Huang da et al, 2009) and/or the online database STRING. RT-qPCR were further
led for validation (n=6/group).
Results
Chronic RF exposure did not affect acquisition or long-term retention of a spatial memory in the
Morris water maze vs Sham rats, whatever the age and the post-acquisition delay (Figure 1).
However, in the Adult group at the longer post-acquisition delay (25 days), rats from the no chamber
condition (Controls) did not remember the platform location, while the rats exposed to the chamber
(RF, Sham) did (Figure 1, Adult). Due to an old age (22.5-23.0 months), the persistence of the
memory in the Aged group was tested at 8 days, not 25 days. Anxiety and locomotor activity tested
3-4 days after the end of the 3-month exposure, were not affected by RF when compared to Sham

524
rats (Figure 2).
RNA-sequencing data showed that RF exposure induced a significant number of differentially
expressed genes (47 up- and 33 down- regulated genes) compared to Sham group, in the medial
prefrontal cortex of resting rats (Home Cage condition). Functional annotation identified an
association of these genes with myelination for up-regulated genes (Benjamini, p<0.05). Different
isoforms of collagen were found in the down-regulated genes. In Learning condition, RF exposure
revealed 2 up-regulated genes and 55 down-regulated genes, but these genes did not show any
significant functional association. Deregulated gene expression measured in these conditions is
currently being validated by either RT-qPCR experiments (n=6) and a new RNA-seq study (n=3), in
order to establish whether these changes in gene expression are linked to the chronic exposure to a
4G signal. Importantly, all groups (Control, Sham, RF) presented higher level of immediate early and
plasticity genes (e.g. c-fos, Arc and Egr1) in response to learning, demonstrating activity-regulated
transcription in the medial prefrontal cortex at this time-point. This in agreement with its implication
during episodic memory encoding (Frankland and Bontempi, 2005; Lesburguères et al., 2011; Bero
et al., 2014), and likely relevant to the good performance observed in all experimental groups
(Control, Sham, RF) during acquisition and recent memory (Figure 1).
Discussion
The originality of our approach was to study for the 1st time in the Rat at different ages, the impact of
a chronic exposure to a 4G signal on memory persistence and to evaluate gene expression
modulation both in resting conditions (no memory test) and during learning of a spatial memory task,
in a brain structure essential for memory persistence (the medial prefrontal cortex, e.g., Frankland
and Bontempi, 2005). Our results confirm some of previous data showing the absence of effect of
RF on memory in young or adult rodents (e.g., Dubreuil et al, 2002; Maaroufi et al, 2014; Zhang et al
2017) and extend them to aged animals. It is to note that the literature in this field is controversial
with data showing also either deleterious memory effect (e.g., Maaroufi et al, 2014; Tang et al, 2015)
or beneficial effects (Kumlin et al, 2007; Wang et al, 2017), the latter one being observed notably in
animal models of neurodegenerative diseases (for Alzheimer’s models, e.g. Arendash et al, 2010;
Son et al, 2018). However, comparisons between studies are difficult because of the great variability
in animal models (rats, mice, including genetically-modified animals), in the type of memory (spatial,
fear conditioning, recognition), of RF signal (UMTS, W-CDMA,Wi-Fi, GSM…), of exposure (head
only vs whole body), of device and dosimetry, as well as exposure duration (from 15 min to 8
months). As our behavioral data did not show any differences as a function of age, we did not
confirm the hypothesis of a particular sensitivity to RF of the young brain or the aged one.
Concerning the impact of RF on brain gene expression, to our knowledge, very few studies were
performed and none in the context of memory formation/persistence. Using cDNA microarray, Zhao
et al (2007) showed some change in gene expression profile of rat neurons in vitro after 24h
exposure to 1800 MHz. Within the 12 000 genes analyzed, they found 24 up-regulated- and 10
down-regulated genes associated with multiple cellular functions, such as cytoskeleton, signal
transduction pathway or metabolism. Likewise, Fragopoulou et al (2012) showed in mice, 143
proteins (brain proteome) whose expression was up- or down-regulated in 3 brain regions after
8-month exposure to either a DECT base (1900 MHz) or a GSM 900 MHz signal. These changes
targeted several neural function-, cytoskeletal- and brain metabolism proteins. Our data, showing
modulation of some genes by RF in both resting and Learning conditions, remain preliminary and
will be confirmed and extended to a second memory-related region, the hippocampus.
Conclusion
A 3-month exposure to a mobile phone RF signal (LTE, 4G at environmental level) had no impact on
spatial learning and memory (recent or remote), as well as on anxiety and locomotion. However, RF
exposure seemed to induce a few changes in gene expression in the medial prefrontal cortex, but
did not alter activity-dependent transcription in response to learning in this structure. Specific gene
expression changes are currently under validation.

525
References
Arendash GW, Sanchez-Ramos J, Mori T, Mamcarz M, Lin X, Runfeldt M, Wang L, Zhang G, Sava
V, Tan J, Cao C (2010) Electromagnetic field treatment protects against and reverses cognitive
impairment in Alzheimer's disease mice. J Alzheimers Dis 19(1):191-210.
Bero AW, Meng J, Cho S, Shen AH, Canter RG, Ericsson M, Tsai LH (2014) Early remodeling of the
neocortex upon episodic memory encoding. Proc Natl Acad Sci U S A. 111:11852-7.
Dubreuil D, Jay T, Edeline JM (2002) Does head-only exposure to GSM-900 electromagnetic fields
affect the performance of rats in spatial learning tasks? Behav Brain Res 129: 203-10.
Frankland PW, Bontempi B (2005) The organization of recent and remote memories. Nat Rev
Neurosci 6:119-30.
Fragopoulou A, Samara A, Antonelou MH, Xanthopoulou A, Papadopoulou A, Vougas K,
Koutsogianopoulou E, Anastasiadou E, Stravodopis DJ, Tsangaris G, Margaritis L (2012) Brain
proteome response following whole body exposure of mice to mobile phone or wireless DECT base
radiation. Electromagnetic Biol Med 31: 250-74.
Huang da W, Sherman BT, Zheng X, Yang J, Imamichi T, Stephens R, Lempicki RA (2009)
Extracting biological meaning from large gene lists with DAVID. Curr Protoc Bioinformatics.
Sep;Chapter 13:Unit 13.11.
Kumlin T, Iivonen H, Miettinen P, Juvonen A, van Groen T, Puranen L, Pitkäaho R, Juutilainen J,
Tanila H (2007) Mobile phone radiation and the developing brain: behavioral and morphological
effects in juvenile rats. Radiat Res 168(4): 471-9.Li HJ, Peng RY, Wang CZ, Qiao SM, Yong Z, Gao
YB, Xu XP, Wang SX, Dong J, Zuo HY, Li Z, Zhou HM, Wang LF, Hu XJ (2015). Alterations of
cognitive function and 5-HT system in rats after long term microwave exposure. Physiol Behav
140:236-46.
Lesburguères E, Gobbo OL, Alaux-Cantin S, Hambucken A, Trifilieff P, Bontempi B (2011) Early
tagging of cortical networks is required for the formation of enduring associative memory. Science.
331:924-8.
Love MI, Huber W, Anders S (2014) Moderated estimation of fold change and dispersion for RNA-
seq data with DESeq2. Genome biology 15: 550.
Maaroufi K, Had-Aissouni L, Melon C, Sakly M, Abdelmelek H, Poucet B, Save E (2014) Spatial
learning, monoamines and oxidative stress in rats exposed to 900 MHz electromagnetic field in
combination with iron overload. Behav Brain Res 258:80-9.
Regel and Ackermann (2011) Cognitive performance measures in bioelectromagnetic research--
critical evaluation and recommendations. Environ Health 10(1) :10.
Son Y, Kim JS, Jeong YJ, Jeong YK, Kwon JH, Choi HD, Pack JK, Kim N, Lee YS, Lee HJ (2018)
Long-term RF exposure on behavior and cerebral glucose metabolism in 5xFAD mice. Neurosci Lett
666:64-9.
Tang J, Zhang Y, Yang L, Chen Q, Tan L, Zuo S, Feng H, Chen Z, Zhu G (2015) Exposure to 900
MHz electromagnetic fields activates the mkp-1/ERK pathway and causes blood-brain barrier
damage and cognitive impairment in rats. Brain Res 1601:92-101.
Wang K, Lu JM, Xing ZH, Zhao QR, Hu LQ, Xue L, Zhang J, Mei YA (2017) Effect of 1.8 GHz
radiofrequency electromagnetic radiation on novel object associative recognition memory in mice.
Sci Rep7:1-12.
Winocur G, Moscovitch M, Bontempi B (2010) Memory formation and long-term retention in humans
and animals: convergence towards a transformation account of hippocampal-neocortical
interactions. Neuropsychologia 48(8):2339-56.
Zhang JP, Zhang KY, Guo L, Chen QL, Gao P, Wang T, Li J, Guo GZ, Ding GR (2017) Effects of 1.8
GHz Radiofrequency Fields on the Emotional Behavior and Spatial Memory of Adolescent Mice. Int
J Environ Res Public Health 14(11): 1-14.

526
Zhao R, Zhang S, Xu Z, Ju L, Lu D, Yao G (2007) Studying gene expression profile of rat neuron
exposed to 1800MHz radiofrequency electromagnetic fields with cDNA microassay. Toxicology
235(3):167-75.
Figures

Figure 1. Effects of a chronic exposure to a RF 4G signal in the Rat on the

formation/persistence of a spatial memory in the Morris water maze. After 8 days of
acquisition (4 trials/day) during which rats learned the localization of a fixed submerged
platform in the NO quadrant (target), 2 probe tests were performed [1 and 25 days (Young,
Adult), 1 and 8 days (Aged)]. Performance during the probe test (platform removed) is
expressed in percentage of time spent in the target quadrant vs the 3 other quadrants (>25%
[chance] meaning a memory of the platform location). All data are expressed as means ±
SEM. Dashed line represents chance level (25%). * p < 0.001, significant differences from
‘others’, # p < 0.05 significant differences between groups.

527
Figure 2. Effects of a chronic exposure to a RF 4G signal in the Rat on locomotion and
anxiety. Results are expressed as means ± SEM of the number of longitudinal cage crossing
per hour during the diurnal and nocturnal phases of the activity test over 24h (Left) and of the
time spent in open & closed arms and in the center of the Plus maze (Right). The elevated
Plus maze test consisted of 4 arms (50 cm long x 10 cm wide) fixed to a central platform (10
x10 cm): 2 arms with 40 cm high side-walls (closed arms) and 2 arms with 1.5 cm high
borders (open arms). The light was adjusted to ensure uniform lighting on each pair of arms
(10 lux on open- and 2 lux on closed-arms). The rat has 5 min to freely explore the maze.

528
PA-115 [14:30]
Effect of rotating magnetic field on oocytes maturation
Yingying Zhao1, Tianying Zhan1 & Xiaomei Wang1
1Physiology Department, Shenzhen, China, 518060
Keywords: In vivo, Static, Completed (unpublished)
Presented by: Tianying Zhan
key word: rotary non-uniform magnetic fields; in vivi; oocytes; ovarian; ATP content; mitochondrial
respiratory chain; completed(unpulished). Presented by: Tianying Zhan.
Biologic effects of magnetic field have been studied many years such as bird migration, tissue
regeneration, and post-operational rehabilitation. However, the magnetic impact on the reproductive
systemis rarely reported. In this research, we investigated the maturity rate and Mitochondrial
respiratory chain function of oocytes from mice exposed to rotary magnetic fields (RMFs).To
investigate the influence of rotary nonuniform magnetic fields(RMFs) on oocytes growth, we used
the 4-6wKMfemale mice divided into the control group and treated group. The experimental group
was treated with 0.4T RMFs exposuresfor 15 days 2 hours a day, each mouse was intraperitoneally
injected with 5-10 IU PMSG, and 48 hours later, an intraperitoneal injection of 5-10 IU HCG was
performed. Eighteen hours later, oocytes were collected and tested with control group. The results
showed higher ovarian volume with less maturity rate of follicles in RMF exposure groups.RMF
caused increased mitochondrial respiratory chain activity inovarianwith elevated intracellular ATP
content.Increased expression of genes function in mitochondrial respiratory chain and ATP synthase
may lead to an increase in ATP content.These results indicate that the RMF mainly affects the
transfer of electrons in the electron transport chain and changes in the amount and activity of the
relevant enzymes leading to an increase in intracellular energy metabolism.

PA-117 [14:30]
Investigation of high frequency electromagnetic field component of cold microwave argon
plasma - Effects of modulation of skin microcirculation wound healing and regeneration in
mice in vivo
Lubomir Traikov1, Todor Bogdanov1, Nicoll-Ellen Kafozoff1, Radka Hadjiolova2, Liubina Vesselinova3,
Margarita Kouzmanova4, Gabriela Atanasova5 & Nikolai Atanasov5
1Dept. Medical Physics and Biophysics, Medical University-Sofia, Sofia, Bulgaria, 1431
2Dept. Pathophysiology, Medical University-Sofia, Sofia, Bulgaria, 1431
3Dept. Physical and Rehabilitation Medicine, Military Medical Academy – Sofia, Sofia, Bulgaria, 1606
4Dept. Biophysics and Radiobiology, Sofia University "St. Kliment Ohridski", Sofia, Bulgaria, 1164
5Dept. Wireless Communication and Broadcasting, College of Telecommunications and Post, Sofia, Bulgaria,
1700
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Lubomir Traikov
Current work present results of health effects of Cold Microwave Argon Plasma-CMAP over
stimulation of processes of skin wound healing and regeneration. In order to explore the influence of
all physical and chemical factors of this physical complex system we separated high frequency
electromagnetic field component from net argon gas flow. In order to evaluate biological effects of
high frequency electromagnetic field HF-EMF, we investigate micro-vessel tone response, using
Laser Doppler Flowmetry-LDF. After HF-EMF action, no significant increase of surface body
temperature of investigated animals was detected using high-resolution FLIR-camera, but an
increase of the skin blood flow was measured by LDF.
Introduction
Presented investigation is a part of scientific project based on investigation of the effects of Cold

529
Microwave Argon Plasma-CMAP over stimulation of processes of skin wound healing and
regeneration. In order to explore in details influence of different physical and chemical factors of
such a complex system we separated high frequency electromagnetic field component from argon
gas flow. In order to evaluate biological effects of high frequency electromagnetic field HF-EMF
(2450 MHz ± 20 MHz; Output power 150 mW/cm2), in this work we investigate micro-vessel tone
response, using Laser Doppler Flowmetry-LDF. Skin blood flow variations are measured by LDF.
Additionally for determination of surface skin temperature we used high-resolution (0.001K) FLIR-
camera. After HF-EMF (2450 MHz ± 20 MHz; Output power 150 mW/cm2) action we do not
registered significant increase of surface body temperature of investigated animals, but we
registered increasing of the blood flow registered by LDF.
By performing such technique, we recover the filtered signal from the LDF signal. Low-frequency
range pulsation show us the functional state of the microcirculatory system. In conclusion, the low-
frequency band monitoring can be used for micro-vessel tone control within the frequency ranges
corresponding to endothelial, neurogenic and myogenic activities. After 14 days of 30 min/daily
exposure with HF-EMF (2450 MHz ± 20 MHz; Output power 150 mW/cm2), postmortem immune
fluorescent tissue analysis was applied, with digital image densitometry was performed estimation of
area of blood vessel network, dermal tissue of exposed animals was abundant of blood vessels
(area of blood vessel network was larger) in comparison of sham exposed.
Material and Methods
System for generation of Cold Microwave Argon Plasma-CMAP including generator of high
frequency electromagnetic field (SOLID STATE MICROWAVE GENERATOR 1-200 W, 2450 MHz
MMS200WSMXPR2IR Sairem, Ltd.; Coupled with resonator type SURFATRON 80 Sairem, Ltd.,
sustained and transformed output power 15 W into 150 mW) In this paper we measured low-
frequency pulsations of LDF by (PeriFlux System PF5, Perimed Ltd. PF-1 probe, with a near-
infrared laser wavelength 780 nm) LDF was measured using a laser Doppler flow meter after every
30 min/daily chronic exposure in continuation of 14 days. LDF was measured in arbitrary perfusion
units (p.u.). The sampling frequency was 40 Hz. The temperature measurements were performed
with an Infrared camera (FLIR Infrared Cameras; Rockwell Automation Ltd.) designed for recording
temperature with a resolution 0.001 K and sampling frequency 20 Hz. Probes (LDF) were attached
to the skin surface of the distal part of mice back. All the subjects had 10 min acclimatization period
in the lab, and the duration of signal collecting was 20 min.
The Wavelet analysis of the LDF signal, registered on 20 ICR (n=20) preliminary depilated mice is
applied. Obtained results revealed a high correlation in the low-frequency range 0.01<ν<0.1 Hz of
vasomotion between control and HF-EMF exposed animals. We performed an inverse wavelet
transformation for the frequency band that corresponds to active mechanisms of vascular tone
regulation (0.01<ν<0.1 Hz).
Results and Discussion
The wavelets application demonstrates the efficiency of this method for the frequency-by-frequency
correlation analysis of biomedical signals, which are typically short and nonstationary signals as a
signals of vasomotion and skin temperature-ST changes. The WCC of LDF and ST signals for 20
ICR (n=20) preliminary depilated mice shows a high correlation in the low-frequency subrange 0.01
< ν < 0.1 Hz, which plays a major role for diagnosing endothelial dysfunction. The WCC function
provides the phase difference φ(ν) used to estimate the effective depth of temperature wave
production. Taking into account the decay rate for temperature oscillations at each frequency and
the LDF-ST phase shift, we have performed an inverse wavelet transform for the frequency band
that corresponds to active mechanisms of vascular tone regulation.
Wounds in the magnet group healed in an average of 14 days, significantly faster than those in
either the sham group (22 days,P= .006) or control group (20.3 days,P< .0001). There was no
statistically significant difference between the sham and control groups (P= .45).

530
Conclusions
An externally applied, low-power, high frequency electromagnetic field increases the rate of
secondary healing. Review of the literature consist a lot of evidences regarding the use of
electromagnetic energy to aid the healing of bone, tendon, and skin.
After 30 min/daily exposure with HF-EMF, we registered increasing of amplitude of low-frequency
range pulsation of vasomotion. These changes show us the functional state of the microcirculatory
system. In conclusion, the low-frequency band monitoring can be used for micro-vessel tone control
within the frequency ranges corresponding to endothelial, neurogenic and myogenic activities. After
applied postmortem immune fluorescent tissue analysis, with digital image densitometry performed
estimation of area of blood vessel network, dermal tissue of exposed animals was abundant of
blood vessels (area of blood vessel network was larger) in comparison of sham exposed.
Key words: HF-EMF; Microcirculation; Cold Microwave Argon Plasma; health effects
Acknowledgments: Bulgarian National Fund for Research and Science, Project №DM03/3/2016

PA-119 [14:30]
Molecular dynamics simulations of the A2A adenosine receptor in presence of magnetic field
Elena della Valle1, Federico Del Signore1, Paolo Marracino1, Davide Cocco1, Stefania Setti2, Ruggero
Cadossi2, Micaela Liberti1 & Francesca Apollonio1
1Department of Information Engineering, Electronics and Telecommunications, University Sapienza of Rome,
Rome, Italy
2IGEA SpA, Via Parmenide 10/A, Carpi, Italy
Keywords: Mechanistic/Theoretical, Static, Work in Progress
Presented by: Francesca Apollonio
In this paper, we present the implementation of an external static magnetic field with the Velocity
Verlet algorithm for performing Molecular Dynamics simulations. Molecular Dynamics simulations
allow to understand at molecular level the interaction mechanisms between atoms under specific
conditions. Here MD simulations of a receptor protein have been performed in presence of an
external magnetic field in order to try to elucidate specific endpoints of interaction with the field.
Introduction
Several studies have shown how the use of low frequency magnetic fields may have biological
effects on different cells functions. It has been reported that the use of pulsed magnetic fields
increases the anti-inflammatory effect of different types of cells such as neutrophils, synoviocytes,
chondrocytes and osteoblasts, with significant reduction in some of the most important inflammatory
cytokines [1-3]. The mechanisms of interaction underlying such effects, pointed out an involvement
of adenosine receptors (ARs). In particular, it has been shown that pulsed magnetic field exposure
mediates a significant upregulation of A2A adenosine receptor. Nevertheless, since direct
experiments on the action of magnetic fields on transmembrane proteins are not easily to be
performed, it still remains to be elucidated which is the final endpoint of the exposure.
To this regard, simulations based on Molecular Dynamics (MD) may become a strategic tool to study
molecules behavior, under different physical conditions. MD simulations are used to observe the
dynamics of organic compounds, protein structures, membrane patches or ionic solutions in
physiological conditions or subjected to the action of external agents such as thermal or electric
perturbations [4-7]. Recently authors have implemented a procedure to introduce a static
homogeneous magnetic field in the Gromacs software, one of the most used environments for MD
[8].
In this work, using MD simulations, we want to show possible effects of magnetic fields on the
behavior of A2A adenosine receptor, modeled as a free protein in a typical buffer solution.
Simulations of more than 60 ns have been performed either in no field condition or with the

531
application of the field.
Materials and Methods
MD simulations have been performed in NVT (number of particles, volume and temperature
constants) ensemble at a temperature of 300 K using the Nose-Hoover thermostat in a box of 12 x
12 x 12 nm3 of dimension and with an integration step of 1 fs (Fig. 1). The gromos45 force field has
been used simulating 114 Na+ and Cl- ions (reproducing the DMEM concentration of 110mM) and
57577 water molecules. For the adenosine receptor A2A molecular model, the 3PWH PDB structure
has been chosen, but in the unbound configuration in order to firstly analyze the protein behavior
with no ligand (see Fig. 1a). The adenosine receptor A2A molecule was immersed in the NaCl buffer
solution and because of its positive charge (+6e) 6 more Cl- ions were added to the simulations box
to neutralize the system (Fig. 1b). To implement the magnetic field, we employed the Velocity Verlet
(VV) algorithm firstly proposed in [9], in which the Lorentz force acts on the charged particles, which
perform Larmor oscillations at the Larmor frequency when an external magnetic field is applied.
Since we used in our simulations a time step sufficiently small, we adopted the inversion algorithm
to solve the equations of positions and velocities [9].
Results
As a first verification, we performed MD simulations considering two different charged particles (Na+
and Cl-) in vacuum, to verify the cyclotron motion of the ions when an external magnetic field is
applied. We applied an intensity of B = 105 T in order to be able to see the particle motion in the
picosecond time scale and in the nm spatial scale, with a time step of 0.1 fs, for a 35 ps of
simulation. The initial velocity of the particle has been settled to v= 0.37 nm/ps. Figure 2 shows the
motion of the Na and Cl ions during the 35 ps of simulation: the charge particle Na+ describes a
clockwise ellipsoid (Fig. 2a) going in the z direction along the B field application, while the charge
particle Cl- shoes a counterclockwise motion (Fig. 2b). In Figure 2c the Na+ motion projected in the
z plane is reported showing a pitch of the motion h equal to 2.9 nm in accordance with the analytical
value.
After a 100 ns of equilibration, molecular simulations have been performed both with no field applied
and with an intensity of the B field equal to 1T, for a total duration of 60 ns. No statistically significant
variations have been observed for the secondary structure of the receptor neither for hydrophobic
and hydrophilic hydrophobic solvent accessible surface area, meaning that the protein structures
itself is not affected by the field (data not shown).
However, a finer analysis has been conducted selecting some specific residues especially those
placed in the extracellular loop of the protein (see Fig. 1a), which seems to play a fundamental role
in ligand binding and receptor activation [10]. Figure 3 shows the behavior of the single components
of the dipole moment of one of the residues of the extracellular loop, Cysteine 71 (CYS71). It can be
noticed looking at Fig. 3a and 3b, that both the x and z components of the dipole moment M show
an appreciable different time course when the B field is applied; moreover, when representing the
three time components in a 3D space (Fig. 3c) it is evident how an overall difference in the dipole
residue is shown with B=1T applied. Such preliminary results should be tested over longer
simulations time in order to assess statistical significance.
Conclusions
In this paper, we used the implementation of a static external magnetic field based on the
introduction of the Lorentz force into the VV algorithm, inside the Gromacs software, in order to
explore the effects of a magnetic field on the A2A receptor molecular model. First results of 60 ns
simulations show a slight effect on specific charged residues of the protein, while no structural
effects on its overall conformation.
References
[1] Katia Varani, Stefania Gessi, Stefania Merighi, Valeria Iannotta, Elena Cattabriga, Susanna
Spisani, Ruggero Cadossi, and Pier Andrea Borea, “Effect of low frequency electromagnetic fields

532
on A2A adenosine receptors in human neutrophils”, British journal of pharmacology, 136 (1):57-66,
2002
[2] Katia Varani, Fabrizio Vincenzi, Annalisa Ravani, Silvia Pasquini, Stefania Merighi, Stefania
Gessi, Stefania Setti, Matteo Cadossi, Pier Andrea Borea, and Ruggero Cadossi “Adenosine
receptors as a biological pathway for the anti-inflammatory and beneficial effects of low frequency
low energy pulsed electromagnetic fields”, Mediators of inflammation, 2017, 2017:2740963. doi:
10.1155/2017/2740963.
[3] Fabrizio Vincenzi, Martina Targa, Carmen Corciulo, Stefania Gessi, Stefania Merighi, Stefania
Setti, Ruggero Cadossi, Mary B Goldring, Pier Andrea Borea, and Katia Varani. Pulsed
electromagnetic fields increased the anti-inflammatory effect of a2a and a3 adenosine receptors in
human t/c-28a2 chondrocytes and hfob 1.19 osteoblasts. PLoS One, 8(5): e65561, 2013.
[4] Apollonio, F., Liberti, M., Amadei, A., Aschi, M., Pellegrino, M., D'Alessandro, M., D'Abramo, M.,
Di Nola, A., d'Inzeo, G “Mixed quantum-classical methods for molecular simulations of biochemical
reactions with microwave fields: The case study of myoglobin”, (2008) IEEE Transactions on
Microwave Theory and Techniques, 56 (11), art. no. 4657388, pp. 2511-2519.
[5] Reale, R., English, N.J., Garate, J.-A., Marracino, P., Liberti, M., Apollonio, F, “Human aquaporin
4 gating dynamics under and after nanosecond-scale static and alternating electric-field impulses: A
molecular dynamics study of field effects and relaxation” (2013) Journal of Chemical Physics, 139
(20), art. no. 205101.
[6] Paolo Marracino, Francesca Apollonio, Micaela Liberti, Guglielmo dInzeo, and Andrea Amadei,
“Effect of high exogenous electric pulses on protein conformation: myoglobin as a case study”, The
Journal of Physical Chemistry B, 117(8):2273-2279, 2013.
[7] Laura Zanetti-Polzi, Paolo Marracino, Massimiliano Aschi, Isabella Daidone, Antonella Fontana,
Francesca Apollonio, Micaela Liberti, Guglielmo D'Inzeo, Andrea Amadei, “Modeling triplet flavin-
indole electron transfer and interradical dipolar interaction: A perturbative approach”, Theoretical
Chemistry Accounts 132(11), 1-10, 2013.
[8] Elena della Valle, Paolo Marracino, Stefania Setti, Ruggero Cadossi, Micaela Liberti, Francesca
Apollonio, “Magnetic molecular dynamics simulations with Velocity Verlet algorithm", 32nd URSI
GASS Conference, Montreal,19 - 20 August 2017.
[9] Q. Spreiter, M. Walter, “Classical Molecular Dynamics Simulation with the Velocity Verlet
Algorithm at Strong External Magnetic fields”, J. Comput. Phys., 152, February 1999, pp. 102-119,
doi:10.1006/jcph.1999.6237
[10] Peeters MC, van Westen GJP, Li Q, IJzerman AP (2011) Importance of the extracellular loops in
G protein-coupled receptors for ligand recognition and receptor activation. Trends Pharmacol Sci
32:35–42
Figures

533
Figure 1. Molecular models used in the simulations. (a) Adenosine receptor model from the
Protein Data Bank (PBD) with indication of the extracellular loop; (b) Simulation box with
Adenosine receptor model solvated in a ionic solution environment.

Figure 2. Test of the MD algorithm; (a) trajectory for a charge (Na), with velocity v, under the
effect of the B field (z direction); (b) trajectory for a negative charge (Cl), with velocity v, under
the effect of the B field (z direction); (c) Na+ motion projected on the z axis showing a pitch of
the motion h equal to 2.9 nm.

Figure 3. Time behavior of the dipole moment of the residue CYS71, for the x component (a)
and z component (b) of the dipole moment; (c) 3-D representation of the dipole moment
components of the CYS71 residue in presence and in absence of the magnetic field.

534
PA-121 [14:30]
STUDENT PAPER
Steepest-entropy-ascent quantum thermodynamic approach to scaling the electric field
parameters and criticality related to cell signaling in electrically perturbed cells:
Experimental evidence and rationale
Ishan Goswami1, Scott S. Verbridge2 & Michael R. von Spakovsky1
1Mechanical Engineering, Virginia Polytechnic Institute and State University, Blacksburg, Virginia, USA,
24061
2Biomedical Engineering and Applied Mechanics, Virginia Polytechnic Institute and State University,
Blacksburg, Virginia, USA, 24061
Keywords: Mechanistic/Theoretical, Pulsed, Work in Progress
Presented by: Ishan Goswami
A successful coupling of standard cancer treatment modalities with electric field-based tumor
ablation techniques requires a general understanding of the mechanism of action of each modality
to evaluate synergistic effects for better treatment efficacies. The caveat, however, is that the
electric field parameter space and the molecular pathways are so large that understanding the
mechanism of action of electric fields via exploring all possible combinations is prohibitive from a
cost and time standpoint. To solve this parameter space problem, this work introduces a new
computational approach that takes advantage of thermodynamic scaling laws applicable to bio-
membranes. We report experimental evidence and the rationale behind this approach.
Introduction
The use of electric fields, be it in the form of low-amplitude tumor treating fields or high-amplitude
pulsed electric fields, is receiving increased traction as an attractive tumor treatment modality to be
used in conjunction with standard treatment regimens. The attractiveness lies in the flexibility of the
modality whereby fine-tuning field parameters (e.g., amplitude, pulse-width) leads to distinct effects
on treated cells. For example, fine tuning the pulse width was shown to target cells with a higher
nuclear to cytoplasmic ratio that is a hallmark of malignant cells. This boon of flexibility accompanies
the bane of not knowing the exact mechanism by which electric fields can target cells and cell
signaling due to the vast parameter space, which makes performing experimental mechanistic
studies untenable. Answers to the mechanistic influence of electric fields and electromagnetic fields,
in general, are essential for FDA approval and broader clinical use of the modality. Thus, a
computational route is needed to solve this vast parameter space problem.
To address this gap in knowledge, we propose a computational approach which derives from the
fundamental findings of Keller and colleagues [e.g. 1] that the physical laws of Ising universality
classes and the scaling laws associated with them can be used to explain how lipid bio-membranes
form “domains” enriched with specific lipid and protein species. The formation and distribution of
these domains influences cellular signaling and cell behavior. Exogenous electric fields can also
influence the lateral reorganization of these domains [e.g. 2]. Therefore, a fundamental approach
may be taken whereby one can use scaling laws to study the effect of electric fields on lipid
reorganization represented by an Ising lattice such that electric field parameters may be scaled. A
scaling law approach may extend the understanding developed in one regime of parameter space
(e.g., that of microsecond pulses) to another regime (e.g., that of nanosecond pulses).
In this article, an experimental study on mammalian cells exposed to clinically used microsecond
pulsed electric field (μsPEF) parameters is reported to argue that a reductionist approach of
studying lateral reorganization in membranes under the influence of electric fields may be the first
appropriate step needed to understand the mechanistic causes. Second, classical methods to solve
the Ising Hamiltonian are unable to temporally track the system, i.e., it is not possible to see the
evolution of the lattice in real time but instead only in statistical time. Therefore, an approach to
solve the Ising Hamiltonian in such a way that a temporal perturbation such as a pulsed

535
electric/magnetic field can be incorporated into the time evolution of the state of the lattice and be
done in real time. To this end, we propose the steepest-entropy-ascent quantum thermodynamics
(SEAQT) framework [3–9] developed by Beretta, von Spakovsky, Li, and colleagues for the
theoretical basis of our modeling approach. Preliminary data demonstrating the use of this
framework to solve for the non-equilibrium evolution of state of an Ising lattice are presented.
Results and Discussion:
Experimental evidences that serve as rationale to study lipid reorganization:
Goswami et al. [10] report that a µsPEF can inhibit cancerous signaling pathways used by triple
negative breast cancer cells (4T1 and MDA-MB-231) to influence the immune system. Moreover, the
PEF had a unique effect on the signaling in comparison with other conditions (thermal shock,
permeabilization alone via detergent treatment) that may accompany electric field exposure. On the
back of this finding, it was hypothesized that the PEFs could influence cell signaling via two routes.
First, PEFs could induce reorganization of the lipid bilayer membrane that has downstream effects
in the signaling pathways. Second, membrane permeabilization due to PEFs and subsequent
exposure of the cytoplasm to exogenous fields can perturb voltage sensitive cell signaling
organelles such as the mitochondria.
To test whether or not µsPEFs influence mitochondrial function, high resolution respirometry was
performed on 4T1 cells immediately after µsPEF exposure as well as after overnight incubation
following treatment. It was observed (data not shown) that µsPEFs alone did not alter mitochondrial
physiology, as was determined via the measurements of the electron transport chain state recorded
through responses to glutamate-malate and ADP and mitochondrial membrane potential via
responses to FCCP [11]. We highlight that contradictions do exist between findings on intact cells
and isolated mitochondria exposed to PEFs and, therefore question if the differences are due to the
complex interactions, both structurally and functionally, that the mitochondria inside the intact cells
have with other cellular structures such as the cytoskeleton on which each mitochondrion is
anchored. Upon compromising the actin cytoskeleton via treatment with drug Latrunculin B (Figure
1A), mitochondrial membrane potential in 4T1 cells were affected upon treatment with a µsPEF as
measured via responses to FCCP (Figure 1B). Therefore, a link exists between the two routes
proposed, whereby membrane reorganization, which also impacts the cytoskeleton, could lead to
downstream effects on the organelles. Note that reorganization events are influenced by membrane
criticality, i.e., local composition and temperature that determine whether or not a part of the
membrane is in a ordered or disordered phase (Figure 1C). Changing the criticality, for example, by
depletion of cholesterol using methyl-β-cyclodextrin (MBCD), leads to adherent cells being more
prone to the PEF (Figure 1D).
We surmise that lateral reorganizations may hold the key to developing a coarse-grained
mechanistic understanding of how PEFs influence cell behavior, although the contributions via other
mechanisms cannot be discounted. However, it is impossible to perform either experimental or
computational investigations using all combinations of the electric field parameter space. A
fundamental route may be taken by studying the effects of electric fields on lipid criticality
represented by an Ising lattice (Figure 2A). Such an approach allows electric field parameters to be
scaled, leading to an extension of the understanding developed in one regime of parameter space
(e.g., that of µsPEFs) to another (e.g., that of nsPEFs).
Demonstration of SEAQT as a candidate theoretical framework for bio-membranes:
A bio-membrane may be represented by an Ising lattice of spins up (+1, ordered) and spins down
(-1, disordered). Above a critical temperature (corresponding to the miscibility temperature of the
lipids), spins are well mixed as is observed at equilibrium and indicated by the absolute value of the
ensemble-averaged magnetization being close to zero (e.g., kT/J = 3.5; Figure 2B). Around the
critical temperature (kT/J=2.2; Figure 2B) spatial correlations develop and phase segregation is
observed. Below the critical temperature, spins are segregated as indicated by the absolute value of
the ensemble-averaged magnetization being close to one. In this work, it is shown that the
measurement of the deviations of the spatial distribution of spins from the Poisson distribution, as

536
measured by the L-statistic derived from Ripley’s K function, can distinguish critical compositions
(i.e., lattices with spin distributions observed at criticality) from non-critical compositions (Figure 2C).
Therefore, a L*-statistic, independent of whether the system is at equilibrium or non-equilibrium, can
be defined as a deviation from the critical composition, i.e., if L* is zero, the lattice is at a non-critical
state and otherwise has spatial correlations indicative of criticality.
It turns out that for an Ising lattice of N nodes, there are 2N possible lattice configurations (or phase
distributions) but only N-1 energy eigenlevels (i.e., the energy of a given lattice configuration), which
means that the eigenstructure of the lattice is highly degenerate and evaluation of L* for each
configuration is impossible. By employing a Wang-Landau density of state method, this investigation
predicts the degeneracy related to each eigenlevel and moreover proves that, like the degeneracy,
the L*-statistic is unique to a given eigenlevel irrespective of the eigenvector (i.e., the lattice
configuration; Figure 2D) associated with the eigenlevel. To prove the value of evaluating the L*-
statistic, this investigation calculates the ensemble-averaged L*-statistic at equilibrium for different
temperatures. The contours of L* in the phase plane of the characteristic length and temperature
experiences a sharp peak near the criticality point (Figure 3A-B), indicative of eigenvectors
corresponding to criticality contributing more in this region.
The scientific question to evaluate is whether or not a scaled PEF pulse parameter is able to
influence the criticality of a bio-membrane as evaluated by the L*-statistic and thereby bring about
lateral reorganizations. To answer this question, a candidate computational framework must be
capable of incorporating non-equilibrium perturbations of the lipid system and track thermodynamic
properties in real-time in a way that enables scaling of the electric field parameters. To this end, we
propose the SEAQT framework as the candidate framework since it allows tracking of system
evolution in real time (as opposed to the statistical time seen in Monte-Carlo techniques) and does
not require the massive computational burden inherent to kinetic Monte-Carlo or molecular dynamic
simulations.
The SEAQT framework [3–9] can be used to predict the non-equilibrium evolution of state of the
lipid Ising lattice interacting with an electric/electromagnetic field. To demonstrate the concept, we
use the approach proposed by Beretta [12] to generate an initial non-equilibrium state. The
relaxation of a simple four-level system towards equilibrium using the SEAQT framework is
illustrated in Figure 3C, which shows the transient evolution of the entropy of the system and the
probabilities associated with the energy eigenlevels (Figure 3D). Once non-equilibrium probabilities
are known, L*-statistics can be easily obtained to evaluate if the system is brought towards or away
from criticality. Detailed analyses in relation to bio-membranes will be presented at the conference.
Conclusion:
The SEAQT framework is expected to provide an elegant framework for parameter space
exploration within a reasonable time-frame by leveraging the scaling laws applicable to bio-
membranes. This in turn can be used to guide more detailed computational and experimental
analyses to develop a mechanistic understanding of how PEFs or even other electromagnetic fields
impact cellular machinery.
References:
[1] S.L. Veatch, P. Cicuta, P. Sengupta, A. Honerkamp-Smith, D. Holowka, B. Baird, ACS Chem.
Biol. 3 (2008) 287–293.
[2] J.T. Groves, S.G. Boxer, H.M. McConnell, J. Phys. Chem. B 104 (2000) 119–124.
[3] G.P. Beretta, O. Al-Abbasi, M. von Spakovsky, Phys. Rev. E 95 (2015) 042139.
[4] G.P. Beretta, Phys. Rev. E 90 (2014) 042113.
[5] S. Cano-Andrade, G.P. Beretta, M.R. von Spakovsky, Phys. Rev. A 91 (2015) 013848.
[6] G. Li, M.R. von Spakovsky, Phys. Rev. E 93 (2016) 012137.
[7] G. Li, M.R. von Spakovsky, C. Hin, Phys. Rev. B 97 (2018) 024308.

537
[8] G. Li, M.R. von Spakovsky, Energy 115 (2016) 498–512.
[9] G. Li, M.R. von Spakovsky, Phys. Rev. E 94 (2016) 032117.
[10] I. Goswami, S. Coutermarsh-Ott, R.G. Morrison, I.C. Allen, R.V. Davalos, S.S. Verbridge, L.R.
Bickford, Bioelectrochemistry 113 (2017) 42–50.
[11] I. Goswami, J.B. Perry, M.E. Allen, D. Brown, M.R. Von Spakovsky, S.S. Verbridge, Rev. (n.d.).
[12] G.P. Beretta, Int. J. Thermodyn. 9 (2006) 117–128.
Figures

Figure 1. Treatment with Latrunculin B compromises the actin cytoskeleton as shown in the
fixed cell staining image (A) of the actin (red) and the nucleus (blue). An A10 μm scale bar is
shown. Shown in (B) is the step rate data for cells with no treatment (control), those treated
with the Lan B vehicle DMSO, those treated with Lan B, and those with PEF exposure post the
Lan B treatment. The lipid domains can be visualized via electron microscopy. Shown in (C) is
the distribution of protein receptor CD44 that is localized in the ordered lipid domains. Altering
the cholesterol in cells lead to altered cell response to the PEF in a dose dependent manner of
MBCD(D). The error bars represent the standard error of the mean. Treatment conditions
connected by different letters within each figure are statistically different as determined by a
one-way ANOVA followed by a Tukey HSD-post test

538
Figure 2. The lipid domains on membranes are highly ordered regions. A visualization of the
distribution of CD44 protein receptors localized in these ordered regions is shown in an
electron microscopy image (A). This distribution of ordered and disordered phases can be
represented using a computational Ising lattice of spins up (black) and spins down (white)
shown alongside the microscopy image in (A). Predictions obtained via the Metropolis scheme
on an Ising Hamiltonian describes bio-membrane behavior and explains phase behavior on
either side of the scaled critical temperature of 2.2 as shown in (B). Critical lattice
configurations have a distinct L-statistic obtained from Ripley’s K function when compared to
non-critical ones (C). The distribution of the L*-statistic for each energy eigenlevel is shown in
(D).

539
 Figure 3. The evaluation of the ensemble-averaged L* statistic is demonstrated in (A). The
ensemble-averaged value peaks around the critical point (~2.2) indicative of spatial
correlations. The area under the curve of positive L* and the characteristic length can be used
to evaluate an area statistic A*, which upon an ensemble averaging also serves as a statistic
to track spatial correlations (B). Also provided in (B) are the estimates of heat capacity per
spin indicative of fluctuations that peak at the critical point. Alongside these statistics, the <L*>
for a characteristic length of 0.1 fraction of the lattice length is shown in (B). Illustration of the
SEAQT approach in predicting the time evolution of a system’s eigenenergy probabilities and
entropy and entropy generation rate are given in (C) and (D), respectively.

PA-123 [14:30]
Down-regulated miR-30a induced promotes neuronal autophagy by activating the AMPK
signaling pathway after microwave exposure
Ruiyun Peng1, Yanhui Hao1, Wenchao Li1 & Li Zhao1
1Department of Experimental Pathology, Institute of Radiation Medicine, beijing, China, 100850
Keywords: Mechanistic/Theoretical, RF/Microwaves, Completed (unpublished)
Presented by: Ruiyun Peng
miR-30a is involved in the pathogenesis and progression of multiple diseases by affecting
autophagy, however, the mechanism under the regulation of autophagy by miR-30a has not been
elucidated. In this study, the regulation mode of autophagy by miR-30a in neurons after microwave
exposure was explored. The result found that microwave exposure reduced the expression of
miR-30a in neurons, which could promote autophagy via activating the AMPK signaling pathway.
miR-30a is involved in the pathogenesis and progression of multiple diseases by affecting
autophagy. And we previously found that miR-30a in rat hippocampus was significantly down-
regulated after 30 mW/cm2microwave exposure. Therefore, as a non-invasive means, whether

540
microwave could regulate autophagy via miR-30a is of great interest to us. However, the mechanism
under the regulation of autophagy by miR-30a has not been elucidated.
In this study, the regulation mode of autophagy by miR-30a in neurons after microwave radiationwas
explored.
First, we exposed rats to 30 mW/cm2microwaveand found that miR-30a of hippocampal neurons
decreased from 7 d to 1 m after exposure. By KEGG analysis of miR-30a target genes, miR-30a
might affect autophagy through AMP activated protein kinase (AMPK). Interestingly, autophagy of
hippocampal neurons was activated from 7 d to 1 m after exposure and phosphorylated AMPK
increased from 7 d to 1 m after exposure.
Then, we set up an in vitro model in PC12-derived neuron-like cells, in which 30 mW/cm2microwave
down regulated miR-30a and activated autophagy at 6 h after exposure andincreased
phosphorylated AMPK was also detected. Furthermore, miR-30a overexpression by mimics
inhibited the occurrence of autophagy in neuron-like cells. And the decreased phosphorylation of
AMPK mediated by miR-30a overexpression was observed, indicating the negative effects of
miR-30a on the AMPK signaling pathway.
Conclusion: microwave exposure reduced the expression of miR-30a in neurons, which could
promote autophagy via activating the AMPK signaling pathway.
Kaywords: microwave;neuron;miR-30a; autophagy; AMPK.

PA-125 [14:30]
Experimental platform for measurement of biological response of cells to weak low
frequency magnetic fields
Michal Teplan1, Martin Bereta1, Ivan Bajla1, Yvonne Haba2, 3 & Michal Cifra4
1Institute of Measurement Science, Slovak academy of sciences, Bratislava, Slovakia (Slovak Republic),
84104
2Department of Orthopaedics, Biomechanics and Implant Technology Research Laboratory Rostock,
University, Rostock, Germany
3MSF/Chair of Ocean Engineering, University of Rostock , Rostock, Germany
4Institute of Photonics and Electronics, Czech Academy of Sciences, Prague, Czech Republic
Keywords: Mechanistic/Theoretical, ELF/LF, Work in Progress
Presented by: Michal Teplan
The goal of our research is to design a complex experimental platform enabling efficient scanning
through magnetic field parameters, while searching for specific response of investigated biosystem.
The methods are focused on the impedance spectroscopy in order to determine the response in
growth rate of cell cultures as well as in the electrical characterization of cell structures.
Introduction
Effects of weak low frequency (LF) magnetic fields (MF) on biological matter are not fully
established. Findings seem to be isolated, comparison between different reports is contradictory and
inconclusive [1]. Some of the difficulties are yield variability in biological experiments together with
experiment time costs. Thus, results are not only poorly replicated, but also sparse in the space of
MF parameters like frequency and amplitude. On the other hand, existing interaction models of
external MF with biological structures are lending legitimacy for experimental investigations.
Possible effects may vary from strong resonant behavior to mild variations in terms of stimulation or
inhibition. Research challenges are still open, with potential applications in diagnostics, therapy as
well as in industry.
The goal of this research is to design a new experimental approach enabling a more efficient
scanning through MF parameters, while searching for specific response of the investigated
biosystem. With this aim, a complex experimental platform for quantitative characterization of the
541
biological response to weak LF MF is under development.
Methods
A cultivation chamber provides suitable and stable conditions for yeast Saccharomyces cerevisiae
cells’ growth in aqueous media. The cell culture is cultivated in quasi-homogenous MF in media
solution placed in Helmholtz coils. The coils generate MF with frequency 0-2 kHz and intensity up to
2 mT. To move cell samples through different measurement stations, a peristaltic circulation pump is
used. At several measurement stations different physical quantities are measured. The cell density
is monitored by flow turbidimetry. Inverted light microscopy provides new monitoring tools in a form
of quantification of cell motility. Moreover temperature, pH, DO, CO2, and nutrient media properties
are monitored as well.
From all physical methods, we are focused on the use of electroimpedance spectroscopy (EIS). EIS
is measured by portable Stemlab and is incorporated in the experimental platform for two-fold use:
For monitoring of cell growth curve and for monitoring of cell electrical parameters.
Results
An innovative tool for measurement of cell motility is being developed. Cells are segmented by
automatic image processing of single video frames obtained from digital camera. A tracking
procedure of cells throughout the frames enables monitoring of single cell behavior. From sample
recording app. 90 cells were located. Their motility was characterized by the following measures:
cell area, cell border length, centroid position in both X and Y direction. These quantities in a form of
time series are further processed by Fourier transform. The obtained spectra reflect periodicity in
cells’ vibrational movement. For the camera sampling rate of 18 Hz, different cells yielded different
peaks in spectra in the range of 0-9 Hz.
For pair experiments, EIS is designed parallely at three stations: directly in the growth chamber with
the use of conductivity cell, by the probe inserted in between tubing, and finally in the EIS microchip
(Micronit). The chip chamber size of 28×30 µm enables single cell EIS under the control of light
microscopy. For the flow application we are developing 3D-printed probe consisting of plastic body
and Ag electrodes. The methodological steps for exploitation of EIS include assessing of geometry
factors for different electrode configurations, extraction of conductivity and permittivity from
impedance measurements for pure nutrient media and for cell solution at different frequencies.
Modelling of electrical field distribution during EIS measurement by numerical FEM analysis in
Comsol Multiphysics is also in development. With the use of equivalent electric circuit model, the
estimation of cell electrical properties like membrane resistivity and capacitance will be enabled.
Discussion and Conclusions
Developed cell motility measures should reflect viability and metabolic activity of cells. Statistical
characterization over sufficiently large cell samples will enable to study subtle changes of cell
behavior under the influence of MF.
EIS sensitivity analysis is under development, both experimentally and analytically. Spherical model
of cells with varying concentration [3,4] and varying electrical properties of subcellular structures like
resistivity and capacity of the cell membrane together with cytosol resistivity [2] are involved.
The main challenges of this research are: 1) finding the weakest MF causing observable effects, 2)
finding the shortest time scale for detecting the effects, and finally, 3) finding frequency and
amplitude windows for biological effects of LF MF.
Acknowledgements
The study was supported by the research grant 2/0138/16 from the VEGA Grant Agency, Czech
Science Foundation, project no. P102/15-17102S, bilateral exchange project between Slovak and
Czech Academies of Sciences, no. SAV-18-11. Authors also participate in COST Actions BM1309
and CA15211 andthefund for theadvancement of women inscienceat the University of Rostock.
References

542
[1] Buchachenko A. Why magnetic and electromagnetic effects in biology are irreproducible and
contradictory?, Bioelectromagnetics, vol. 37, pp. 1-13, 2016.
[2] Valero A. A unified approach to dielectric single cell analysis: Impedance and dielectrophoretic
force spectroscopy. Lab on chip, 10(17), 2216-2225, 2010.
[3] Foster KR and Schwan, HP. Dielectric Properties of Tissues and Biological Materials: A Critical
Review. Critical Reviews in Biomedical Engineering, 17: 25-104, 1989.
[4] Asami K.,Yonezawa T. Dielectric analysis of yeast cell growth. Biochim Biophys Acta, 1245(1):
99-105, 1995.

PA-127 [14:30]
EMF exposures near the electrical gastronomy devices – low and radiofrequency evaluation
Krzysztof Gryz1, Jolanta Karpowicz1 & Patryk Zradziński1
1Central Institute for Labour Protection - National Research Institute (CIOP, Warszawa, Poland, 00-701
Keywords: Occupational, All Frequencies, Work in Progress
Presented by: Krzysztof Gryz
The aim is to identify and evaluate combined EMF near electric gastronomy devices (common
private and business use) - ELF magnetic and RF electric components, both covered by IARC 2B
classification. Both components recorded near microwave ovens significantly exceed average level
of population exposure - ELF up to 10μT; RF up to 14V/m. Special attention during EMF
measurements and evaluation is needed for exposure time-pattern - devices of continues EMF
emission, as well as pulsed (keyed) were found. Results shown that professional activities near
electrical gastronomy devices may play dominant role in the EMF cumulative exposure.
Introduction
Today, variety of electrical devices is used in thermal processes in the field of gastronomy. They may
be found in private kitchens, social rooms in offices, hospitals, schools, etc., as well as in the
commercial gastronomic businesses. The devices used for private purposes and in the small
gastronomic businesses may be very similar (or the same type). Such devices always emit
electromagnetic field (EMF) of low frequency (ELF) related to electric supplying installation and
some internal circuits, and usually also EMF of higher frequencies from the kHz, MHz or GHz
frequency bands, depend on the thermic technology applied. Taking into account, that in many
cases such devices are located near the users body, their contribution to the EMF exposure needs
attention.
The aim of the presented investigations is identifying and evaluating the ELF magnetic field and RF
electric field components (both covered by 2B IARC classification) of combined EMF exposure near
the electric gastronomy devices of common use, for private and business use.
Methods
The investigations covered the recordings of root-mean-square (RMS) values of magnetic flux
density, B, and electric field strength, E. The investigations were worked out near (30 cm away) to
various types of electrical gastronomy devices. The ELF and RF components of exposure have
been measured using:
− recordings by data logger EMDEX II Standard (Enertech Consultans, Campbell, CA, USA) –
measurement range of magnetic flux density (0.01–300) μT in the frequency band (40–800) Hz,
sampling rate 1.5 s
− recordings by data logger EME SPY 121 frequency selective (Satimo, Brest, France) -
measurement range of electric field strength (0.05-10) V/m in the frequency band (80-2500) MHz,
spread into 12 pre-defined frequency measurement ranges, corresponding to the most common RF-
EMF applications currently in use in the public environment, sampling rate 4 s

543
− spot measurement of E-field spatial distribution by broadband field meter NBM-550 (Narda Safety
Test Solutions, Pfullingen, Germany) with probe EF0392 – measurement range of electric field
strength (0.8- 1300) V/m in the frequency band 100 kHz – 3 GHz
− waveform identification by scopemeter Fluke 190-104 (Fluke, USA), with a Fast Fourier Transform
(FFT) analyze option for ELF or by Selective Radiation Meter SRM-3006 (Narda Safety Test
Solutions, Pfullingen, Germany) for RF
Studying the EMF near the gastronomic devices the results of recordings from the predefined
frequency band WLAN/ WiFi 2G have been used to characterize microwave exposure at frequency
2.45 MHz using by microwave ovens.
The investigations of exposure cover the frequency composition, level and variability in time.
The criteria for measurement results covers variety of health and safety aspects: short term
exposure evaluation (workers, general public, sensitive groups – users of medical implanted devices
[1-5], as well as long term exposure evaluation, which consider the exposure level with respect to
the averaged level in population – usually below 1 μT @ power frequency (ELF component) and
below 1 V/m @ radiofrequency (RF component) [6].
Results and Discussion
Data collecting is ongoing. Both exposure components, ELF and RF, are evaluated near selected
gastronomy devices.
In order to standardize the measurement results, recordings were done while heating of approx. 200
ml and 1000 ml water, respectively in ovens (during 3 minutes) and kettles (during 2-5 minutes -
depended on device). Microwave ovens output power was set at 600-650 W, supplying power of
electric kettles was from the range 2000-2400 W. In both cases – typical parameters of the common
use devices of such type. Figure 1 presents example of the range of ELF and RF components of
EMF recorded near microwave ovens. Both components, ELF (up to 10 μT) and RF (up to 14 V/m),
significantly exceeded mentioned average level of the population exposure, as well as varied
significantly between devices. Various time-pattern of exposure were found - devices of continues
EMF emission, as well as pulsed (keyed) in time, what may need in consequences special attention
during EMF measurements and evaluation. The only formal exposure limits which overexposure
was found is 3 V/m (in RF range) limit used in the immunity tests for the medical electronic devices,
including medical implants. It suggest that population of implant users’ needs to be informed about
possible malfunctions of their devices near microwave ovens.
ELF exposure near kettles was found at lower level, however also varying significantly between
devices – 0.1-2 μT, as well as in some cases exceeding averaged level of the population exposure.
Conclusions
Results of the study shown that in case of professional activities near electrical gastronomy devices,
their use may play dominant role in the EMF cumulative exposure. In case of vulnerable population
of medical implant users, such exposure needs also attention. Further investigations are ongoing in
order to evaluate EMF exposures near wider group of gastronomy devices.
Acknowledgments
Research supported in Poland within the National Programme “Improvement of safety and working
conditions” (2017-2019) - within the scope of state services - by the Ministry of Family, Labour and
Social Policy, Poland (2.G.07). The CIOP-PIB is the Programme’s main co-ordinator.
References
[1]. European Recommendation 1999/519/EC; [2]. Directive 2013/35/EC; [3]. ICNIRP 2010. Health
Phys. 2010;99(6):818-836; [4]. ICNIRP 1998. Health Phys. 1998;74(4):494-522; [5]. IEC
60601-1-2:2014; [6]. SCENIHR 2015.
Figures

544
 Figure 1. The components of ELF (B-field) and RF (E-field) recorded near the microwave
ovens

PA-129 [14:30]
Single nucleotide polymorphisms in 5-HTR and GRIN2B associated with risk of cognition
dysfunction in electric workers?
Xiangjun Hu1, Lifeng Wang1 & Haijuan Li1
1Beijing Institute of Radiation Medicine, Beijing, China, 100850
Keywords: Occupational, ELF/LF, Completed (unpublished)
Presented by: Xiangjun Hu
There were some abnormal characterizations of the brain and cardiac function in workers engaged
in electric company for long time.The latency of N2 and P3 in TT genotype of rs6313, the N2 latency
and the contents of LENK in GG genotype of rs6295, the N2 latency in CC genotype of
ENS10557853 were prolonged, with the increase of exposure dosage respectively compared to CC,
CG and TT genotypes. 5-HT1AR in rs6295 G allele, GRIN2B in ENS10557853 C allele and
5-HT2AR in rs6313 T allele might increase the risk of cognition retardation in the electric workers.
Background
EMF could cause disorder in the occurrence of neurodegenerative disorders, cardiovascular
disease and cancers, whereas other studies have not confirmed these results. The CNS is a
potential target of interaction between biological systems and ELF-MF. Some studies found that
there were differences in the change of brain function between individuals after electromagnetic
radiation.
Methods
In this study, Health of workers in electric company was assessed by questionnaires, peripheral
blood cells, biochemical parameter, radioimmunity indexes, electrocardiogram (ECG) and the P300
event-related brain potential (ERP). The associations of eight single nucleotide polymorphisms
(SNPs) of5-HT1AR, 5-HT2AR, 5-HT2CR, GRIN2B, BDNF and COMT with harmful effects to the
health of workers in electric company were investigated.
Results
There was no significant abnormality in the questionnaire and routine physical examination

545
indicators including peripheral blood cells and biochemical parameter. Some abnormal
characterizations of ECG in operating and repairing groups were observed compared to control
group, which was shown as arrhythmia, bradycardia, T wave abnormalities, etc. The N2 latency of
P300 was prolonged and the contents of LENK were increased in repairing group (electric field
intensity≥50V/m, magnetic field strength≥10μT) compared to control and operating groups, the
contents of BNP in operating group (electric field intensity<50V/m, magnetic field strength<10μT)
were increased compared to control group. The latency of N2 and P3 in TT genotype of rs6313, the
N2 latency and the contents of LENK in GG genotype of rs6295, the N2 latency in CC genotype of
ENS10557853 were prolonged, with the increase of exposure dosage respectively compared to CC,
CG and TT genotypes.
Conclusions
There were some abnormal characterizations of the brain and cardiac function in workers engaged
in electric company for long time. 5-HT1AR in rs6295 G allele, GRIN2B in ENS10557853 C allele
and 5-HT2AR in rs6313 T allele might increase the risk of cognition retardation in the electric
workers.
Key Words: ELF-EMF; 5-HTR; GRIN2B; polymorphism; cognition
(This work supported by National Natural Science Foundation of China (61571455))

PA-131 [14:30]
A systematic review on biological effects of magnetic, electric and electromagnetic fields in
the intermediate frequency range (300 Hz–1 MHz)
Lambert Bodewein1, Kristina Schmiedchen1, Dagmar Dechent1, Dominik Stunder1, David Graefrath1, Lukas
Winter1, Thomas Kraus1 & Sarah Driessen1
1Research Center for Bioelectromagnetic Interaction (femu), University Hospital RWTH Aachen, Aachen,
Germany, 52074
Keywords: Public Health Policy, IF, Completed (published)
Presented by: Dagmar Dechent
The aim of this systematic review was to provide an update of the state of research on effects of
electromagnetic fields in the intermediate frequency range on biological systems. Fifty-six
experimental studies were eligible with only few human studies and no epidemiological study.
Included studies examined several in vivo and in vitro systems and different endpoints and found
inconsistent evidence for field effects. No indications for effects of specific applications using
intermediate frequency fields were found. Weak intensity fields and frequencies >100 kHz have
been scarcely investigated. Methodological limitations lowered the credibility of studies. Future work
should consider the identified shortcomings and gaps of knowledge.
Background
Many novel technologies, such as induction hobs or electric vehicles produce electromagnetic fields
in the intermediate frequency (IF) range. In comparison to research on the effects of extremely low
and radio frequency fields, the effects on biological systems of IF fields, however, have been poorly
investigated. The aim of this systematic review was to provide an update of the state of research
and to evaluate the possible effects of electric, magnetic and electromagnetic fields in the IF-range
(300Hz–1MHz) on biological systems.
Methods
The review was prepared in accordance with PRISMA (Preferred Reporting Items for Systematic
Reviews and Meta-Analyzes) guidelines [1]. Methodological limitations in individual studies were
assessed using the Office of Health Assessment and Translation (OHAT) Risk of Bias Rating Tool
for Human and Animal Studies [2].

546
Results
Fifty-six studies exposing humans, animals and in vitro systems were eligible for this review. Most of
the findings were obtained in studies with exposure to magnetic fields. Furthermore, many different
endpoints were examined, but for most of them, the reviewed studies yielded inconsistent evidence
for effects of exposure in the IF-range. While some studies indicated no effect of exposure, some
linked exposure with adverse effects or effects with unclear relevance on different endpoints.
Studies explicitly examining the potential for adverse health effects from induction hobs, inductive
charging and magnetic resonance imaging (gradient fields) did not find evidence for effects of
exposure on investigated endpoints, e.g., reproductive and developmental parameters. Most of the
included studies applied field strengths above the proposed ICNIRP reference levels for the general
public [3, 4] and potential effects for frequencies >100 kHz have been scarcely investigated, so far.
Furthermore, many of the reviewed studies suffered from methodological limitations which lowered
credibility in the reported results.
Conclusion
Due to the large heterogeneity in study designs, investigated systems and endpoints, the
inconsistent results as well as methodical limitations in many studies, evidence for biological effects
of IF fields remains unclear. Emphasis in future studies should be placed on effects on humans,
including epidemiological studies, and on investigating a larger spectrum of the IF range and field
strengths.
References
[1] Moher, D., Liberati, A., Tetzlaff, J., Altman, D. G. and The, P. G. (2009): Preferred Reporting
Items for Systematic Reviews and Meta-Analyses: The PRISMA Statement. PLOS Medicine, 6(7)
[2] National Toxicology Program (NTP), Office of Health Assessment and Translation (OHAT)
(2015): OHAT Risk of Bias Rating Tool for Human and Animal Studies.
[3] ICNIRP (1998) Guidelines for limiting exposure to time-varying electric, magnetic, and
electromagnetic fields (up to 300 GHz). Health Physics, 74(4)
[4] ICNIRP (2010) Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz
to 100 kHz). Health Physics, 99(6)

PA-133 [14:30]
Specific absorption rate reduction design of dual-band planar monopole antenna with
defected ground structure for WLAN applications
Min-Joo Jeong1, Ji Woong Park1, Byeong-Nam Kang1, Niamat Hussain1, Hanul Bong1 & Nam Kim1
1Chungbuk National University, Cheongju-si, Korea, 28644
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Niamat Hussain
In this paper, we proposed a planar monopole antenna with dual band characteristics that can be
used in 2.4 GHz band and 5 GHz band of WLAN. We also calculated the Specific Absorption Rate
(SAR) for 1 g tissue at 0 mm distance from head instead of 10 mm, which is stricter than the
international standard, and compared it with SAR values with and without reflector at 2.4 GHz and
5.8 GHz. Compared with the international standard 1 g tissue SAR value (1.6 W/kg), the proposed
antenna with reflector reduces about 86.38 % (@ 1.382 W/kg difference) at 2.4 GHz and 81.18 %
(@ 1.299 W/kg difference) at 5.8 GHz.
Introduction
Recently, because of the development of information communication technology and devices, the
wireless network field uses various frequency bands. Accordingly, wireless communication devices
require the use of multiple antennas or antennas capable of covering multiple frequency bands.

547
WLAN (Wireless Local Area Network) is mainly used in industrial fields as well as in homes. In this
paper, we propose a planar monopole antenna with dual band characteristics for 2.4 GHz band and
5 GHz band of WLAN and analyze the SAR. In addition, a reflection plate was installed under the
antenna to reduce the SAR value.
Methods
The proposed antenna composed of planar monopole antenna and a feed line to the front side, the
dual-band (WLAN) was implemented using DGS (Defected Ground Structure). To reduce the SAR
of the proposed antenna, a reflector was installed at the λ/4 from the antenna. The SAR at 2.4 GHz
and 5.8 GHz was compared and analyzed with and without reflector. In the simulation, the distance
between the head and the antenna is 0 mm and the SAR for 1 g tissue at input power of 1 W is
calculated at 2.4 GHz and 5.8 GHz. SAR analysis was performed using Sim4Life (SEMCAD-X), a
simulation software.
Results
The bandwidth of the proposed antenna is 2.01 - 2.62 GHz and 4.76 - 5.92 GHz. The size of the
antenna is 51 × 44 × 1.6 mm3. The antenna substrate was FR-4 having a relative dielectric constant
of 4.3 and a loss tangent of 0.025. To reduce the SAR of the proposed antenna, a reflector was
installed at the λ/4 from the antenna. The material of the reflector is PEC and the size is 70 x 70 x
0.017 mm3. The optimum distance between the antenna and the reflector is 41 mm. The bandwidth
of the antenna with reflectors is 2.17 - 2.88 GHz, 4.77 - 5.94 GHz. Because of comparing 1 g tissue
SAR of the proposed antenna, 24.5 W/kg (without reflector) and 0.218 W/kg (with reflector) at 2.4
GHz and 25.4 W/kg (without reflector) and 0.301 W/kg (with reflector) at 5.8 GHz.
Conclusion
In this paper, we propose a dual - band planar monopole antenna for WLAN applications. We also
compared the SAR values of 1g tissue with and without reflector at 2.4GHz and 5.8 GHz. The SAR
value was decreased around 99 % as compared to the SAR without reflector at both bands. This is
because the SAR value is reduced or shielded by the reflector. This result is also as low as about
86.38 % (@ 1.382 W/kg difference) for 2.4 GHz and about 81.18 % (@ 1.299 W/kg difference) for
5.8 GHz when compared to the international standard value of 1.6 W/kg for 1 g tissue SAR.
Therefore, proposed antenna with reflector can effectively reduce the SAR for WLAN applications.
Figures

Figure 1. Reflection loss of proposed antenna with and without reflector

548
Figure 2. Radiation pattern of proposed antenna with and without reflector

Figure 3. The average SAR for 1g tissue of proposed antenna with and without reflector,
placed at 0mm distance from head.

549
PA-135 [14:30]
Risk perception of manufacturing industry labors exposure to ELF-MF at work
M.D Rajitha Kawshalya1, Yun Jin Lee2 & Seung-Cheol Hong1, 2
1Emergency and Disaster management, Inje University, Gimhae-Si, Korea, 621-749
2Occupational Health & Safety Engineering, Inje University, Gimhae-Si, Korea, 621-749
Keywords: Public Health Policy, ELF/LF, Completed (published)
Presented by: M.D Rajitha Kawshalya
The questionnaire survey was conducted among the industrial labors to investigate the risk
perception of exposure to extremely low-frequency magnetic fields (ELF-MF) at the workplace. The
study subject was the foreign labors who currently occupied in Korea as labors under the employee
permits system by the Korean government. These workers are involved with heavy industrial work
that rejected by many locals to do due to the dangerous, dirty and difficulty. There have been very
fewer studies conducted to investigate the safety measurement of these works in non-ionizing
radiation. This study is to collect valuable information of perspective to the exposure of ELF-MF.
Introduction
This study focused on the labors who work in the industry such as die-casting, electroplating,
molding, welding etc. as shown in table 1. There is a possible risk for these labors to exposure in the
extremely low-frequency magnetic field (ELF-MF) in the workplace. The exposure to extremely low-
frequency field occurs whenever the electrical energy is used in a workplace. Due to the variety of
tasks in the working environments, ELF-MFis present and exposure invariably occurs. The basic
properties ofELF-MFare related to the wavelength and field polarization, characterized by so-called
near- or far-field conditions of exposure.ELF - MF in highly exposed work environments are usually
near fields. ELF-MFs were categorized as being possibly carcinogenic to humans (group 2B) by the
International Agency for Research on Cancer (IARC) and there is evidence that ELF-MF may be
associated with childhood leukemia (Zhao et al., 2014). Also, there is studies foreground that
prognosticatesthe association of psychological factors in the attributed to electromagnetic fields
(EMF) such as environmental worries, environmental stressor, anxiety, and depression. The main
research questions address were; Do the labors know what are they surrounded in the working
environment and what sort of possible risk they can be exposed to during the work?
Methodology
Study design and participants
The present study is performed within the framework of the risk perception of manufacturing industry
labors exposure to ELF-MF as a potential health risk. To investigate the matter questionnaire survey
was conducted among the labors. The survey was distributed among the labors in the
manufacturing industry as shown in table 1. According to the Korean standard statistical
classification, 6 categories of manufacturing industries were selected. The questionnaire was shared
through online in the southern part of the Korean peninsula.
Results and Discussion
As an ongoing study following results was based on the collected sample The initial target was to
collect 150 samples of the EPS workers but the respondent's samples were 110 with a 73.3% of
return rate. Among the survey, 100% respondents turn out to be male. The survey indicated highest
respondent of 22.98% (n=20) of labors were from the grinding sections while 20.68% (n=18) were
from construction labors.
Table 1. Demographic profiles of the survey respondent

Variable Nationality

550
 South Asian Southeast Asian East Asian
(n=87) (n=21) (n=2)

Gender

Male 87 (79.0%) 21 (19.0%) 2 (1.8%)

Age

20-25 19 (17.2%) 5 (4.5)

26-30 53 (48.1%) 13 (11.8)

31-35 15 (13.6%) 3 (2.7%) 2 (1.8)

Occupation

Electroplating 14 (12.7%) 6 (5.4%) 1 (0.9%)

Die-casting 16 (14.5%) 12 (10.9%) 1 (0.9%)

Molding 10 (9.09%) 1 (0.9%)

Welding 9 (8.1%)

Construction 18 (16.3%)

Grinding 20 (18.1) 2 (1.8%)

Upon the respondent, the figure 1 indicates 84.5% of labors were not aware of the ELF-MF emission
in their respective workplaces and uncertain about personal exposure to ELF-MF. Also up to now
100% of the participants answered none of the companies had measured ELF-MF exposure levels
in plants or acknowledge the potential exposure to ELF-MF of the labors at the workplace.Among
the participants, 70.90% of them answered that the dormitories are not located in the same building
as workplace while 25.50% of them answered they are living in the same place as the workplace as
indicated in figure 2. Due to 24 hours of plant operation, there is a possible risk to exposure to the
ELF-MF when living in the same building as a workplace due to the exposure emission from the
plant during the plant time spent in residence. Also, 69.10% of participants are uncertain about living
in the same building as the workplace is dangerous or not, due to lack of awareness about the
possible ELF-MF emission and there is a possible health risk may be occurred due to the chronic
exposure.
Questioner survey indicated the social and personal determinants of ELF-MF risk perception among
labors are lower and participants were somewhat exposed to the extent of occupational and
residential ELF-MF. The data from the survey indicated participated labors awareness as an
occupational exposure to ELF-MF as a possible health threat is low due to the lack of awareness.

551
Conclusion
As for the respondents, none of the workplaces has measured the exposure of ELE-MF at the
workplace.Understanding this, there is a social responsibility for both authorities and respective
companies to maintain safety standards for the labors in Korea and make awareness about their
working environments by conducting environmental impact surveys’. The current results indicate
another issue that living in the same place as workplace may cause to the long-term exposure of
ELF-MF to labors. This needs to study further to investigate potential long-term exposure by living in
the same place as working place.
References
R.Baan, Y.Grosse, B.Lauby-Secretan, F.El Ghissassi, V.Bouvard, L.Benbrahim-Tallaa, et al.
Carcinogenicity of radiofrequency electromagnetic fields. Lancet Oncol.,12(2011), pp.624-626
L.Zhao, X.Liu, C.Wang, K.Yan, X.Lin, S.Li, et al. Magnetic fields exposure and childhood leukemia
risk: a meta-analysis based on 11,699 cases and 13,194 controls. Leuk. Res., 38(2014), pp.269-274
U.Frick, J.Rehm, P.Eichhammer Risk perception, somatization, and self-report of complaints related
to electromagnetic fields—a randomized survey study Int. J. Hyg. Environ. Health, 205(2002), p.353
World health organization (WHO). 2004. Available at http://www.who.int/peh-emf/about/WhatisEMF
/en/index1.html.
KOSTAT- Statistic Korea: 2015.11.30
Figures

Figure 1. Awareness of ELF-MF among labors in manufacturing industry

552
 Figure 2. Labors living condition and ELF-MF exposure perception

PA-137 [14:30]
Development and operation of movable EMF experience camp for efficient risk
communication to general publics
Seungwoo Lee1, Geun Teak Yeo1, Kyungku Nah1 & Hong Lae Kim1
1Transmission & Substation Construction Dept., Korea Electric Power Corporation, Naju-si, Korea, 58217
Keywords: Public Health Policy, ELF/LF, Work in Progress
Presented by: Seungwoo Lee
We developed and managed new type of experience house which is called Movable EMF
Experience Camp for movable and easy access. The movable EMF experience camp which
consists of experience contents in the container house. The size of a camp is 7m x 3m x 2.5m, and
it contained 10 kinds of contents in 4 areas.We are operating the camp in the places of civil
complaint, public relations, construction of electric power facilities, etc.
To relieve anxiety of the electromagnetic fields radiated by the electric power facilities, KEPCO
provided and operating various activities for risk communication. However, the policy still lack the
belief, responsibility and acceptance to the general publics, and the risk communication of EMF has
limitations. In addition, the general publics do not trust the objectivity and justice of KEPCO who is a
stakeholder for construction and management of the electric power facilities for risk
communication.Most of all, the risk communication of government and academia for reducing the
anxiety from the electric power facilities are effective, but it needs high expense and effort. KEPCO
is one of the public corporations for the infrastructure, but also it is a kind of private company.
To prevent the complex and long-term problems for EMF of electric power facilities, we have to
introduce new risk communication activities as recommended in WHO. Consequently, the risk
communication should be based on the communication and trust. Previous risk communication is
one way message, but we need a special solution to reinforce the bidirectional
communication.Therefore, we developed and managed new type of experience camp which is
movable and easy access.
Figure 1 shows a movable EMF experience camp which consists of experience contents in the
container house. We are operating the camp in the places of civil complaint, public relations,
construction of electric power facilities, etc. The size of a camp is 7m x 3m x 2.5m, and it contained
10 kinds of contents in 4 areas.Figure 2 shows the contents in the EMF camp. All contents are very

553
interesting and could be experience by oneself. The experience devices are the principle of
electricity generation, electric and magnetic fields, the measurement of electric devices, the
comparison of transmission lines and electric devices, the kiosk of EMF information, melody chair
using micro current, etc.The list of contents and specific explanation of the EMF experience camp is
shown in Table 1.
Figures

Figure 1. The exterior and perspective image view of the movable EMF experience camp.

554
Figure 2. The installed experience contents in the camp.

Figure 3. List of contents in the EMF experience camp.

555
PA-139 [14:30]
Measurements of magnetic fields generated by home electric appliances
Kiyotaka Hayashi1, Shigeru Ootsubo1, Toshihisa Kadoya1, Chiyoji Ohkubo1 & Yukio Mizuno2
1Japan EMF Information Center, Japan Electrical Safety & Environment Technology Laboratories, Tokyo,
Japan, 105-0014
2Department of Electrical and Mechanical Engineering, Nagoya Institute of Technology, Nagoya, Japan,
466-8555
Keywords: Standards, ELF/LF, Completed (published)
Presented by: Kiyotaka Hayashi
We conducted measurements of magnetic flux density generated at more than one frequency by the
latest home electric appliances (total of 131 models of 42 products) within the frequency range of 10
Hz - 400 kHz.All the measurement results were well below the reference levels for the general
public of ICNIRP Guidelines (2010).
Introduction
Home electric appliances generate magnetic fields at various frequencies. The Association for
Electric Home Appliances in Japan (AEHA, 2013) [1] published the measurement results of
magnetic flux density generated by the appliances at several frequencies, not the actual measured
values but the proportional values to the reference levels for the general public established by the
International Commission on Non-Ionizing Radiation Protection (ICNIRP, 2010) [2], according to the
measurement procedures described in the IEC standard 62233 [3]. However, lay people tends to
difficult for understanding proportional values and we have been asked to show actual measured
values. In order to respond their request, we conducted measurements at more than one frequency
generated by the latest appliances.
Methods
We used Narda S.T.S. ELT-400 suitable for measurement of low frequency magnetic fields (10 Hz -
400 kHz) complying with JIS C 1912 [4], a Japanese standard identical to IEC 62233, by conducting
time-domain evaluation. We selected generally used 131 models of 42 products (about 3 models
per product; each model was most popular according to web research in Japan) such as TV,
washing machine, vacuum cleaner, PC, heater, lighting equipment, etc. Measurements were taken
around the appliances, and at the point the highest magnetic flux density was found, detailed
measurements were conducted at 0 to 50 cm from the appliances with 10 cm interval.
Results and Conclusions
An example of the measurement results of magnetic fields generated by home electric appliances is
shown in figure 1. The graph shows the range (the lowest and the highest models) of magnetic flux
density (RMS) by each appliance measured at the point with the highest value. (Measurement
distance was 0 cm for laptop computer, electric toothbrush, electric carpet, electric blanket, hot
water washing toilet seat; 10 cm for hair dryer; 30 cm for others). We measured magnetic fields
generated by total of 131 models of 42 products of the latest home electric appliances around us. All
the measurement results were well below the reference levels for the general public established by
ICNIRP (200 μT at 50 Hz; 27 μT at 30 kHz).
Acknowledgement
We gratefully appreciate Professor Emeritus Katsuo Isaka, Tokushima University, and Dr. Ken-ichi
Yamazaki, Deputy Associate Vice President, Central Research Institute of Electric Power Industry,
for their supervisions and guidance to our measurements.
References
[1] Association for Electric Home Appliances (2013). Measurement survey of electromagnetic wave
(10 Hz - 400 kHz) generated by the home electric appliances, Fiscal Year 2013 (available in

556
Japanese only).
[2] International Commission on Non-Ionizing Radiation Protection (2010). Guidelines for Limiting
Exposure to Time-Varying Electric and Magnetic Fields (1 Hz - 100 kHz). Health Phys
99(6):818-836.
[3] International Electrotechnical Commission (2005). IEC 62233:2005. Measurement methods for
electromagnetic fields of household appliances and similar apparatus with regard to human
exposure.
[4] Japanese Standards Association (2014). JIS C1912:2014. Measurement methods for
electromagnetic fields of household appliances and similar apparatus with regard to human
exposure (available in Japanese only).
Figures

Figure 1. Magnetic flux density by the types of products

PA-141 [14:30]
Enhanced method for caculating the area of protection in the vicinity of radio base stations
Werner Wiesbeck1, Mario Pauli3, Thorsten Kayser3 & Josef Opitz2
1KarlsruheInstitute of Technology, KIT, Karlsruhe, Germany, 76131
2Federal Network Agency, BNetzA, Mainz, Germany, 55122
3PKTEC GmbH, Schutterwald, Germany, 77746
Keywords: Standards, RF/Microwaves, Completed (published)
Presented by: Josef Opitz
The method presented in this poster for calculating the area of protection in the near field of a
mobile radio transmitting antenna is based on a field theoretical method. It works for any antenna,
without having to know the structure or details of the antenna. Only one data set is needed with the
near field data of the antenna on an envelope surface (e.g. sphere). It is thus suitable for both
current and future antenna designs.
557
Introduction
With its site certification procedure the Federal Network Network Agency ensures that the limits
applicable in Germany for the protection of persons exposed to electromagnetics fields of radio
installations are consistently applied without fail.
A fixed radio system with an equivalent isotropic radiated power of 10 Watts and more may only be
put into operation if the Federal Network Angency has evaluated the planned operation of the radio
system and has set safety distances to be adhered to in a site certificate. The determination of the
safety distances to be maintained is usually done by a conservative far-field calculation. However,
the calculation methods used so far are increasingly reaching their limits, so that especially in urban
areas the needs for complex field strength measurements are continuously increasing.
This results in the need for a precise method for calculating the electromagnetic field emitted by the
antenna and the resulting protection areas.The less accurate the procedure, the higher the
necessary safety margins. Consequently, to minimize the safety margins, the quality and accuracy
of the calculation must be high.Previous methods often use approximations based on very simple
calculation methods that do not take into account the actual characteristics of the antennas,
especially the near fields. These simple models often also require parameters of the antenna, which
are not easy to estimate. The required high accuracy and reliability is thus often not achievable.
Calculation of the protection area
The method presented here is based on a field-theoretical calculation method and thus achieves a
high accurarcy and uniqueness of the calculation.The area of protection around an antenna or an
array of antennas is the 3-dimensional area in which the limits for the electric or magnetic field are
exceeded. This results directly from the calculated electromagnetic field. The accuracy of the
protected area directly depends on the accuracy of the underlying calculation method for the
electromagnetic field calculation. The relevant protection zone is usually in the vicinity of the
antenna. It cannot be determined considerung only the far field properties of the antenna.
Description of the procedure
An essential requirement of the novel method is that, despite improved computational accuracy, it
must not presuppose detailed knowledge of the antenna. The exact structure of the antenna is
usually known only to the manufacturer and must not be published. Consequently, a method based
on a detailed simulation would be difficult to implement and very expensive.
The method presented here is based on the equivalence principle. Thus, the field radiated by a
source can be accurately calculated even without knowledge of the source (here, antenna), if the
tangential electromagnetic field on a closed envelope around the antenna is known. This shell is
expediently chosen quite small in the vicinity for the antenna. This can be, for example, a spherical
surface on which the electromagnetic field inclunding polarization and phase is determined by
measurement (near-field measurement). Corresponding measured values are generally available at
the antenna manufacturer. The mode of operation of the method is based on the fact that the field
can be reconstructed exactly from the near-field data outside of the envelope. Within this envelope,
the reconstructed field of the antenna always returns zero. As a result, the boundary conditions
inside the shell and thus the structure of the antenna itself no longer matter. It is only important that
the entire antenna structure is within this envelope surface. If the tangential field strength
components of E and H are known on the entire sphere surface, the field outside the sphere can be
determined by applying the Green´s function via integration over the sphere surface. According to
the uniqueness theorem, the field calculated that way agrees with the original field of the antenna at
every point outside the sphere.
It´s desirable to have the envelope as tight as possible around the antenna housing, to keep the non
computable area inside the envelope small. Hence, a sphere would not always be the best choice
esspecially for the form factor of typical base station antennas. To adress this problem the
transformation of the envelope sphere into a more conform equivalent slim cylinder is currently in
development.

558
Implementation of the calculation method
The calculation method described is implemented in software ("WattWächter Mobilfunk").This
consists of a core which can be integrated into Java Virtual Machine (JVM). For performance
reasons, the core is implemented in C ++. The control of the core and the communication with users
and clients, as well as the input-output are implemented by Java routines utilizing JNI (Java Native
Interface). As a result, the software is largely platform-independent and easy to maintain.
Calculation example
Proof on the function and practicality of the method is provided by the frequently used mobile radio
antenna 7422155V01 from Kathrein at the frequencies 1855 MHz and 2140 MHz. Based on
measured near-field data from Kathrein, the electric field is calculated on the one hand in sectional
planes and compared with measured data. Environment and ground influence are not taken into
account. In the chosen scenario, these influences only play a subordinate role.
For the comparative measurement data, an antenna of this type was set up and the electric field
strength measured at various points in the room using a Narda E-field probe.The calculated values
are compared with the measured data.
The following pictures (Fig.1) show the electric field strength (RMS) of the test antenna as cutting
planes for a polarization of + 450 and a set electrical down tilt of 00 . The near field used for the
calculation is based on near field measurements on a spherical surface with a radius of about 3
meter and was provided by the manufacturer. The inside of the sphere has no field strength and is
therefore shown black in the illustration.
The field pictures show the characteristic of a mobile radio antenna typical course, with a broad lobe
in horizontal section and a narrow, directed down tilt lobe in vertical section. Because of 00 down tilt
the lobe runs horizontally here. The vertical cuts also show the rather weak side lobes. The bright
points on the spherical surface in the vertical sections are numerical artifacts, due to the singularity
of the Green´s function used. Though, it is possible to avoid these artifacts showing up by applying
small modifications to the algorithm.
Verification measurement
For verification of the calculation results based on the data available from the manufacturer,
comparative measurements are performed on an exemplary test antenna. For this purpose, the
electric field was recorded at eleven defined points around the antenna at two test frequencies. The
measurements were carried out in three different heights. At one position, the field is additionally
measured above the height. To measure the field strength, an E-field measuring probe from Narda
of the measuring service of the Federal Network Agency is used.
Conclusion for comparison with measured values
The local course of the field strengths is reflected very well by the calculation. The difference
between calculated and measured values is mostly within the range 0.7 to 1,7 dB with few
exceptions. The measurement setup as well as the graphics for the comperison
measurement/calculation are available for download on the EMF internet pages of the Federal
Network Agency (http://www.emf3.bundesnetzagentur.de)
The method is thus very well suited for calculating the electromagnetic field of a mobile radio base
station antenna in the near field (an far field) and to deduce protection areas from it.
Calculation of protection areas
In order to be able to calculate protection zones, the calculated field strength in the vicinity of the
antenna is compared with the permissible limit value. At the two frequencies used for the
measurement, the ICNIRP limits for the electric field strength are:
1855 MHz - 59,2 V/m
2140 MHz -61 V/m

559
Due to the similarity of the different operating cases, only the case with supply at 2140 MHz with
polarization "+" is considered here. The fed-in power amounts to 100 Watt. Figure 2 shows the
protection area around the antenna in horizontal and vertical section. The calculated field strength
was increased by 1,6 dB safety markup.
The dimensions of the protected area can be read from the figure 2: In the main beam direction
there is a distance of 8.4 meter to the antenna. The vertikal extend of the protection area is 0.8
meter. The horizontal extension 5.50 meter across the atenna. The situation is also not completely
symmetrical, neither in the vertical nor in the horizontal direction. Without a safety margin, the main
beam direction would has a safety distance of 6.80 meter and a horizontal extension of 4.50 meter.
A free version of the calculation tool (WattWächter Mobilfunk) and a detail description of the
procedure can be downloaded from the Federal Network Agency´s EMF website
(http://www.emf3.bundesnetzagentur.de).
Figures

560
Figure 1. Calculated E-field "WattWächter Mobilfunk"

561
 Session: PB
Poster Session B
Tuesday June 26, 2018 • 14:30 - 16:00
Mediteranea

PB-2 [14:30]
Effect of Transcranial Magnetic Stimulation in a model of rat´s Hemiparkinson
Leticia Verdugo-Diaz1, Idalia Medina-Salazar1 & David Elias-Viñas2
1Departamento de Fisiologia, Universidad Nacional Autonoma de Mexico, Ciudad de Mexico, Mexico, 04510
2Seccion Bioelectrònica, Instituto Politecnico Nacional, Ciudad de Mexico, Mexico, 07360
Keywords: Behavioural, ELF/LF, Work in Progress
Presented by: Leticia Verdugo-Diaz
The effectiveness of Transcranial Magnetic Stimulation (TMS) experiments depends on TMS coil
position and frequency of stimulus. This work presents results with a device and a coil constructed
especially for the head size of the animals used. We tested repetitive TMS at two frequencies in the
rat model of Parkinson's disease. Significant differences were found in different motor behaviors if
the stimulation was either inhibitory (1Hz) or stimulatory (5Hz).
Introduction
TMS is a non-invasive therapy that can alter the excitability of the cerebral cortex and other parts of
the brain (Pascual-Leone et al., 2001). Repetitive TMS (rTMS) has been used as a treatment in
different neuropsychiatric and neurodegenerative pathologies; however, its mechanism of action has
not been fully elucidated. rTMS induces some recovery of motor and cognitive function, as well as
an increase in dopamine (Vandala el al, 2015) in patients with Parkinson's disease (PD). It is known
that the frequency of rTMS stimulation plays a critical role in the induced response; so, frequencies
greater than 5 Hz typically induce facilitation, whereas those less than or equal to 1 Hz induce
plasticity inhibition (Chervyakow et al., 2005). We wanted to learn about the effects TMS of different
frequencies has on the animal model of PD. Thus, the aim of the study was to evaluate motor
behaviors related to the animal model of PD, induced by low (1 Hz) or high frequency (5 Hz) rTMS.
Methods
A hemiparkinson model in adult male Wistar rats (200-220 g) was used. The rat model was
produced with unilateral 6-hydroxydopamine (6-OHDA) lesions in the medial forebrain bundle. The
lesion was evaluated 10 days later with the apomorphine-induced rotation test. The following
experimental groups were (n = 6 in each): 1) control (non-injury); 2) false injury (sham); 3) sham +
rTMS; 4) sham + false stimulation; 5) Hemiparkinson (6-OHDA); 6) 6-OHDA + rTMS; and, 7)
6-OHDA + false stimulation. The rTMS was performed with the EMAGPRO12 device and an eight-
shaped coil. The device allows to provide magnetic stimuli to rodents, in the same way that an rTMS
does in humans. The stimulation parameters selected were 1 or 5 Hz, 10 min/day for 30 days.
Stimulation was initiated 15 days after the injury. Before, and one day after the last stimulation, the
behavioral tests performed were: balance beam, curling, open field and elevated plus maze.
Results
Any significant difference was observed on behavioral tests induced by rTMS or false stimulation in
both control and sham groups. 6-OHDA-lesioned rats exhibited a decrease of the exploratory
activity (open field test) and an increase of the total time to cross the beam. 6-OHDA-lesioned
animals receiving rTMS of 1 Hz showed significant differences in the balance beam (increased
travel time) and the elevated plus maze tests (greater entrances and time spent in the open arms).
The other behavioral tests do not show significant differences in these groups.
Discussion
rTMS produces electrical currents in cortical regions beneath the stimulation coil, that can change
562
excitability of cortical and subcortical structures depending on the frequency of the stimulation
applied (Pascual-Leone et al., 2001). Despite the diverse clinical applications of TMS in the
treatment of neurodegenerative diseases including PD, the molecular basis of the neuron response
to the electromagnetic stimulus is far from being understood. These therapeutic strategies mainly
alleviate the symptomatology of PD patients, and a lack of knowledge in terms of cellular, molecular,
and epigenetic determinants of the brain’s reaction to TMS significantly limits the improvement of
these techniques. Few reports have explored the effect of TMS on PD animal model. Most of these
studies used TMS devices and coils designed for the human brain; therefore, the stimulation is not
focal (Kanno eta., 2004, Keck et al 2001; Yang et al 2010). Yang et al (2010) reported that low-
frequency rTMS treatment ameliorated 6-OHDA-induced brain injury of rats and prevented the loss
of dopaminergic neurons in rat substantia nigra. They assumed that these effects are related to
antioxidant and anti-inflammatory activities created by low-frequency rTMS. Studies in animals non-
lesioned, rTMS increased the release of dopamine in the mesolimbic and mesostriatal system
(Kerck et al 2001) and on the dorsolateral striatum (Kanno et al, 2004). Thus, the behavioral results
presented in this report may be due to changes in neuronal excitability and/or in the release of
dopamine.
Conclusion
The preliminary results of the present work showed differential behavioral effects between 1 and 5
Hz rTMS on the model of rat’s Hemiparkinson. Other frequencies must be tested in addition to the
correlation of the behavioral data with immunohistochemically and to the expression of specific
neuronal markers involved in the response to rTMS.
References
Chervyakov AV, Chernyavsky AY, Sinitsyn DO, Piradov MA. 2015. Possible mechanisms
underlying the therapeutic effects of transcranial magnetic stimulation. Front Hum Neurosci
9:303.
Kanno M, Matsumoto M, Togashi H, Yoshioka M, Mano Y. Effects of acute repetitive transcranial
magnetic stimulation on dopamine release in the rat dorsolateral striatum. J Neurol Sci 2004;
217: 73-81.
Keck ME, Welt T, Müller MB, et al. Repetitive transcranial magnetic stimulation increases the
release of dopamine in the mesolimbic and mesostriatal system. Neuropharmacology 2002; 43:
101-9.
Pascual-Leone A, Davey NJ, Rothwell J, Wassermann E, Puri BK. 2001. Handbook of
Transcranial Magnetic Stimulation. London, United Kingdom: Arnold.
Vadala M, Vallelunga A, Palmieri L, Palmieri B, Morales-Medina JC, Iannitti T. 2015.
Mechanisms and therapeutic applications of electromagnetic therapy in Parkinson’s disease.
Behav Brain Funct 11:26.
Yang X, Song L, Liu Z. The effect of repetitive transcranial magnetic stimulation on a model rat
of Parkinson’s disease. NeuroReport 2010; 21: 268-72.
Acknowledgement.This work was supported by PAPIIT grant IN211417 to LV-D.

PB-4 [14:30]
Early exposure of rotating magnetic fields promotes central nervous regeneration in
planarian Girardia sinensis
Xiaomei Wang1 & Xiaoyun Zhang1
1Physiology Department, Shenzhen , China, 518060
Keywords: Behavioural, Static, Completed (published)
Presented by: Xiaomei Wang
Key words: rotary non-uniform magnetic fields; in vivi; planarian; neural regeneration; stimulating

563
effect; window effect; completed(pulished). Presented by: Xiaomei Wang
Magnetic field exposure is an accepted safe and effective modality for nerve injury. However, it is
clinically used only as a supplement or salvage therapy at a later stage of treatment. Here, we used
a planarian Girardia sinensis decapitated model to investigate beneficial effects of early rotary
nonuniform magnetic fields (RMFs) exposure on central nervous regeneration. Our results clearly
indicated that magnetic stimulation induced from early RMFs exposure significantly promoted neural
regeneration of planarians. This stimulating effect is frequency and intensity dependent. Optimum
effects were obtained when decapitated planarians were cultured at 20 8C, starved for 3 days
before head-cutting, and treated with 6 Hz 0.02 T RMFs. At early regeneration stage, RMFs
exposure eliminated edema around the wound and facilitated subsequent formation of blastema. It
also accelerated cell proliferation and recovery of neuron functionality. Early RMFs exposure up-
regulated expression of neural regeneration related proteins, EGR4 and Netrin 2, and mature nerve
cell marker proteins, NSE and NPY. These results suggest that RMFs therapy produced early and
significant benefit in central nervous regeneration, and should be clinically used at the early stage of
neural regeneration, with appropriate optimal frequency and intensity.

PB-6 [14:30]
Electromagnetic interference with medical devices due to radio waves of TD-LTE using 3.5
GHz band and candidate frequencies for 5G
Takahiro Iyama1, Teruo Onishi1 & Keisuke Nagase2
1NTT DOCOMO, INC., Yokosuka, Kanagawa, Japan, 239-8536
2Kanazawa University Hospital, Kanazawa, Ishikawa, Japan, 920-8641
Keywords: Clinical (diagnostics), RF/Microwaves, Completed (unpublished)
Presented by: Teruo Onishi
Examination of EMI with medical devices is conducted due to radio waves of TD-LTE using 3.5 GHz
band and candidate frequencies for 5G. As frequency increase, EMI tends to be less occurred and
category of EMI tends to decrease. At 28.5 GHz, all medical devices do not cause EMI in this
examination.
Introduction
The electromagnetic interference (EMI) with medical devices due to mobile phones using third
generation mobile communication systems (3G)/frequency division duplex (FD)-long term evolution
(LTE) and wireless local area network (WLAN) technologies has been investigated [1]-[3]. This
paper describes the results of such EMI from radio waves of time division duplex (TD)-LTE using 3.5
GHz band and candidate frequencies for fifth generation mobile communication systems (5G).
Materials and Methods
Examination parameters are listed in Table 1. For TD-LTE using 3.5 GHz band, radio waves with
modulated uplink signals defined in the Third Generation Partnership Project (3GPP) are emitted
from a dipole antenna. By using a signal generator and a power amplifier, the emitted power can be
controlled at arbitrary levels. Mobile phones which emit radio waves at the maximum power are also
used as a radio source. For 5G candidate frequencies, radio waves with non-modulated signals at
3.9 GHz and 4.65 GHz are emitted from tuned dipole antennas. A horn antenna is also used at 28.5
GHz. Totally 23 medical devices including external cardiac pacemakers, electrocardiographs, and
electroencephalographs are used for this examination.
As the basic examination, dipole or horn antennas scan on each surface of medical devices and
their attachments including sensors and cables. The conditions in the examination are as follows.
1. Transmission with the period of 1 second (On: 0.5 s, Off: 0.5 s),
2. Maximum transmission power, and
3. Scanning antenna while touching medical devices.

564
In the case that the EMI occurred on the conditions above, the optional examination is conducted
while changing the conditions described below separately.
1. Continuous transmission,
2. Adjusting transmission power till EMI disappeared, and
3. Adjusting distance between antenna and medical devices till EMC disappeared.
The medical engineers classify the degree of EMI. The criteria for such classification, that is called
“category”, are described in [4]. It is noted that category “1” means “non- EMI”. Categories “2” and “3
or more” mean that “reversible and disturbed in medical treatment” and “irreversible, misdirected,
and/or declining etc.”, respectively.
Results and Discussion
Table 2 summarizes the EMI with medical devices for each frequency. In 16 medical devices in
which EMI occurred for WLAN signals (2.4 GHz, 5.2 GHz, and 5.6 GHz band) in the previous
examination, the ratio of medical devices which still cause EMI at 3.5 GHz (TD-LTE), 3.9 GHz, and
4.65 GHz is less than 40%.
In all 23 medical devices, such ratio is almost 30%. At 28.5 GHz, all medical devices do not cause
EMI in this study. Figure 1 shows the category of EMI for 19 medical devices which are commonly
used in the previous examination for WLAN and newly used frequencies in this study. The ratio of
category “2 or less” increases with increasing frequency. The maximum distance between medical
devices causing EMI and antenna is approximately 40 cm.
Conclusions
Examination of EMI with medical devices is conducted due to radio waves of TD-LTE using 3.5 GHz
band and candidate frequencies for 5G. As frequency increase, EMI tends to be less occurred and
category of EMI tends to decrease. At 28.5 GHz, all medical devices do not cause EMI in this
examination.
Acknowledgements
The authors thank the medical engineers in Kanazawa University Hospital for their technical
assistance with this study.
References
1. K. Nagase et al., IEICE ComEX, vol. 1, pp. 222-227, 2012.
2. S. Ishihara et al., IEICE EMC’14, 14P1-H6, 2014.
3. S. Shimojima et al., IEICE General Conference 2018, B-20-2, 2017 (in Japanese).
4. MIC, "Report on research and study of effects of radio waves on medical equipment, etc., " 2002
(in Japanese).

T 1. E .

565
 T 2. N EMI.

F 1. C EMI ( WLAN).

PB-8 [14:30]
Device development for application of pulsed electromagnetic fields for hand arthritis
therapy
Andrea Fernandez1, Marco Rivera-Gonzalez1, Nazario Felix-Gonzalez1 & Ceferino Maestu1
1Laboratorio de Bio-electromagnetismo, Universidad Politécnica de Madrid, Madrid, Spain, 28223
Keywords: Clinical (therapy), Pulsed, Work in Progress
Presented by: Ceferino Maestu
Arthritis defines hundreds of musculoskeletal disorders, such as Rheumatoid Arthritis (RA) and
Osteoarthritis (OA). Nowadays, Pulsed electromagnetic Fields (PEMFs) are being applied within the
field of Rheumatology due to its promising results in tissue regeneration, joint damage and pain
relief. This work reflects the obtained results that identified the optimal parameters of PEMF, which
enhance cell proliferation; and the development and results of a new and effective musculoskeletal
treatment based on a portable therapeutic device.
Introduction
In recent years, the field of Rheumatology has been progressively favoured by medical advances
due to the increasing cases of musculoskeletal diseases in patients of adult age. Arthritis is a
concept that encompasses the whole set of inflammatory pathologies of the joints, being the most
affected the knees, hip and hands, being Osteoarthritis (OA) and Rheumatoid Arthritis (RA) the most
prevalent ones.

566
OA is caused due to joint wear because of excess load. On the other hand, in RA the immune
system triggers an inflammatory cascade that culminates in the destruction of the joint and, as it is a
systemic disease, affects multiple joints. Both pathologies are characterized by the irreversible
destruction of cartilage, tendon and bone of the damaged joint. This degradation of the connective
tissue can lead to a substantial loss of functionality, inflammation and pain1. As a consequence of
this, many patients suffer a serious loss of their independence and quality of life.
Studies carried out during the last decades have evaluated the promising benefits of the application
of PEMFs in arthritis therapy, among which chondroprotection, pain relief, anti-inflammation and
increased mobility stand out. Both in vivo studies carried out with animals2,3 and in vitro studies with
human cells4-7 analysed PEMF parameters in order to maximize the results. These parameters are:
frequency, exposure time, intensity and waveform.
Methods
Device Development
The aim of this study is the development of a device based on a combined dual therapy both
magnetic and thermal at the same time in independent circuits. Also, first results applying this novel
device are presented.
The parameters of the electromagnetic fields applied have been established as optimal in the
previous cell culture experiment. These parameters are:
-Frequency: 60Hz.
-Field intensity: 0.84mT.
-Waveform: square.
-Induced temperature: 38.5ºC
The patented prototype (ES2575886(A1) ― 2016-07-01) which has been developed, figure1,
consists of three gloves, one on top of the other, figure 3. Within the first level, the heating coils are
placed between the joints, figure 3-left. The second level is composed by the magnetic coils, figure
3-centre, which rest on a plastic support. Also, this level contains the generator circuit, the batteries
and the battery charging circuit, figure 2. Finally, both levels are covered with a third glove for
protection, figure 3-right.
In each finger, the magnetic coils need to be fixed on the interphalangeal joints, figure 3-centre, in
order to reinforce the homogeneity of the field, therefore, the field lines will be distributed
longitudinally along the finger and cell proliferation can occur. Also, the thermal circuit temperature
(38.5ºC) is the one commonly used in arthritis treatments that oscillates in between 36ºC and 41ºC,
figure 3-left. The device has a security system to prevent overheating and would turn itself off
instantly in case of exceeding this temperature. The Spanish Patent and Trademark Office granted
the device patent on 2017: “Portable therapeutic device for the treatment of arthritis by magneto-
thermal stimulation” Patent Number: ES2575886(A1) ― 2016-07-01.
Osteoarthritis and Rheumatoid Arthritis patients were selected to participate in the trial. The
participants were divided into two homogenous groups, so one of them could perform as a control.
The therapy that was proposed took place twice a day and lasted two hours per session.
Throughout the study, a researcher tracked patients’ progress through a HAQ test, an EVA test and
the force measure of each finger by means of a dynamometer.
Results
After completing the trials, most patients reported significant pain relief and increased finger mobility.
However, a few participants did not experience any improvement of their medical condition. Any of
the patients suffered from adverse effects form the treatment.
Discussion
Even though the participants of the trial observed a great improvement, further medical tests are

567
required to prove the findings.
Conclusions
All the results given can be used to affirm that PEMF therapy could be an effective alternative
treatment for arthritic patients. However, additional studies should be designed to obtain more
quantitative and qualitative data, in order to accurately determine the effectiveness of low frequency
pulsed electromagnetic fields therapy to totally clarify its mechanism.
The development of a portable therapeutic device could seriously benefit patients with
musculoskeletal disorders allowing them to being treated with a non-invasive method that would
efficiently relieve the symptoms. However, its effectiveness remains unsettled, until further tests are
performed.
References
[1] Jahanban-Esfahlan, R., Mehrzadi, S., Reiter, R. J., Seidi, K., Majidinia, M., Baghi, H. B., Khatami,
N., Yousefi, B., and Sadeghpour, A. (2017) Melatonin in regulation of inflammatory pathways in
rheumatoid arthritis and osteoarthritis: involvement of circadian clock genes. British Journal of
Pharmacology, doi: 10.1111/bph.13898
[2] Pulsed electromagnetic fields reduce knee ostearthritic lesion progression in the aged Dunkin
Hartley guinea pigs. M. Fini, G. Giavaresi, P. Torricelli, F. Cavani, S. Setti V. Cane et al., 2005
[3] Modification of osteoarthritis by pulsed electromagnetic field—a morphological study.D. McK.
Ciombor Ph.D., R. K. Aaron M.D., S. Wang M.D. and B. Simon Ph.D.2003.
[4] The effect of pulsed electromagnetic fields on chondrocyte morphology. M. E. Jahns, E.Lou, N.
G. Durdle, K. Bagnall, V. J. Raso, D. Cinats, R. D. C. Barley, J. Cinats, N. M. Jomha. 2007.
[5] Effect of Pulsed Electromagnetic Fields on the Bioactivity of Human Osteoarthritic Chondrocytes.
Patrick Sadoghi, MD; Andreas Leithner, MD; Ronald Dorotka, MD; Patrick Vavken, MD, MSc.
[6] In vitro exposure of human chondrocytes to pulsed electromagnetic fields. V. Nicolin, C. Ponti, G.
Baldini, D. Gibellini, R. Bortul, M. Zweyer, B. Martinelli, P. Narducci.
[7] Effects of Pulsed Electromagnetic Fields on Human Chondrocytes: An In Vitro Study. F. Pezzetti,
M. De Mattei, A. Caruso, R. Cadossi, P. Zucchini, F. Carinci, G. C. Traina, V. Sollazzo.
Figures

Figure 1. Drawing of the device from patent ES2575886.

568
Figure 2. Electronic Device Circuit.

569
 Figure 3. Different levels of the glove.

PB-10 [14:30]
STUDENT PAPER
Development and evaluation of microwave antenna for transcatheter denervation
Shohei Matsuhara1, Kazuyuki Saito1, Hiroshi Kawahira1, Nobuyoshi Takeshita2 & Tomoyuki Tajima3
1Chiba University, Chiba, Japan
2National Cancer Center, Kashiwa, Japan
3The University of Tokyo Hospital, Tokyo, Japan
Keywords: Clinical (therapy), RF/Microwaves, Work in Progress
Presented by: Shohei Matsuhara
In this study, a catheter, which ablates extravascular nerve by the thermal effect of microwave, is
introduced. In addition, heating characteristics of the catheter are evaluated by the numerical
calculation and the experimental investigation. As a result, temperature distributions around the
catheter were revealed and the usefulness of a denervation catheter using microwave was
suggested.
Introduction
Transcatheter denervation is a treatment in which a catheter inserted to a treatment site ablates the
nerve outside the blood vessel (Fig1). It is mainly aimed at treating hypertension, but it has not
necessarily been effective [1]. One reason is that it uses radio frequency (RF) current. RF current
has inherently limited depth of heating and ability to provide circumferential nerve ablation, which
could lead to incomplete denervation and inconsistent clinical efficacy [2]. Microwave energy may
have suitable characteristics for transcatheter denervation [3]. In this study, we consider this
technology for clinical application. In this study,we evaluate the heating ability of a microwave
denervation catheter.
Methods
Heating characteristics of the catheter is evaluated by numerical calculation and validity of the
calculation is confirmed by experimental investigations. In this study, coaxial-slot antenna is

570
introduced for the purpose. The operating frequency of the antenna is 2.45 GHz which is one of the
industrial, scientific and medical (ISM) frequencies in Japan.
In the numerical calculation, first, specific absorption rate (SAR) distribution around the antenna is
calculated. The SAR is one of the most important indexes for evaluation of heating characteristics
and is heating source inside the biological tissue. After that, temperature distribution in biological
tissue is calculated. It can be obtained by solving bioheat transfer equation numerically.
As a calculation model, an antenna with a diameter of 1.5 mm and blood with a diameter of 8 mm
are concentrically arranged in a uniform blood vessel. The cooling effect by the blood flow is
substituted by keeping the temperature of the blood area 37 °C.
In the experiment, biological tissue equivalent solid phantom with a hole with a diameter of 8 mm is
connected to small electrical pump and refluxed with 37 °C of water. (Figure 2) The antenna is
placed in the center of the hole, and the temperature distribution of the cross section of the phantom
is measured by an infrared camera.
Results
Figure 3(a) shows the calculated temperature distribution. Here input power and heating time are
35W and 300 s, respectively. Figure 3(b) shows the measured temperature distribution. Here input
power and heating time are approximately 33W and 300 s, respectively.
As the result of both calculation and experiment, a circumferential heating pattern was observed.
Temperatures exceeding 60 °C are also observed in the vicinity of 7 mm from the intima. Moreover,
the intima is not heated, and it is understood that the vicinity of the intima is substantially cooled
down by the blood flow. Therefore, theburn injury of intima can be avoided.
Discussion
Similar temperature distribution is observed in calculation and experiment, and the validity of the
calculation can be confirmed.
The nerve is inactivated above 60 °C, and more than 90 % of the nerve exist within 7 mm from the
intima. Therefore, it can be said that the obtained temperature distribution is sufficient for ablating
nerves.
Conclusion
The denervation catheter using the microwave energy is introduced. Moreover, heating pattern
around the antenna is evaluated by the numerical calculation and validity of the calculation can be
confirmed by the experimental investigations. As a further study, practical device based on this
technique should be developed.
References
[1] Symplicity HTN-3 Investigators, G.L. Bakris, R.R. Townsend, J.M. Flack, S. Brar, S.A. Cohen, R.
D’Agostino, D.E. Kandzari, B.T. Katzen, M.B. Leon, L. Mauri, M. Negoita, W.W. O’Neill, S. Oparil, K.
Rocha-Singh, D.L. Bhatt, “12-month blood pressure results of catheter-based renal artery
denervation for resistant hypertension: the SYMPLICITY HTN-3 trial,” J AM Coll Cardiol, vol. 65, no.
13, pp. 1314-1321, April 2015.
[2] D.E. Haines, “Determinants of lesion size during radiofrequency catheter ablation: the role of
electrode-tissue contact pressure and duration of energy delivery,” J Cardiovasc Electrophysiol, vol.
2, no. 6, pp. 509-515, Dec 1991.
[3] P. Qian, M.A. Barry, T. Nguyen, D. Ross, P. Kovoor, A. McEwan, S. Thomas, A. Thiagalingam, “A
Novel Microwave Catheter Can Perform Noncontact Circumferential Endocardial Ablation in a Model
of Pulmonary Vein Isolation,” J Cardiovasc Electrophysiol, vol. 26, no. 7, pp. 799-804, Jul 2015.
Figures

571
Figure 1. Transcatheter denervation

Figure 2. Experimental model and coordinate system

Figure 3. (a) Calculated temperature distribution (b) Measured temperature distribution

572
PB-12 [14:30]
The multi-phase procedure for EMF limit values excesses verification
Cestnik Breda1 & Karol Grabner1
1Elektroinštitut Milan Vidmar, Ljubljana, Slovenia
Keywords: Dosimetry (computational), ELF/LF, Completed (published)
Presented by: Breda Cestnik
A multi-phase procedure to determine the area of exceeding EMF limit values in the living
environment of overhead electric power lines is explained. We present a detailed micro location
analysis with combining the EMF calculation results and geo-referenced data graphic components in
AutoCAD Map 3D. It enables a good visualisation of the field at the micro location.
1. Introduction
Slovenia is one of the rare countries in the world in which precautionary legislation regarding
electromagnetic fields (EMF) in the natural and living environment was adopted (in 1996) [1, 2].
Stricter limits for the electric and magnetic fields in the environment apply to new electric power
facilities in increased (1st) category area of protection against EMF. These limits are below the limits
set by ICNIRP guidelines. Slovenian legislation in a specific manner stipulates the maximum
permitted levels of EMF, the degree of protection against EMF in individual areas of natural and
living environment, the methodology of assessing EMF values and establishing excesses as well as
the measures for mitigating and preventing excessive values. The prescribed limit values for EMF
are presented in Table 1.
Table 1: Limit values for the electric and magnetic fields of the 50 Hz frequency [1]

2nd EMF protection area

1st EMF protection area
for new and reconstructed
EMF sources
for new and reconstructed
EMF sources
and 1st and 2nd EMF
protection area for existing
EMF sources

Electric field strength (E) 500 V/m 10000 V/m

Magnetic flux density (B) 10 µT 100 µT

Assessments or expert evaluations of EMF are always performed for new electric power facilities
prior to their siting in the environment.
In the abstract we present certain phases of EMF limit values excesses verification and E and B
calculation in which we involve detailed spatial data using the following commercially available
softwares: EFC-400 (Narda STS) and AutoCAD Map 3D (Autodesk).
Detailed spatial data are: geometry of overhead electric power lines, ground profile, buildings,
spatial information (buildings cadastre, land register of plots, intended space use etc.), digital
orthophoto.
With siting the new high-voltage transmission line in the natural and living environment the multi-
phase procedure for EMF limit values excesses verification consists of five phases: determination of
approximate corridor, identification of areas for micro location analysis, construction of 3D model for
calculation of E and B with EFC-400, calculation of E and B with EFC-400, evaluation and
visualization (with the AutoCAD Map 3D).
2. Determination of approximate corridor

573
In the first phase the determination of the approximate corridor for the entire overhead power line is
established. The approximate corridor is determined with general calculations of E and B with
EFC-400 concerning the least favourable situation in normal operations of the overhead power line
as the maximum distance to the 1st EMF protection areas and to buildings on 1st EMF protection
areas. The approximate corridor is based on the maximal distance from overhead power line axis to
the limit value curve.
3. Identification of areas for micro location analysis
The next phase is the identification of areas for micro location analysis. It is based on merging of
two layers - land register of plots with data of intended space use and axis of overhead power line
with its approximate corridor in SHP format - in AutoCAD Map 3D. This allows us to verify where the
approximate corridor overlaps with the 1st EMF protection areas.
4. Micro location analysis – construction of 3D model for calculation of E and B with EFC-400
We construct 3D calculation models of spans in the EFC-400 for recorded micro locations. Detailed
3D model includes data of ground profile and geometry of conductors (GK coordinates, angles,
heights, types of towers, the height of the lowest conductor, span length, the angle of the tower,
catenaries etc.)
5. Micro location analysis – calculation of E and B with EFC-400
With the EFC-400 we perform the calculations of the EMF 1 m above the ground or in section
planes at buildings. The calculations results of fields are then exported in the AutoCAD Map 3D and
are then transformed back to the natural coordinate system. Calculation results are prepared for the
merger with detailed spatial data.
For electric power line we usually calculate the unperturbed field, which can, in case of electric field,
be quite higher than the real electric field. Sometimes conductive objects are included in the 3D
model.
6. Micro location analysis – evaluation and visualization (with the AutoCAD Map 3D)
The evaluation and visualisation of calculated field values to appropriate EMF limit values is
possible with the merger of geo-referenced calculations and appropriate geo-referenced data-based
graphical platforms (buildings cadastre, land register of plots, intended space use, LIDAR (3D model
of ground profile, buildings, conductors, vegetation etc.), digital relief model of Slovenia, digital
orthophoto (DOF in TIFF format) as an aerial photograph for 2D view of the location, protective
zone…).
Figure 1 presents a 3D visualisation of high-voltage overhead electric power line model with
residential buildings on the terrain profile, generated from the LIDAR point cloud. Figure 2 presents
ground plan visualisation of the magnetic field calculation results (B) 1 m above ground with
included layers: DOF, intended space use, protective zone, buildigs with protective zone, 3D model
of high-voltage overhead electric power line, B calculation results. Figure 3 presents a 3D
visualisation of the model with calculation results of electric field strength (E) and included layers:
DOF, LIDAR Slo buildings and high vegetation, 3D model of high-voltage overhead electric power
line, E calculation results.
7. Conclusion
The purpose of involving more detailed spatial data is achieving greater calculation precision, faster
automation of processes and lower possibility of errors generation during data entry and data
processing, as well as a more plastic visualisation of fields on micro locations. The construction of a
detailed 3D calculation model for the entire overhead power line would be a very time-consuming
task in most cases. Therefore, only where the 1st EMF protection areas and buildings are situated
within the approximate corridor, it is reasonable to perform a micro location analysis of fields with a
3D situation model on the micro location. We have presented the method of combining the 3D
model with the EMF calculation results and geo-referenced data graphic components in AutoCad

574
Map 3D, which provides the basis for evaluating the calculated field values to appropriate limit
values. It also enables a good visualisation of the field at the micro location.
8. References
[1] Decree on electromagnetic radiation in the natural and living environment, Official Gazette of the
Republic of Slovenia, no. 70/96
[2] Regulation on initial measurements and operational monitoring for sources of electromagnetic
radiation and on condition for their execution, Official Gazette of the Republic of Slovenia, no. 70/96
[3] Cestnik, B. Electromagnetic influences of electric power transmission system on the
environment. Vidmar's day, 2002, Ljubljana: EIMV
[4] Grabner, K., B. Cestnik, N. Ljubijankić. New approach in assessing EMF in the vicinity of high-
voltage overhead power lines, 10th CIGRE Conference, 2011, Ljubljana
Figures

Figure 1. 3D model of a high-voltage overhead electric power line and residential buildings

575
Figure 2. Ground plan calculation of magnetic field 1 m above the ground (included layers:
DOF, intended space use, protective zone, buildigs with protective zone, 3D model of high-
voltage overhead electric power line, B calculation results)

Figure 3. 3D presentation of electric field calculation results at micro-location (included layers:

DOF, LIDAR Slo buildings and high vegetation, 3D model of high-voltage overhead electric
power line, E calculation results)

576
PB-14 [14:30]
Numerical estimation of SAR enhancement due to implantable medical device inside human
body exposed to external electromagnetic waves - 5G cellular candidate frequency bands -
Takashi Hikage1, Keita Sakakibara1, Toshio Nojima1, Takahiro Iyama2 & Teruo Onishi2
1Hokkaido University, Sapporo, Japan, 0600814
2NTT DOCOMO inc., Yokosuka, Japan
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Takashi Hikage
The RF exposure guidelines do recognize that an increased local SAR (Specific Absorption Rate)
due to metallic implants is possible. In this study, the effect of implantable medical device in human
torso region is examined. A simple layer flat torso phantom is used which the implantable cardiac
pacemaker is implanted in. The local SAR dependencies due to exposure frequency bands were
presented. Here, we estimate localized SAR at several candidate frequency bands for 5th
generation wireless systems (5G).
Introduction
With regard to the electromagnetic field (EMF) emitted from wireless communication devices, radio
radiation protection guidelines for human exposure to EMF have been established [1-3]. These
guidelines provide no quantitative discussions about their relevance to humans with metallic objects
embedded in their bodies. However, given the progress in biomedical technologies, the number of
such users continues to increase, such as active implantable pacemakers / ICDs and medical
metallic plates, upper limb prostheses, and prosthetic legs. It is important to estimate the amount of
exposure that users with metallic implants will experience.
Some papers regarding interaction of radio frequency (RF) EMF and metallic implants have been
published, and studies on the EMF exposure of users with passive metallic implants have reported
measurements and numerical estimations based on numerical simulations [4-7]. A basic metric used
as the reference value is the specific absorption rate (SAR), which is defined as averaged energy
absorbed by the human body in any 6 minute period divided by the weight of tissues.
The main objective of this paper is to assess whether implantable medical device such as
implantable-cardiac pacemaker and cardioverter defibrillator could be expected to cause SAR
enhancements at several candidate frequency bands for 5th generation wireless systems (5G). A
torso phantom model [8] is used in which a cardiac pacemaker and single lead are implanted.
Methods
The human torso phantom for estimation, which contains the implantable-cardiac pacemaker and
lead, is constructed, as shown in Figure 1. As shown in the figure 2, the single unipolar lead is
modeled as is connected to the pacemaker’s terminal. The depth of implantation was 10 mm. Here,
we numerically estimated local SAR around implanted can of pacemaker and lead exposed to RF
plane wave at 5G cellular candidate frequency bands, 3.9 GHz, 4.65 GHz and 28.5 GHz. The
dielectric constants and electric conductivities of the biological material used in the phantom model
are assumed 2/3 muscle equivalent tissue [9]. The peak 1-g averaged SAR dependencies are
quantitatively investigated with metallic implants of various sizes and shapes. Finite-difference time-
domain (FDTD) analysis is applied [10].
Results and Conclusions
Local SARs for the torso phantom model with pacemaker and lead were evaluated. By using a torso
phantom implanted with pacemaker and lead model, the effect of metallic implants, were simulated
under far-field exposure condition at 5G cellular candidate frequency bands. In one example,
obtained frequency characteristics of local SAR distributions due to horizontally polarized RF wave
exposure in 5G cellular candidate frequency bands are shown in Figure 3. From the figure, we can

577
observe, at the higher frequency band exposure local SAR enhancement due to metallic implant
became lower. RF wave at these higher frequency band could not propagate into the biological
tissue deeply since penetration depth decreased. Then, the results indicated that local SAR
enhancements due to the implanted pacemaker and lead were enough small, and the most
increased rate was approximately 4% at 3.9 GHz exposure.
References
[1] ICNIRP Guidelines, “Guidelines for limiting exposure to time-varying electric, magnetic, and
electromagnetic fields (up to 300 GHz),” Health Phys., vol.74, no.4, pp.494–522, 1998.
[2] ANSI/IEEE C95.1-1999, “IEEE standard for safety levels with respect to human exposure to
radio frequency electromagnetic fields, 3 kHz to 300 GHz”
[3] “Measurement of SAR from mobile phone terminals and other terminals that are intended for use
in close proximity to the side of the head,” Telecommunications Technology Council Report,
Deliberation no/118, 2000.
[4] McIntosh RL, Anderson V, McKenzie RJ., “A numerical evaluation of SAR distribution and
temperature changes around a metallic plate in the head of a RF exposed worker,”
Bioelectromagnetics, vol.26, no.5, pp.377-388, 2005.
[5] Virtanen H, Keshvari J and Lappalainen R, “The effect of authentic metallic implants on the SAR
distribution of the head exposed to 900, 1800 and 2450 MHz dipole near field”, Phys. Med. Biol. 52
pp.1221-36, 2007.
[6] Matikka(formerly Virtanen) H, Keshvari J and Lappalainen R, ”Temperature changes associated
with radiofrequency exposure near authentic metallic implants in the head phantom—a near field
simulation study with 900, 1800 and 2450 MHz dipole”, Phys. Med. Biol. 55, pp. 5867–5881, 2010.
[7] A. Kyriakou, A. Christ, E. Neufeld, and N. Kuster, “Local Tissue Temperature Increase of a
Generic Implant Compared to the Basic Restrictions Defined in Safety Guidelines,”
Bioelectromagnetics, vol.33, no.5, pp.366–374, 2012.
[8] Irnich W.,et al., PACE, vol.19, pp.1431-1446, 1996.
[9] C.Gabriel, “Compilation of the dielectric properties of body tissues at RF and microwave
frequencies,” Brooks Air Force Technical Report AL/OE-TR-1996-0037, 1996.
[10] Schmid & Partner Engineering AG, SEMCAD X (http://www.speag.com/products/semcad
/overview/)
Figures

Figure 1. Torso phantom (ANSI/AAMI PC69)

578
 Figure 2. Numerical model of torso phantom for SAR estimation

Figure 3. Local SAR distributions around pacemaker lead (Simulation)

PB-16 [14:30]
Simulating the power deposition in a simple brain model loaded with magnetite
macrospheres and exposed in the near field of a 440 MHz monopole antenna
Cora Iftode1 & Simona Miclaus2
1Department of Measurements and Optical Electronics, Politehnica University of Timisoara, Timisoara,
Romania
2Department of Technical Sciences, "Nicolae Balcescu" Land Forces Academy, Sibiu, Romania
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Cora Iftode
In order to observe power deposition differences between a pure dielectric tissue and a tissue
containing magnetic material when exposed in the near field of a monopole antenna, we designed
an approach in ANSYS HFSS software. SAR was computed based not only on electric power
deposition, but also taking into account the magnetic absorption. Organs like human brain contain
minute quantities of magnetite nanocrystals whose role and presence in connection to RF field
exposure is still completely neglected. Although available computational dosimetric tools offer just
macroscopic description of magnetic power dissipation, it is worth analyzing simplified situations like
579
an ellipsoidal dielectric brain containing small magnetite spheres.
1. Introduction
Practically all radiofrequency (RF) computational dosimetry in the ultrahigh frequency (UHF) range
usually used today, completely neglects the presence of trace amount of ferrimagnetic material in
some tissues. In consequence, specific absorption rate (SAR) of energy absorbed in the exposed
tissue is computed based solely on the internal electric (E) field strength determination, SAR_E, and
on the dielectric behavior of tissues. With new data from 2016, we know that human brain contains
not only nanometric biogenic magnetite – whose role is still unknown, but also exogenic magnetite
particles arising from environment pollution and deposited there. Magnetite is ferrimagnetic and
absorbs well the magnetic component (H) of a UHF wave. During use of various communication
devices in front of, or lateral to the head, an open question persists: which is the contribution to the
total absorbed power, SAR_TOT, of the H-field component separately, SAR_H, and which is the
impact of the fact that deep local magnetic hot-spots may appear on the surface of magnetite
particles spread in the brain due to magnetic energy dissipation? In order to approach such a topic,
we developed gradual dosimetric situations by using the macroscopic-scale computational
capabilities of ANSYS HFSS software.
2. Materials and methods
Since all software programs in use today are automatically calculating just SAR_E, we have used
here the HFSS capability to compute the mean volume loss density (VLD) based on the expression
of the real part of the time-dependent Poynting vector:

and from it we could then determine SAR_TOT, which is the comprehensive value of SAR when
both the electric and magnetic losses are considered. The radiation source was a thin monopole
antenna (2 mm thickness) with the length L=18 cm connected to a ground plane with dimensions 50
cm x 90 cm and fed with 1W input power when frequency was set at f=440 MHz. It mimics well the
portable transceivers antennas which are held during communication at few centimeters in front of
the face.
The dosimetric approach developed in three stages:
1) A single sphere (radius rsf=1.6 cm) was placed at a distance D=13 cm in front of the antenna and
its material was set in three different ways:
a1) as a pure dielectric (owing just dielectric parameters);
b1) as a pure magnetic (owing just the magnetic properties of magnetite, but no dielectric
parameters);
c1) as real magnetite (owing mixed features – dielectric and magnetic). The material properties are
presented in Table 1.
2) A homogeneous brain was modeled as an ellipsoid with volume= 1130 cm3 and assigned the
material parameters given in Table 2. The mass density of the brain tissue = 1100 kg/m3 while for
the magnetite= 5200 kg/m3. One magnetite sphere was inserted in the brain, in a non-symmetric
position regarding the ellipsoid axes. The antenna was situated at a distance D=13 cm from the
brain center, while the closest point of brain surface to the antenna was at 8 cm. Computation of
mean VLD in the sphere and in the brain tissue allowed determination of the mean SAR_TOT over
“non-magnetic brain” and over “magnetic-brain” (Table 2).
3) Five magnetite spheres of a radius r=1 cm each, were inserted in the ellipsoidal brain model (Fig.
1a). The volume fraction of magnetite was 3.7‰. The spheres were placed in various positions,
three of them having centers in the same vertical plane (parallel) to the antenna axis. Average
SAR_TOT was computed again comparatively, by not taking into account their presence (“non-
magnetic brain”) and by taking them into account (“magnetic brain) (Table 2).

580
3. Results and discussions
Table 1 – Power loss of a sphere in air placed in the near field of the monopole antenna emitting at
440 MHz with a radiated power of 0.93 W:

Computed Pure Pure

Sphere type Mixed
situation dielectric magnetic

Relative
εr’=50 εr’=1 εr’=50
permittivity

Conductivity σ=3 S/m σ=0 S/m σ=3 S/m

Dielectric loss
Input material Electric/magnetic δe=2.5 δe=0 δe=2.5
tangent
data properties

Relative
µr=1 µr=1.9 µr=1.9
permeability

Magnetic loss
δm=0 δm=0.3 δm=0.3
tangent

Mean Volume_loss_density
50.63 31.43 99.77
[W/m3]

Case 1 = One
Mean Local SAR_E [mW/kg] 9.74 0 14.20
sphere in air

Mean Local SAR_TOT

9.74 6.04 19.18
[mW/kg]

Considering both dielectric and magnetic losses in the sphere in air, increases the average SAR by
35% if compared with the situation when only dielectric losses are considered.
Table 2 – Power loss of a “non-magnetic brain” and of a “magnetic brain” model placed in the near
field of the monopole antenna emitting at 440 MHz with a radiated power of 0.93 W:

Computed Brain components Brain tissue With Spheres

situation (dielectric) (magnetite)

Input material Electric/magnetic Relative εr’=43.48 εr’=50

data properties permittivity

Conductivity σ=0.87 S/m σ=3 S/m

Dielectric loss δe=0.49 δe=2.5

tangent

581
 Relative µr=1 µr=1.9
permeability

Magnetic loss δm=0 δm=0.3

tangent

Case 2 = One Average SAR_TOT [mW/kg] 51.10 54.40

sphere inside
ellipsoidal brain
model

Case 3 = Five Average SAR_TOT [mW/kg] 53.78 55.44

spheres inside
ellipsoidal brain
model

If 3.7 ‰ of the brain volume has also ferrimagnetic properties, the relative increase of average SAR
is 3.4 % due to magnetic losses (at the macroscopic scale).
Not the relative increase of average SAR is so significant, but the location of energy dissipation. In
case of five magnetite spheres spread in the brain volume, it is possible that VLD to be much higher
locally, in the vicinity of their surface (Fig. 1b).
4. Conclusion
The presence of macroscopic magnetite particles in a pure dielectric tissue may increase the
average SAR by few percents, depending on dimensions, number, position and material properties
of the magnetic particles. The most important aspect is however the possible occurence of deep hot
spots at the separation surface between magnetic and dielectric material. The simulations of
electromagnetic power dissipation in tissues containing magnetic particles with current software are
restricted to macroscopic scale situation since magnetic dissipation mechanisms that govern the
micro- and nano-scale magnetic crystals absorption are not mathematically implemented. That is
why RF micro-dosimetry of magnetic-content tissues remains a great challenge.

Figures

582
 Figure 1. a. The “magnetic brain” model with five unevenly distributed magnetite spheres
highlighted for volume loss density observation; b. A section plane in the “magnetic brain”
containing three out of five magnetite spheres and showing areas with intensified power
dissipation.

PB-18 [14:30]
Analysis of the core temperature distribution of the organs exposed to LTE signals
Sangbong Jeon1, Ae-kyoung Lee1 & Hyung-Do Choi1
1Radio & Satellite Research Division, ETRI, Daejeon, Korea, 34129
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Sangbong Jeon
In this paper, we simulated the thermal effects for rats in order to investigate the core temperature
elevation due to whole body averaged specific absorption rate of 4 W/kg at 1.76 GHz long term
evolution (LTE) RF signal bands. The temperature rise due to RF exposure depends on the location
of the organ, even the same organ. Therefore, the core temperature distribution of the organs in RF
exposure was calculated, and the average temperature and the standard deviation were obtained.
Introduction
The public concerns about the adverse effects of human exposure from mobile devices have been
steadily increasing. Although there are some differences in each country, we limit exposure to
mobile devices based on specific absorption rate (SAR) values. The SAR is related to the body
temperature rise caused by RF exposure, but the results are different depending on the researchers
[1-3].
Reverberation chambers are widely used as systems for animal experiments on RF exposure [4-5].
This system not only provides a statistically homogeneous field distribution when averaged over the
time of stirrer rotation within a specific volume in the chamber, but also allows the animals to ream
freely within their cages.
Modeling the field distributions in reverberation chambers has been typically obtained by reducing
the random incident wave to only 12 equal waves from six directions [5]. In this paper, the SAR
values used to evaluate core temperature were obtained by averaging over 12 plane waves in order
to realize the SAR in the reverberation chambers as a simulation model.

583
Results
We analyzed the core temperature rise of Sprague-Dawley rats, which the weight is 225 g, exposed
at whole body averaged SAR (WBA-SAR) of 4 W/kg in 1.76 GHz LTE bands. The core temperature
distribution showed the simulation results of analysis of the heart and large intestine in tissues of
rats in this paper. Fig. 1 shows the temperature distribution in heart organs exposed from six
directions and two polarizations, respectively. For the comparison of the temperature distributions,
the temperature range for the results was the same range from 37.2 °Cto 38.1 °C. As shown in Fig.
1, it can be seen that the distributions of temperature are different for every incident fields. Fig. 2
also shows the temperature distribution in the large intestine as in Fig. 1. The temperature range in
Fig. 2 was from 37.2°Cto 39.6 °C. At the core temperature of the large intestine organs, the hot spot
temperature was higher than that of the heart organs. As shown in Fig. 1 and Fig. 2, the tissue
temperature due to RF exposure varies depending on the position. The temperature of average and
standard deviation for heart and large intestine organs is shown in table 1. In Fig. 1 and Fig. 2, the
hot spot temperature showed a high large intestine, but the average temperature in table 1 shows
that the temperature of the heart and the large intestine are almost the same.
Table 1. The core temperature of average and standard deviation for heart and large intestine
organs.

Heart Large intestine

avg std avg std

Ex to y 37.309 0.113 37.359 0.159

Ex to –y 37.338 0.123 37.405 0.252

Ex to z 37.357 0.156 37.440 0.298

Ex to –z 37.391 0.175 37.414 0.135

Ey to x 37.320 0.117 37.391 0.188

Ey to –x 37.311 0.107 37.385 0.185

Ey to z 37.360 0.119 37.458 0.342

Ey to –z 37.363 0.125 37.386 0.145

Ez to x 37.385 0.175 37.417 0.331

Ez to –x 37.402 0.180 37.483 0.351

Ez to y 37.366 0.125 37.422 0.358

Ez to –y 37.399 0.160 37.452 0.375

584
 Total 37.358 0.145 37.418 0.278

Conclusions
We simulated the temperature for rats exposed at WBA-SAR of 4 W/kg in 1.7 GHz LTE bands. The
core temperature was analyzed for the heart and large intestine, and the temperature of the tissue
hot spot was different. However, the average temperature of the heart is 37.358 °C and the average
temperature of the large intestine is 37.418 °C. The average temperature of the both tissues was
almost the same, but the standard deviation was somewhat higher in the large intestine.
References
[1] H. J. Lee, Y. B. Jin, T. H. Kim, J. K. Pack, N. Kim, H. D. Choi, J. S. Lee, and Y. S. Lee, “The
effects of simultaneous combined exposure to CDMA and WCDMA electromagnetic fields on rat
testicular function,” Bioelectromagnetics, no. 33, pp. 356-364, 2012.
[2] J. Chakarothai, J. Wang, O. Fujiwara, K. Wake, and S. Watanabe, “Dosimetry of a reverberation
chamber for whole-body exposure of small animals,” IEEE Trans. Microw. Theory Techn., vol. 61,
no. 9, pp. 3435-3445, Sep. 2013.
[3] S. Ohtani, A. Ushiyama, M. Maeda, K. Hattori, N. Kunugita, J. Wang, and K. Ishii, “Exposure
time-dependent thermal effects of radiofrequency electromagnetic field exposure on the whole body
of rats,” J. Toxicol. Sci., vol. 41, no. 5, pp. 655-666, 2016.
[4] K. B. Jung, T. H. Kim, J. L. Kim, H. J. Doh, Y. C. Chung, J. H. Choi, and J. K. Pack, “Development
and validation of reverberation chamber type whole-body exposure system for mobile-phone
frequency,” Electromagn. Biol. Med., vol. 27, no. 1, pp. 73-82, 2008.
[5] M. H. Capstick, S. Kuehn, V. Berdinas-Torres, Y. Gong, P. F. Wilson, J. M. Ladbury, G. Koepke,
D. L. McCormick, J. Gauger, R. L. Melnick, and N. Kuster, “A radio frequency radiation exposure
system for rodents based on reverberation chambers,” IEEE Trans. Electromagn. Compat., vol. 59,
no. 4, pp. 1041-1052, Aug. 2017.
Acknowledgment
This work was supported by Institute for Information & communication Technology Promotion(IITP)
grant funded by the Korea government(MSIT) (2017-0-00961, Study on the EMF Exposure Control
in Smart Society).
Figures

585
Figure 1. The temperature distributions in heart organs exposed from six directions and two
polarizations.

Figure 2. The temperature distributions in large intestine organs exposed from six directions
and two polarizations.

586
PB-20 [14:30]
Alvar: Adult whole-body anatomic phantom for computational dosimetry
Ilkka Laakso1, Tuukka Lehtinen1 & Marco Soldati1
1Department of Electrical Engineering and Automation, Aalto University, Espoo, Finland, 02150
Keywords: Dosimetry (computational), All Frequencies, Work in Progress
Presented by: Ilkka Laakso
This paper presents an anatomically realistic voxel model of an adult male and its application to
modelling power-frequency electric and magnetic field exposure. The phantom, Alvar, is based on
anatomic atlases, weights 72 kg, and is 176 cm tall. Alvar and three other anatomic adult male
phantoms were used for dosimetric analysis of uniform electric and magnetic field exposure at the
frequency of 50 Hz, and the induced electric fields were shown to be in good agreement. Alvar is
directly compatible with freely available BodyParts3D anatomic atlas, which allows easy
identification of anatomic structures for dosimetry modelling.
Introduction
International guidelines for limiting human exposure to electromagnetic fields base largely on data
obtained from dosimetric studies (ICNIRP 2010). Accurate modelling of the human anatomy is
essential for dosimetric analysis. Several highly-detailed whole-body anatomic phantoms have been
developed (Dimbylow 1998, Nagaoka et al 2004, Kim et al 2008, Christ et al 2010, Wu et al 2011).
Typically, whole-body anatomic phantoms are voxel models based on segmented medical or
cryosection images, and thus, are specific to a single human subject. Also, they are based on
subjects in supine position. Dosimetry modelling can also be very sensitive numerical errors arising
from segmentation of tissue boundaries, for example, in the armpits and groin. To solve these
problems, we have developed a generic very human-like anatomic phantom that has been
specifically designed for robust numerical modelling.
Description of the body model
The phantom, Alvar, is based on freely available online anatomic atlas, BodyParts3D
(http://lifesciencedb.jp/bp3d/), which consists of 3D surface meshes of over 2000 organs or their
parts.
For making the Alvar phantom (pictured in Figure 1), each surface mesh was voxelized with a
uniform resolution of 0.5 mm x 0.5 mm x 0.5 mm. After voxelization, the size of the voxels was
increased to 0.5116 mm x 0.5116 mm x 0.5116 mm to match the height of the model with that of the
ICRP reference man (ICRP 2002). The final model consists of 513.4 million non-air voxels.
The model is divided into 82 tissues, organs, anatomic regions, and/or bodily fluids, which are (in
alphabetical order): accumbens, adrenal gland, air, amygdala, artery, bile, bladder, blood, brainstem,
bronchi, cancellous bone, cartilage, caudate, cavernous body, cerebellar grey matter, cerebellar
white matter, cortical bone, CSF, diaphragm, duodenum, dura, ear cartilage, epididymis, esophagus,
eye anterior chamber, eye choroid, eye cornea, eye humour, eye iris, eye lens, eye retina, eye
sclera, eye suspensory ligament, fat, gall bladder, gingiva, gland, grey matter, heart, hippocampus,
intervertebral disk, kidney, lacrimal duct, lacrimal gland, large intestine, large intestine contents,
larynx, ligament, liver, lung, meninges, mesentery, muscle, nerve, pallidum, pancreas, prostate,
putamen, salivary gland, skin, small intestine, small intestine contents, spinal cord, spleen, stomach,
stomach contents, teeth, tendon, testis, thalamus, thymus, thyroid gland, tongue, trachea, ureter,
urethra, urine, vas deferens, vein, ventral diencephalon, ventricular CSF, and white matter.
To make the model suitable for electromagnetic dosimetry, several modifications were applied when
the original surface-based models were voxelized. The most important are listed below.
Brain
The brain of Alvar is based on ICBM 2009a nonlinear asymmetric template MR images, which
represent the average brain of 152 healthy individuals (Fonov et al 2009, 2011). The FreeSurfer

587
image analysis software (Dale et al 1999) was used to segment the template images to white and
grey matter, brainstem, and subcortical nuclei. An affine transformation was applied to map the
template brain to the coordinates of Alvar.
Skin and fat
The original skin model of the BodyParts3D database does not fully follow the boundaries of other
tissues. At several occasions, other tissues ”pop out” from the skin, and at some anatomic sites, the
subcutaneous space is excessively thick. To correct these problems, the original skin was
completely replaced. The new skin is 2 mm thick and smoothly follows the boundaries of muscles,
bones, vasculature, and internal organs.
The BodyParts3D atlas does not include subcutaneous fat or fat that surrounds internal organs. In
Alvar, all cavities between tissues and the subcutaneous cavity are filled with fat. Consequently, all
tissues that are not included, e.g., peripheral nerves, are automatically modelled as fat.
Other changes
Because the muscle mass of the BodyParts3D model was less than that of the ICRP reference man,
the size of all muscles was slightly increased. Approximations of facial muscles were also added
because they were not included in the BodyParts3D atlas.
A model of the spinal canal was generated based on the data of the cross sectional areas of the
spinal subarachnoid space and the spinal cord (Edsbagge et al 2011). Furthermore, the model of
the spinal cord was replaced.
Bones were segmented to cortical bone and cancellous bone (including bone marrow). The outer
layer, with a thickness of 1.5 mm was segmented as cortical bone, and the rest was treated as
cancellous bone or bone marrow.
Comparison with ICRP reference man
Alvar weights 72 kg, is 176 cm tall, and has a surface area of 1.83 m2. The ICRP reference values
are 73 kg, 176 cm, and 1.90 m2, respectively (ICRP 2002). The below table shows a comparison of
the tissue masses of Alvar to those of the ICRP reference male.
Table 1. Some organ masses (g) of ICRP reference male (ICRP 2002) and Alvar sorted in the
descending order by tissue mass.
ICRP Alvar

Skeletal muscle 29000 27620

Total adipose tissue 18200 18630

Total skeleton 10500 11610

Cortical bone 4400 4630

Skin 3300 3680

Liver 1800 1570

Brain 1450 1690

Dosimetry of 50 Hz magnetic and electric field exposure

The model was exposed to a uniform vertical electric field of 1 kV/m or a uniform magnetic field of 1
mT in the anterior-posterior direction. The electric fields were compared to three existing whole-body
adult male models: Taro (Nagaoka et al 2004), Duke (Christ et al 2010), and Norman (Dimbylow
1998). These models had a resolution of 2 mm. To make the Alvar phantom compatible with the

588
other models, it was downscaled to the resolution of 2.047 x 2.047 x 2.047 mm.
The internal electric fields were modelled using the quasi-static finite element method with cubical
first-order elements (Laakso and Hirata 2012). The method of successive nested grids (Dimbylow
2000) was used for modelling electric field exposure. Tissue conductivities were based on the study
of Dimbylow (2005).
For analysing the electric fields, models were divided into three regions: (1) brain, (2) all other
tissues in the torso and head, and (3) limbs. For each region, the 999th permille electric field values
were calculated. The most exposed tissue in each region was determined by finding the tissue with
the largest number of voxels in the highest permille.
The results for electric and magnetic field exposure are shown in Figures 2 and 3 and Tables 2 and
3. Overall, all four phantoms provide similar electric field distributions (Figures 2 and 3). In the brain
tissue, all phantoms produce very consistent electric fields. Other regions show larger differences. In
terms of the electric fields, the Alvar phantom is very similar to the Taro model.
Table 2. 999th permille in-situ electric fields (mV/m) in human body models exposed to vertical 1
kV/m external electric field.
Alvar Taro Duke Norman

1.20 (cerebellar 1.23 (white 1.18 (grey

Brain 1.19 (brain)
grey matter) matter) matter)

1.75 (cortical 2.50 (cortical

Head and torso 2.21 (fat) 1.89 (fat)
bone) bone)

3.51 (cortical 4.88 (cortical

Limbs 3.36 (fat) 5.69 (fat)
bone) bone)

Table 3. 999th permille in-situ electric fields (mV/m) in human body models exposed to anterior-
posterior 1 mT magnetic flux density.

Alvar Taro Duke Norman

Brain 33 (white matter) 35 (grey matter) 33 (grey matter) 44 (brain)

118 (cortical 140 (cortical

Head and torso 108 (fat) 109 (fat)
bone) bone)

Limbs 63 (fat) 63 (fat) 42 (fat) 116 (skin)

Discussion
The main advantage of Alvar is that the model is directly compatible with BodyParts3D, a freely-
available online anatomic atlas. Therefore, all anatomic structures, e.g. bones, blood vessels, and
muscles, can be individually identified. This is especially useful when highly detailed dosimetry is
needed. For instance, we have previously used an earlier version of Alvar to create a very high
resolution (0.1 mm) model of the eye and its surrounding tissues for dosimetry of millimetre waves
(Laakso et al 2017). Further, the model has been recently used for investigating the electric fields
induced in blood vessels of the arm (to be presented at BioEM 2018).
There are still limitations in the Alvar phantom. The peripheral nervous system is missing, which is a
problem also in other whole-body human models. Similarly, arteries and veins of the head are
missing. And, as the pictures show, the face of Alvar is not especially pretty.
Alvar will be made freely available under a Creative Commons Attribution–ShareAlike licence, which
permits copying, redistributing, remixing and transforming the model data in any medium or format

589
for any purpose.
Acknowledgements
Alvar is based on BodyParts3D (http://lifesciencedb.jp/bp3d/) developed at the Database Center for
Life Science, Japan.
The brain of Alvar is based on ICBM 152 Nonlinear atlases version 2009
(http://www.bic.mni.mcgill.ca/ServicesAtlases/ICBM152NLin2009) developed at McConnell Brain
Imaging Centre, Montreal Neurological Institute, McGill University, Canada.
References
Christ A, Kainz W, Hahn EG, Honegger K, Zefferer M, Neufeld E, Rascher W, Janka R, Bautz W,
Chen J, Kiefer B, Schmitt P, Hollenbach HP, Shen J, Oberle M, Szczerba D, Kam A, Guag JW and
Kuster N 2010 The Virtual Family—development of surface-based anatomical models of two adults
and two children for dosimetric simulations Phys Med Biol 55(2)N23–38
Dale AM, Fischl B and Sereno MI 1999 Cortical surface-based analysis. i. segmentation and surface
reconstruction. Neuroimage 9(2)179–94
Dimbylow P 2005 Resonance behaviour of whole-body averaged specific energy absorption rate
(SAR) in the female voxel model, NAOMI Phys Med Biol 50(17)4053–63
Dimbylow PJ 1998 Induced current densities from low-frequency magnetic fields in a 2 mm
resolution, anatomically realistic model of the body Phys Med Biol 43(2)221–30
Dimbylow PJ 2000 Current densities in a 2 mm resolution anatomically realistic model of the body
induced by low frequency electric fields Phys Med Biol 45(4)1013–22
Edsbagge M, Starck G, Zetterberg H, Ziegelitz D and Wikkelso C 2011 Spinal cerebrospinal fluid
volume in healthy elderly individuals. Clin Anat 24(6)733–740
Fonov V, Evans AC, Botteron K, Almli CR, McKinstry RC, Collins DL and Brain Development
Cooperative Group 2011 Unbiased average age-appropriate atlases for pediatric studies.
Neuroimage 54(1)313–27
Fonov V, Evans A, McKinstry R, Almli C and Collins D 2009 Unbiased nonlinear average age-
appropriate brain templates from birth to adulthood. Neuroimage 47, Supplement 1(0)S102
ICNIRP 2010 Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz to
100 kHz) Health Phys 99(6)818–36
ICRP 2002 Basic anatomical and physiological data for use in radiological protection: reference
values ICRP Publication 89
Kim CH, Choi SH, Jeong JH, Lee C and Chung MS 2008 HDRK-Man: a whole-body voxel model
based on high-resolution color slice images of a Korean adult male cadaver Phys Med Biol
53(15)4093–106
Laakso I and Hirata A 2012 Reducing the staircasing error in computational dosimetry of low-
frequency electromagnetic fields Phys Med Biol 57(4)N25–34
Laakso I, Morimoto R, Heinonen J, Jokela K and Hirata A 2017 Human exposure to pulsed fields in
the frequency range from 6 to 100 ghz. Physics in medicine and biology 626980–6992
Nagaoka T, Watanabe S, Sakurai K, Kunieda E, Watanabe S, Taki M and Yamanaka Y 2004
Development of realistic high-resolution whole-body voxel models of Japanese adult males and
females of average height and weight, and application of models to radio-frequency
electromagnetic-field dosimetry Phys Med Biol 49(1)1–15
Wu T, Tan L, Shao Q, Zhang C, Zhao C, Li Y, Conil E, Hadjem A, Wiart J, Lu B, Xiao L, Wang N, Xie
Y and Zhang S 2011 Chinese adult anatomical models and the application in evaluation of rf
exposures. Physics in medicine and biology 562075–2089
Figures

590
Figure 1. Alvar model. Left: Overview of the model. Tissues have been cut to make internal
organs visible. Right: Panels show the skin, muscles, skeleton, vasculature and central
nervous system.

591
Figure 2. Exposure to uniform magnetic flux density of 1 mT in the anterior-posterior direction
at 50 Hz. Comparison of induced electric fields in Alvar and three other anatomic adult male
models. The induced electric fields are visualized at a depth of 2 mm below the skin.

Figure 3. Exposure to uniform vertical electric field of 1 kV/m at 50 Hz under isolated

conditions. Comparison of induced electric fields in Alvar and three other anatomic adult male
models. The induced electric fields are visualized at a depth of 2 mm below the skin.

592
PB-22 [14:30]
Evaluation of heating effects due to millimeter wave exposure in oblique incidence
Kun Li1, Kensuke Sasaki1 & Soichi Watanabe1
1National Institute of Information and Communications Technology, Koganei, Japan, 184-8795
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Kun Li
In order to realize the ultra-high data wireless communication, millimeter wave frequencies over 6
GHz are anticipated to be used in many applications, such as 5G terminals, wireless gigabit and
automotive radar. As the basic restriction in safety guideline, power density needs to be related with
the temperature elevation of skin tissue exposed to electromagnetic waves at MMW bands. In this
work, we analyzed the heat potential of incident and absorption power density from 6 GHz to 1 THz
for cases of oblique incidence. The results show that absorption power density indicates a high
correlation with skin surface temperature elevation regardless of frequency, incident angle and
polarization effects.
Background
Owing to the potential high-data-rate transmission and usable bandwidth, millimeter wave (MMW)
frequencies are expected to be used in future wireless system [1]. In these frequency regions,
however, the opportunities of human exposure to electromagnetic waves will be increased in a
general public environment [2]. Today, the international safety guidelines have been published by
IEEE [3] and ICNIRP [4] [5], where the incident power density is used as the dosimetric quantity for
MMW exposure. In previous studies, the correlation between the incident power density and the
resultant temperature elevation in MMW frequency bands have been investigated [6] [7].
On the other hand, absorption power density into the skin tissue is also an appropriate candidate as
dosimetric quantity because the electromagnetic waves at MMW bands are mainly absorbed within
the human skin surface. However, absorption power density cannot easily be measured owing to a
shallow penetration depth. Moreover, it is significantly depended on the frequency, incident wave
angle, and polarization effects. This study aims to clarify the relationships between the incident
power density, absorption power density, and skin temperature elevation for cases of obliquely
incident electromagnetic waves at frequencies from 6 GHz to 1 THz.
Analytical results
In this study, we used an analytical model composed of epidermis (i = 1), dermis (i = 2), fat (i = 3),
and muscle (i = 4) used to represent the skin tissue in the forearm [8]. An obliquely-incident plane
wave with an incident power density of p = 10 W/m2 from air to the skin surface is assumed. The
steady-state temperature analysis upon electromagnetic wave absorption is calculated by solving
the boundary conditions of each tissue interface using a bioheat equation [8].
Figure 1 shows the surface temperature elevation ΔT as a function of frequency from 6 GHz to 1
THz. In Fig. 1, when the incident power density p is 10 W/m2, the curves of ΔT for both TE and TM
waves at θ0 = 60 deg, are lower than those in the case of normal incidence. Even if at the incident
angles with the maximum transmittance (θ0 = θmax), the temperature elevation does not exceed that
at normal incidence, indicating that the normal incidence represents the worst-case exposure
condition if incident power density p is defined in an area normal to the incident direction.
Figure 2 shows the ratio of the incident power density p and absorption power density pa to the
surface temperature elevation ΔT in the case of exposure to oblique incident waves with the incident
wave angles of θ0 = 0 deg, 60 deg, and θmax. As can be seen in Fig. 2, with increasing frequency
from 6 GHz to 1 THz, the ratio of p to ΔT changes obviously with the frequency, incident angles and
polarization components, which suggests that the correlation between the incident power density
and the surface temperature elevation is relatively low. In the entire frequency region from 6 GHz to

593
1 THz, when θ0 = 60 deg, the TE and TM waves show average differences of 4.8 and 1.9 dB,
respectively, from that of the normal incidence.
On the other hand, in the cases of the ratio of the power density at z = z1 inside the skin pa to ΔT, all
the results show the flat profiles in the entire frequency region from 6 GHz to 1 THz and coincide
well with each other regardless of incident angles. Note that the results for TE and TM waves at θ0 =
60 deg are in good agreement, indicating that the effects of polarization have no impact on the
correlation between the absorption power density and the surface temperature elevation.
Conclusion
The relationships between the incident power density, absorption power density, and skin
temperature elevation for cases of oblique incidence is analyzed using a fat multilayer model. The
results indicate that absorbed power density into the skin tissue has a high correlation with the
temperature elevation on skin surface regardless of incident angle, frequency and polarization.
Thus, it can be considered as the basic restriction in the safety guidelines for radio wave exposure
at MMW and THz bands.
Acknowledgement
This work was partly supported by the Ministry of Internal Affairs and Communications, Japan.
References
1. T. S. Rappaport, S. Sun, R. Mayzus, H. Zhao, Y. Azar, K. Wang, G. N. Wong, J. K. Schulz, M.
Samimi, and F. Gutierrez, “Millimeter wave mobile communications for 5G cellular: It will work!” IEEE
Access, vol. 1, no. 1, pp. 335-349, 2013.
2. M. Zhadobov, N. Chahat, R. Sauleau, C. L. Quement, and Y. L. Drean, “Millimeter-wave
interactions with the human body: state of knowledge and recent advances,” Int. J. Microwave
Wireless Technol., vol. 3, no. 2, pp. 237-247, 2011.
3. IEEE Standard for Safety Levels with Respect to Human Exposure to the Radio Frequency
Electromagnetic Fields, 3 kHz to 300 GHz, IEEE Std., C95.1, 2005.
4. International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines for limiting
exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz),” Health
Phys., vol. 74, pp. 494-522, 1998.
5. International Commission on Non-Ionizing Radiation Protection (ICNIRP), “Guidelines on limits of
exposure to laser radiation of wavelength between 180 nm and 1000 mm,” Health Phys., vol. 105,
no. 3, pp. 271-295, 2013.
6. K. R. Foster, A. Lozano-Nieto, and P. J. Riu, “Heating of tissues by microwaves: a model
analysis,” Bioelectromagnetics, vol. 19, no. 7, pp. 420-428, 1998.
7. Y. Hashimoto, A. Hirata, R. Morimoto, S. Aonuma, I. Laakso, K. Jokela, and K. R. Foster, “On the
averaging area for incident power density for human exposure limits at frequencies over 6 GHz,”
Phys. Med. Biol., vol. 62, no. 8, pp. 3124-3138, 2017.
8. K. Sasaki, M. Mizuno, K. Wake, and S. Watanabe, “Monte Carlo simulations of skin exposure to
electromagnetic field from 10 GHz to 1 THz,” Phys. Med. Biol., vol. 62, no. 17, pp. 6993-7010, 2017.
Figures

594
Figure 1. Surface temperature elevation for normal and oblique incidence scenarios as a
function of frequency from 6 GHz to 1 THz when incident power density p =10 W/m2 is defined
in an area normal to the incident direction.

Figure 2. Heat potential of incident and absorption power density as a function of frequency
from 6 GHz to 1 THz for incident power density p =10 W/m2 is defined in an area normal to the
incident direction.

595
PB-24 [14:30]
Ad hoc simplified dosimetry analysis of in-vivo EMF exposure of rats and mice in
reverberation chamber
Antonio Sarolic1
1Chair of Applied Electromagnetics - EMLab, FESB, University of Split, Split, Croatia, HR-21000
Keywords: Dosimetry (computational), RF/Microwaves, Concept
Presented by: Antonio Sarolic
We previously proposed a novel method for calculating the whole-body SAR of biological objects in
reverberating environment, based on reverberation chamber and power balance theory. This
method is applied here to demonstrate an alternative approach to dosimetry analysis of the setup
described in recent publications on the study conducted by the National Toxicology Program of the
National Institute of Environmental Health Sciences (two-year cancer bioassay study on rodents
exposed in reverberation chambers), as an interesting example where this method perfectly fits to
the problem.
The novel method of calculating the whole-body specific absorption rate (WBSAR) of lossy objects
in reverberating environment, using reverberation chamber and power balance theory, was first
proposed at BioEM 2015 [1]. WBSAR can be determined from the power transmitted into the
chamber Pt, unloaded reverberation chamber quality factor Qunloaded and volume V, the object
average absorption cross section <ACS> and mass m:
We verified this novel method by thermal
measurements of the actual SAR in liquid
phantoms, and also applied the method to
consider human exposure in reverberating
environments [1]–[3]. Considering the recently
completed and published two-year cancer
bioassay study on rodents exposed in reverberation chambers, conducted by the National
Toxicology Program of the National Institute of Environmental Health Sciences, this method can be
used in conjunction with the recently published dosimetry [4, 5], to determine e.g. the ACS of rats
and mice, and the required transmitted power to achieve the target WBSAR values in the described
setup.
The method is used here only as an ad hoc simplified calculation. The unloaded chamber quality
factor Qunloaded is estimated based on the published geometrical dimensions (but could be actually
measured to obtain the accurate value). The quality factor of a rectangular cavity Qunloaded is given
by [6] as:
where V is the cavity volume (V = width x length x height) and A is the
cavity surface (A = 2 x length x width + 2 x width x height + 2 x length x
height), and δ is the skin depth of the wall material. According to the
publications [4, 5] the chamber dimensions are 2.2 x 3.7 x 2.6 m3, and it
is made of stainless steel so the skin depth can be calculated at the frequencies used in the study
(900 MHz for rats and 1900 MHz for mice). The obtained quality factors are Qunloaded900 = 48511 for
900 MHz and Qunloaded1800 = 70484 for 1800 MHz. These theoretical quality factors will be used as
the quality factors for the unloaded chamber, i.e. the chamber without the rodents, although the
actual quality factor without the rodents can actually be lower, due to presence of other losses in the
chamber (e.g. openings, lossy objects, antennas etc.). The presence of such losses is not
accounted for in this simplified analysis.
ACS of rodents as a function of their mass can be implicitly derived from the results published in [4,
Fig.4] (data extracted using software extraction tool [7]). Starting from the definition of ACS as the
ratio of the power Pr absorbed by the object and the incident power density Sinc, we derive ACS as
the function of (WBSAR/Einc2) ratio vs. rodent mass m:

596
 The ratio (WBSAR/Einc2) is provided in [4, Fig.4]
(noting and taking into account that [4, Fig.4] is
actually in (μW/kg)/(V/m)^2 as can be observed in [4,
Table VI]). The obtained function of ACS vs. mass is
given in Fig.1 for rats and mice.
Finally, the required power transmitted in the chamber to achieve the target WBSAR values from the
said study can be calculated as:
and it is given in Fig.2 (target WBSAR of
1.5, 3.0, and 6.0 W/kg for rats; 2.5, 5.0, and
10 W/kg for mice). The required power is
calculated by entering total mass and total
ACS of all rats (N=112) and all mice (N=216)
into Eq.4, assuming the exposure setup is engineered so that each animal is exposed from all sides
without any shadowing. However, due to large quality factor of the chamber, the role of rodent ACS
(even in total) in this calculation is minimal, and the required power is linearly dependent on the total
mass of the exposed objects.
The obtained results are illustrative. The calculated power is within ranges mentioned in [4, 5],
although the exact actual values were not provided in [4, 5] so the actual comparison cannot be
performed in this analysis. It is worth noting that this is the theoretically minimal required power -
any aforementioned additional losses in the chamber would further decrease the chamber quality
factor, thus further increasing the required transmitted power.
Acknowledgements:
This study is performed within the framework of the research projects "Measurements in
Bioelectromagnetics (M-BEM)" supported by the University of Split, "Measurements in EMC and EM
health effects research" supported by Ministry of Science, Education and Sports of the Republic of
Croatia, and COST Action BM1309 (COST EMF-MED).
References:
[1] A. Sarolić, D. Senić, "Novel method for practical SAR determination in indoor exposure scenarios
without the need for EMF measurements, based on reverberation chamber theory", BioEM 2015,
Asilomar, 14-19 July 2015, pp. 324-328.
[2] D. Senić, A. Sarolić, J. M. Joskiewicz, and C. L. Holloway, “Absorption cross-section
measurements of a human model in a reverberation chamber,” IEEE Trans. Electromagn. Compat.,
vol. 58, no. 3, pp. 721– 728, June 2016.
[3] D. Senić, A. Sarolić, C. L. Holloway, and J. M. Ladbury, "Whole-body specific absorption rate
assessment of lossy objects exposed to a diffuse field inside a reverberant environment," in IEEE
Transactions on Electromagnetic Compatibility, vol. 59, no. 3, pp. 813-822, June 2017.
[4] M. H. Capstick et al., "A Radio Frequency Radiation Exposure System for Rodents Based on
Reverberation Chambers," in IEEE Transactions on Electromagnetic Compatibility, vol. 59, no. 4, pp.
1041-1052, Aug. 2017.
[5] Y. Gong et al., "Life-Time Dosimetric Assessment for Mice and Rats Exposed in Reverberation
Chambers for the Two-Year NTP Cancer Bioassay Study on Cell Phone Radiation," in IEEE
Transactions on Electromagnetic Compatibility, vol. 59, no. 6, pp. 1798-1808, Dec. 2017.
[6] D. A. Hill, Electromagnetic fields in cavities, Piscataway: Wiley IEEE Press, 2009.
[7] https://automeris.io/WebPlotDigitizer/

Figures

597
 Figure 1. Absorption cross section of a rat (left) and mouse (right)

Figure 2. Required transmitted power in the chamber for 112 rats (left) and 216 mice (right)

PB-26 [14:30]
Realistic EMF exposure assessment of 5G massive MIMO base stations
Bo Xu1, Björn Thors1, Davide Colombi1, Elif Degirmenci1, Paramananda Joshi1 & Christer Tornevik1
1Ericsson Research, Ericsson AB, Stockholm, Sweden, 164 80
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Bo Xu
The traditional approach to assess RF EMF exposure, which uses the theoretical maximum output
power, would lead to very conservative compliance boundaries for a 5G base station (BS)
employing massive MIMO. Here, we start by giving the analytical box-shaped compliance boundary
for theoretical maximum output power. Then, a statistically-conservative power reduction factor is
introduced considering TDD downlink/uplink ratio, BS utilization, and scheduling time. As an
example, the theoretical compliance boundaries of a 5G massive MIMO BS product are given.
Considering the proposed power reduction factor, the actual compliance boundaries are also given
for the 95th percentile of all possible exposure scenarios, according to IEC 62232:2017.

598
Introduction
The fifth-generation mobile communication networks (5G) are currently being researched and
standardized. Advanced antenna technologies, such as massive multi-input-multi-output (MIMO)
and beamforming with phased array antennas, will play important roles [1], [2]. A differentiating
factor between 5G New Radio (NR) [3] and previous radio access technologies is the ‘beam centric
design’, which aims to transmit energy in the directions where it is needed rather than to constantly
transmit energy in a wide angular sector. This focusing of energy in different directions will also
impact the realistic exposure of 5G radio base station (BS) products.
A radio frequency (RF) electromagnetic field (EMF) exposure assessment of a 5G NR system
employing massive MIMO based on the traditional approach would assume that the theoretical
maximum power is transmitted in each possible direction. This is very unrealistic and would lead to
a very conservative compliance boundary consistent with the envelope of all possible array
excitations/beams, where each beam is transmitting at the theoretical maximum output power for a
time-period in the order of minutes corresponding to the averaging time of the relevant RF exposure
limits. Factors such as BS utilization, time-division duplex (TDD), scheduling time, and spatial
distribution of served users are then not considered.
In the recently adopted international standard IEC 62232:2017 [4], it is specified that RF EMF
exposure assessments may be conducted for ‘actual maximum’ exposure conditions corresponding
to the 95th percentile of all possible exposure scenarios. In [5], a far-field model for time-averaged
actual maximum power levels for assessment of RF EMF exposure for 5G BS employing massive
MIMO is presented. The model is based on a statistical approach and developed to provide a
realistic conservative RF exposure assessment for a significant proportion of all possible downlink
exposure scenarios (95th percentile) in-line with requirements in IEC 62232:2017 [4].
In this abstract, the approach in [5] is used to obtain both theoretical and actual maximum box-
shaped compliance boundaries of a 5G BS product operating at 3.5 GHz. Compliance is assessed
with respect to the ICNIRP general public (10 W/m2) and occupational (50 W/m2) reference levels.
Methods
Theoretical Box-Shaped Compliance Boundary
A box-shaped compliance boundary is used characterized by the BS dimensions and the
compliance distances in front of, behind, above/below, and to the sides of the BS (see Figure 1).
Outside this box, the RF exposure is below the exposure limits. Due to the highly directive array
antenna designed to provide coverage in a certain angular range from the broadside direction, the
compliance distance in the back direction is 0 cm (touch compliance) for both occupational and
general public exposure conditions.
For larger scan angles, the gain loss normally increases due to the reduced effective aperture and
an increased impedance mismatch. The realized gain, for a large aperture with area A, when
scanned an angle α from the broadside in a plane, may be written as [6]

where λ, εL, εA, and Γ denote the wavelength, the loss efficiency, the aperture efficiency, and the
active reflection coefficient, respectively. With the conservative assumptions of no losses and a
perfect impedance match, irrespective of scan angle, the gain for the large aperture becomes

With (Eq.2) inserted in the far-field formula, the front compliance distance may be written as

599
where P and Slim denote the theoretical maximum transmitted power and the relevant RF exposure
limit, respectively. To estimate the side compliance distance, the applicable scan range and beam
shape of the scanned beams need to be considered according to Figure 1. The beam shape may be
approximated by a Gaussian beam and the side compliance distance in the horizontal direction for a
beam scanned towards α may be written as [6]

where αmax,H and HHPBW denote the maximum scan angle in the horizontal plane and the
horizontal half-power beamwidth, respectively. Furthermore,

where Ny and Δy denote the number of elements and the inter-element distance in the horizontal
direction, respectively. The compliance distance above/below the BS can be expressed in terms of
beam-scanning in the vertical direction in a similar way.
Actual Maximum Power Levels
In addition to effects of beam steering, factors such as TDD downlink/uplink ratio, BS utilization, and
scheduling time are reasonably foreseeable and will impact the actual maximum transmitted power.
For massive MIMO BS products, the actual maximum transmit power depends on the expected
distribution of connected users and thereby on how the time-averaged transmitted power is
distributed within the scan range of the array antenna. As stated in [5], the expectation of the
statistically conservative fraction of the total power contributing to the EMF exposure in the
broadside beam direction can be written as

where ρ is the system utilization, is the power fraction, n is the number of users, Ts is the total
average scheduling time per user, T is the averaging time, kBB is the number of users served
simultaneously by the broadside beam, kADJ is the number of users that may be scheduled
simultaneously by the most adjacent beams, kSL is the number of users that may be scheduled
simultaneously by other than the broadside beam and its adjacent beams, and δADJ and δSL are the
weighting factors of the corresponding beams.
For a required level of certainty, e.g. 95th percentile, the maximum value of (Eq.8) for all possible
system utilizations ρ may be used to decide the statistical power reduction factor, FSTAT. Moreover,
considering a deterministic power reduction factor for the downlink/uplink ratio, FTDD, the total
power reduction factor, which is the ratio of the actual maximum power to the theoretical maximum
power, can be written as

600
Results
As an example, a BS product operates in the 3.5 GHz frequency band and consists of 8 × 8 antenna
elements. The theoretical maximum output power is 200 W and the peak beam gain is 24.3 dBi. It
can scan ± 60° in the horizontal plane and between 73° and 113° in the vertical plane.
CDFs corresponding to (Eq.8) for rural and urban scenarios are provided in Figure 2. For the rural
scenario, users are distributed horizontally with the highest user density towards broadside and with
the assumption of no vertical beam scanning. For the urban scenario, users are distributed both
horizontally and vertically with the highest user density towards broadside. Based on the results for
this product shown in Figure 2, a statistical power reduction factor FSTAT of 0.3 may be used for
rural installations and a factor of 0.2 for urban sites (95th percentile) [5].
In Figure 3, a lateral view of the theoretical and actual maximum compliance boundaries for ICNIRP
general public and occupational EMF exposure in the rural scenario is shown. The front compliance
distances for the theoretical maximum power (200 W) is 20.7 m and 9.3 m for the general public and
occupational EMF exposure, respectively. Considering a total power reduction factor PRF = FTDD ×
FSTAT = 0.75 × 0.2 = 0.15, the compliance boundary for the actual maximum power (30 W) is 8.0 m
and 3.6 m for the general public and occupational EMF exposure, respectively.
Discussion
The gain of the scan pattern in this study, according to (Eq.1)–(Eq.7), is based on analytical
expressions, and have some discrepancies with the realistic scan pattern. However, measuring all
the beam patterns for massive MIMO is time-consuming and sometimes unrealistic, e.g., in case of
reciprocity-based beamforming. For small cell BS products operating relatively close to people, the
far-field approximation may not be applicable. Instead, the embedded pattern approach may be
introduced in exposure assessment for such kind of BS products. In practice, other factors affecting
the actual power levels can also be added to the total power reduction factor (Eq.9), such as power
tolerance, control channel transmission, etc.
Conclusion
In this abstract, the box-shaped compliance boundaries of a typical 5G massive MIMO BS product
are given with respect to the theoretical and actual maximum power levels using an analytical
approach. The use of theoretical maximum compliance boundaries is extremely conservative and
may make the installation of new 5G BS products difficult. By basing the compliance assessment on
actual maximum power levels (95th percentile), considering effects of TDD downlink/uplink ratio, BS
utilization, and scheduling time, the compliance boundary dimensions may be reduced with a factor
of at least two. The presented compliance boundaries are given with respect to the ICNIRP general
public and occupational reference levels.
References
[1] Ericsson, “5G systems,” Ericsson, Stockholm, Sweden, White Paper 284 23-3251 Uen, Jan.
2017.
[2] T. L. Marzetta, “Massive MIMO: An introduction,” Bell Labs Tech. J., vol. 20, pp. 11-22, Mar.
2015.
[3] E. Dahlman et al., “5G wireless access: Requirements and realization,” IEEE Commun. Mag.,
vol. 52, no. 12, pp. 42-47, Dec. 2014.
[4] Determination of RF Field Strength and SAR in the Vicinity of Radio Communication Base
Stations for the Purpose of Evaluating Human Exposure, IEC Standard IEC 62232:2017, Aug. 2017.
[5] B. Thors, A. Furuskär, D. Colombi, and C. Törnevik, “Time-averaged realistic maximum power
levels for the assessment of radio frequency exposure for 5G radio base stations using massive
MIMO,” IEEE Access, vol. 5, pp. 19711 – 19719, Sep. 2017.

601
[6] Mailloux, “Phased array antenna handbook”, Artech House, 2005.

Figures

Figure 1. Box-shaped compliance boundary of massive MIMO BS

Figure 2. CDFs of the statistical power reduction factor for rural and urban installation
scenarios of the 5G massive MIMO BS product.

602
 Figure 3. Lateral view of compliance boundaries using the theoretical maximum power of 200
W and the actual maximum power of 30 W, to be compliant with ICNIRP general public and
occupational EMF exposure limits for an urban installation scenario.

PB-28 [14:30]
Evaluation of exposure to electromagnetic field emitted by RFID HF desktop readers
Patryk Zradziński1, Jolanta Karpowicz1 & Krzysztof Gryz1
1Central Institute for Labour Protection, National Research Institute (CIOP-PIB), Warszawa, Poland, 00-701
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Patryk Zradziński
Radio frequency identification (RFID) is the most common and the fastest evolving wireless
identification technology. The thermal effects of exposure to electromagnetic field were evaluated by
numerical simulations focused on the specific absorption rate (SAR) values in virtual phantom
present in the vicinity of RFID HF desktop readers operating at 13.56 MHz, at various reading
ranges. The results of numerical simulations showed that SAR values from RFID HF desktop
readers may exceed general public and occupational exposure limits, during at least 6 minute
exposure, near the readers operating at maximum reading range of 150 cm.
Introduction
Radio frequency identification (RFID) is the most common and the fastest evolving wireless
identification technology, used worldwide in various areas of daily life and work as for example:
access control to buildings or rooms, road toll systems, contactless payment cards, public transport
cards, control of time in sport. It is also successfully used in libraries and shops.

603
One of the most common used RFID system is high frequency (HF) system operating in frequency
range (3 – 30) MHz – typically 13.56 MHz. There are two standards related to RFID HF readers:
ISO/IEC 14443-2:2016 – for proximity cards with reading range up to over a dozen cm (short range
(SR) readers) and ISO/IEC 15693-1:2010 – vicinity cards with reading range up to 1.5 m (medium
range (MR) readers).
RFID HF desktop readers are hand operated devices. Their users head and torso are often present
close to working reader. In such cases direct biophysical effects of exposure should be assessed.
The main parameter characterizing direct biophysical result of exposure to high frequency
electromagnetic field in the environmental safety considerations is called specific absorption rate
(SAR). SAR values are typically estimated by numerical simulations with the use of human body
models. The evaluation the most frequently uses international exposure limits for general public
(SAR averaged over the whole body, SARwb = 0.08 W/kg, and SAR localised values in any 10g of
tissues in head and trunk, SARht = 2 W/kg, both averaged over any 6 minutes of exposure) or 5-fold
higher values for workers exposure.
The aim of this study was to evaluate the SAR values in the users of RFID HF desktop readers in
various exposure cases.
Methods
The source of exposure to electromagnetic field of 13.56 MHz was modelled as loop antenna of
dimensions of 25 cm x 25 cm located in front of model of human body at distance of 5 cm to chest.
Investigations were carried out for horizontal or vertical readers positions, and operating at settings
of maximum or one-third of maximum reading ranges: for SR readers 18 cm or 6 cm, and for MR
readers 150 cm or 50 cm.
Simulations of SAR were carried out by Sim4Life (Zurich Med Tech, Switzerland) software with
expanded uncertainty estimated as ± (30-40)%.
Results and Discussion
The highest values of whole-body average SAR (SARwb = 7.5 W/kg) and localised 10g SAR in
head and trunk (SARht = 168 W/kg) were found near MR readers operating at maximum reading
range of 150 cm and vertical reader position. Corresponding values calculated in model with SR
readers operating at maximum reading range of 18 cm were equal to: 0.01 W/kg (SARwb) and 0.25
W/kg (SARht).
The SAR values obtained in models with horizontal readers position were 4-fold lower than values
obtained near vertical readers.
SAR values obtained in models with readers operating at reading ranges equal to one-third of
maximum reading ranges were 10-fold (SR) and 400-fold (MR) lower than values obtained for
maximum reading ranges.
Conclusion
Obtained results showed that SAR values caused by electromagnetic field exposure from RFID HF
(13.56 MHz) desktop readers may exceed general public and occupational exposure limits, during at
least 6 minute exposure, near the MR readers operating at maximum reading range of 150 cm. In
such case it is necessary to take protective measures to prevent being directly near the reader or to
reduce the duration of exposure to significantly shorter than 6 minutes. SAR values obtained for
exposure from SR readers and MR readers operating at 50 cm reading range did not exceed the
general public exposure limits.
Acknowledgments
This work was based on the results of a research task (II.N.19) carried out within the scope of the
third stage of the National Programme “Improvement of safety and working conditions” partly
supported in 2017–2019 — within the scope of research and development — by the Ministry of
Science and Higher Education/National Centre for Research and Development. The Central Institute

604
for Labour Protection – National Research Institute was the Programme’s main co-ordinator.

PB-30 [14:30]
STUDENT PAPER
Numerical analysis of a three-compartment head model subjected to variation of input
parameters
Anna Šušnjara1, Mario Cvetkovic1, Dragan Poljak1 & Hrvoje Dodig2
1Department of Electronics and Computer Science, University of Split, Split, Croatia, 21000
2Department of Naval Electronics and Information Technology, University of Split, Split, Croatia, 21000
Keywords: Dosimetry (computational), RF/Microwaves, Work in Progress
Presented by: Anna Šušnjara
A stochastic framework for the assessment of the induced electric field in the three-compartment
model of human head is presented. The relative permittivity and conductivity of scalp, skull and
brain are modelled as random variables. The statistical moments are calculated by means of a
stochastic collocation method. The sparse grid interpolation in the multidimensional random space
resulted in a small number of simulations. The sensitiy analysis of input parameters indicate the
higher overall impact of relative permittivity over the tissue conductivity on the induced field in the
brain. When considering the induced field along head axis, the results show the impact of
parameters’ variability to be distributed unevenly.
Introduction
The presence of electromagnetic (EM) fields in the environment gives a rise to constant public
concern regarding their potential harmful effects on human health. The dominant effect of the human
exposure to high frequency (HF) EM fields is tissue heating. International standards and guidelines
for HF exposure consider the human brain as one of the most sensitive organs to temperature
variations. Since the measurement in healthy humans is rather difficult, if not impossible, one has to
rely on the use of computational models and related numerical solution methods as a principal tool
for the assessment of HF exposure [1]. However, such models are extremely complex and input
variables are often uncertain, therefore using average values leads to rough approximations of the
realistic scenario [2]. The traditional approach to obtain the stochastic response of given output
values is Monte Carlo sampling which is robust but very slow. The emerging development of
stochastic methods offers different approaches to stochastic modelling. In this paper the human
head is modelled as a three-compartment model, consisting of homogeneous compartments of
brain, skull, and scalp whose electric conductivity and relative permittivity are modelled as uniformly
distributed random variables [1]. The head is exposed to HF EM field and the electric field induced
inside the brain is calculated. The uncertainties from the input variables are propagated to the output
value of interest by means of stochastic collocation (SC) method. The main advantage of the
chosen method is non-intrusive approach which means that the previously validated deterministic
model is considered as a “black box.” Hence, the method is based on sampling similar to Monte
Carlo method but the input points are chosen on the basis of a strong mathematical foundation. In
addition to the statistical moments, a sensitivity analysis is obtained by establishing the relationship
between the input and output variables. The work presented here is a continuation of some previous
efforts [3-4] undertaken for homogeneous models and serves as a basis for the stochastic analysis
of more realistic head models.
The paper is divided in two parts. Firstly, a methodology including deterministic and stochastic
approaches is outlined and then some computational results are presented. Finally some
conclusions are given.
Methodology
Deterministic Head Model
The most realistic head model routinely used in experimental magnetoencephalography is the so-

605
called three-shell or the three-compartment model, consisting of homogeneous compartments of
brain, skull, and scalp [1]. This model is freely available.
The deterministic model features the hybrid finite element/boundary element method (FEM/BEM)
useful in the assessment of the induced fields in the anatomically realistic human models [5].
Incident plane wave is treated as an unbounded scattering problem formulated via the Stratton-Chu
equation, i.e. the time-harmonic electric field at the exterior domain is expressed by the following
boundary integral equation [1]:

where Ei is the incident electric field and G is free space Green’s function.
The interior domain is governed by the vector Helmholtz equation:

with subscripts A and B denoting the exterior and interior regions, respectively.
The interior domain of the three compartment model is discretized using 1.201.246 tetrahedral
elements.
Stochastic Collocation Method
The theoretical basis of SC method is the polynomial approximation of the considered output for a
certain number of random input parameters. The interpolation of the output of interest is carried out
by solving the deterministic problem at a precomputed set of N points chosen as the sparse grid
points generated using the Smolyak algorithm [6]. For the vector of d input random parameters X =
[x1 x2 … xd] the approximation of the stochastic solution is given as:

where Y(Xi) is the solution of the deterministic model for the i-th set of input collocation points and
Li(X) is Lagrangian type basis function. The collocation points are chosen as the extreme points of
Chebyshev polynimials by using the Clenshaw-Curtis formula. These points are generated in a
nested fashion which is suitable for the Smolyak algorithm.
Following the statistical definitions, the mean and variance for univariate case are computed as
follows:

where wi is the weight for each corresponding collocation point. Weights are precomputed and
stored for later use:

The quantity ρ(X) represents the joint pdf of the random input variables X.
A more detailed description of SC method and sparse grid interpolation can be found in [6].
Sensitivity analysis approaches
Sensitivity analysis (SA) analyzes how uncertainty in the output of a model (numerical or otherwise)
can be apportioned to different sources of uncertainty in the model input [7]. In this work a Sobol-like
indices are used as SA indicators. In the first step the influence of a single input parameter’s
606
uncertainty is computed by varying one parameter at a time:

The influence of e.g. group of two parameters is calculated by varying two parameters at a time:

A more detailed description of sensitivity analysis approaches can be found in [7].

Computational results
In order to demonstrate the applicability of stochastic framework in HF dosimetry of human brain the
following test case is considered: the horizontally polarized incident EM wave of frequency 900 MHz
and amplitude 1 V/m is directed toward the anterior part of the brain/head model. The parameters of
the three-shell model are modelled as random variables uniformly distributed in the range defined as
mean value +/-20%. The mean values of parameters are given in Table 1.
Table 1. Mean value for the tissue parameters used in the three-compartment model

Tissue Tissue number Conductivity, σ (S/m) Relative permittivity,εr

Scalp 1 0.899 40.936

Brain 2 0.985 52.258

Skull 3 0.354 20.584

Given the set of uncertain input parameters X = [εr1, σ1, εr2, σ2, εr3, σ3] the effect of uncertainty in
these parameters on induced electric field in the human brain is quantified. The sparse grid
interpolation procedure is employed to approximate the quantity of interest as a function of the input
parameters. Using the Clenshaw-Curtis formula up to order of 4 for the generation of points, the
total number of simulation points is 389. A traditional Monte Carlo approach would require over a
100.000 simulations and the corresponding full tensor model 117.649 simulations.
The expectation with the variance and standard deviation is depicted on Figure 1. Maximal value of
the field expectation is 0.242 V/m with the confidence interval of +/- 0.1782 V/m. The ventral parts of
brain surface are affected more by the variation of input parameters, but the most part of the brain
surface is influenced by the input variations in uniform manner.
The impact factors are shown on Figure 2. The nature of their influence is different for the different
parts of brain surface. In general the relative permittivity of tissues has higher impact than
conductivity. The skull permittivity has the lowest impact with respect to permittivity of other tissues
while its conductivity has no significant impact. The joint influence of scalp parameters [εr1, σ1] is
comparable to joint influence of brain parameters [εr2, σ2]. Although the brain parameters have
bigger impact, there are some parts of brain surface where the effect of variation of scalp
parameters is more pronounced.
The impact factor of input variables on induced electric field along the sagittal axis of head is
depicted on Figure 3. In the interior part of brain the brain parameters have much higher impact
when comparing to impact of the other tissues’ parameters. The relative permittivity of scalp has a
somewhat noticeable influence. The difference among the impact factors decreases in the vicinity of
brain-skull interface. The Figure 3 clearly demonstrates the higher impact of electric permittivity with
respect to the conductivity of the three tissues.
Conclusion
A stochastic framework for the assessment of the induced electric field inside the human brain in

607
case of high frequency exposure is presented. The uncertainty inherent to tissue parameters is
incorporated into the three shell model of the human head. The uncertainty is propagated to the
output by means of stochastic collocation method. A sparse grid interpolation resulted in a relatively
small number of simulations needed for obtaining the stochastic response. Besides the ability to
produce the stochastic mean and standard deviation, and thus providing the confidence intervals for
the induced field, a sensitivity analysis is carried out. The sensitivity analysis clearly stresses out the
parameters for which a deeper knowledge on the corresponding values and distribution is needed in
order to build a good and reliable computational model. The presented stochastic approach will be
used in the analysis of more realistic and thus complex head/brain models. The results presented in
this paper together with the findings published in previous work [1] and [3-4] may help in deciding
the appropriate approach for the uncertainty quantification of input parameters thus lowering the
random dimension.
References
[1] M. Cvetkovic, H. Dodig, D. Poljak, “A Study on the Use of Compound and Extracted Models in
the High Frequency Electromagnetic Exposure Assessment”, Mathematical Problems in
Engineering, Vol 2017, pp 12, 2017.
[2] J. Wiart, “Radio-Frequency Human Exposure Assessment: From Deterministic to Stochastic
Methods”, John Wiley & Sons, 2016.
[3] M. Cvetkovic, S. Lallechere, K. El Khamlichi Drissi, P. Bonnet, D. Poljak, “Stochastic sensitivity in
homogeneous electromagnetic-thermal dosimetry model of human brain”, Uncertainty Modeling for
Electro-Magnetic Applications (UMEMA), Clermont-Ferrand, France, 2015.
[4] A. Susnjara, M. Cvetkovic, D. Poljak, S. Lallechere, K. El Khamlichi Drissi, “Stochastic sensitivity
in thermal dosimetry for the homogeneous human brain model,” BioEM2016, Ghent, Belgium, June
2016.
[5] D. Poljak, D. Cavka, H. Dodig, C Peratta, A. Peratta, “On the use of the boundary element
analysis in bioelectromagnetics”, Engineering analysis with boundary elements, Vol. 49, pp. 2-14,
2014.
[6] D. Xiu, J. S. Hesthaven, “High-order collocation methods for differential equations with random
inputs”, SIAM Journal on Scientific Computing, Vol. 27, No. 3, pp. 1118-1139, 2005.
[7] A. Saltelli et al., “Variance based sensitivity analysis of model output. Design and estimator for
the total sensitivity index”, Computer Physics Communications, Vol. 181, pp. 259-270, 2010.

Figures

Figure 1. The a) mean, b) standard deviation and c) variance of the induced electric field on
the surface of the brain

608
Figure 2. The impact factor of input variables on induced electric field on the surface of the
brain: figures a)-f) correspond to impact of [εr1], [εr2], [εr3], [σ1], [σ2], [σ3] respectively,
according to eq. (6); figures g)-i) correspond to impact of [εr1, σ1], [εr2, σ2] and [εr3, σ3]
respectively, according to eq. (7)

609
 Figure 3. The impact of input variables on induced electric field along the sagittal axis of the
head

PB-32 [14:30]
Quantification of thermoelastic phenomena generated from high peak power radio frequency
exposure utilizing the probe beam deflection technique (PBDT)
Ronald Barnes1, Kaitlin Nelson1, Samuel Johnson1, Jeffrey Whitmore1, Jason Payne1 & Bennett Ibey1
1711 Human Performance Wing, Human Bioeffects Division, Radio Frequency Branch, Ft. Sam Houston, TX,
USA, 78234
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Ronald Barnes
High average power (HAP) radio frequency exposures illicit a thermal response in tissue. High peak
power microwave (HPPM) exposures with pulse durations less than a few µs have been shown to
produce thermoelastic responses. In HPPM exposure conditions, the use of piezoelectric
transducers becomes difficult as the transducer is susceptible to EMI and perturbs the
electromagnetic field to be measured. An all optical sensor is desired to provide passive EMI
immune measurement of HPPM induced thermal and acoustic responses. In this work, we have
demonstrated the use of an optical dosimetry technique, namely the probe beam deflection
technique (PBDT), to measure and quantify the acoustic phenomena generated by HPPM exposure.
Introduction
High average power radio frequency exposures illicit a thermal response in tissue. In most cases
thermal response is negligible i.e. < 1˚ C. High peak power microwave (HPPM) exposures with
pulse durations less than a few µs have been shown to produce thermoelastic responses in tissue
[1,2]. In HPPM exposure conditions, the use of traditional piezoelectric transducers becomes difficult
as the transducer is susceptible to EMI and perturbs the electromagnetic field to be measured. An
all optical thermal and acoustic sensor is desired to provide passive EMI immune measurement of
high peak power electromagnetic field induced thermal and acoustic responses. In this work we
have demonstrated the use of an optical dosimetry technique, namely the probe beam deflection
technique (PBDT) [3,4], to measure and quantify the acoustic phenomena generated by HPPM

610
exposure.
Theory
Radiofrequency heating is a well understood process resuting from the combination of joule heating
and dielectric heating. HPPM sources can heat a volume of liquid very rapidly, and the thermoelastic
effect produced by rapid heating can be described by eq. 1 [5, 6].

(Eq.1)
Where k is the isothermal compressibility, β is the thermal coefficient of volume expansion, p is
pressure and t is temperature. In the case of a 1.8 GHz HPPM source the heated volume of saline
solution is expected to be approximately 3 cm in diameter. To meet the stress confinement condition
shown in eq. 2, the HPPM pulse must be less than 20 µs in duration.

(Eq. 2)
Where is the stress confinement time, d is the diameter of the volume heated and v is the speed of
sound in the media (1500 m/s). By utilizing a pulse with a duration shorter than 20 µs. The initial
pressure generated can be described by eq. 3.

(Eq. 3)
Methods
The measurement of HPPM exposure induced thermoelastic effects was accomplished utilizing a
variable frequency (1.8GHz-8GHz) HPPM transmitter with a max output power of 1.44MW and
variable pulse durations ranging from 500ns-2µs. A tank filled with 0.13 S/m saline solution was
placed 1 cm from the opening of the waveguide. The PBDT sensing system consisted of a 22 mW
Thorlabs® HeNe laser with an output wavelength of 670nm, a variable gain Thorlabs® Si
Photodiode, a pair of lenses with a 70mm focal length and a razor blade. A diagram of PBDT
acoustic measurement mechanism can be seen in fig. 1.
The PBDT system was calibrated utilizing an Onda® Hydrophone HNC-1500 with a sensitivity of
10.12 µV/Pa. An Olympus piston transducer with a 10 MHz center frequency was used as a source.
The Olympus transducer drive voltage was varied from 100V-400V in 100V increments. A PBDT
and hydrophone signal was acquired for each drive voltage. The sensitivity of the PBDT system was
calculated to be 3.37 µV/Pa. The calibration signals can be seen in fig. 2
A preliminary dataset was acquired with a transmitter frequency of 1.8 GHz. The HPPM exposure
duration was 2µs with an output power of 1.44MW. Signals were acquired with a 200MHz sample
rate from the Thorlabs® Si Photodiode with a rise time of 0.5ns. The PBDT signals were averaged
from 64 RF transmitter pulses with 2Hz repetition rate.
Results
The acoustic waves generated through thermoelastic mechanisms can be seen in fig 3a. The
deflection of the laser beam is recorded as an intensity fluctuation at the photodiode. A Fourier
transform of the PBDT signals was computed and can be seen in fig.3b.
Discussion
Thermoelastic response generated by HPPM exposure is a result of RF heating within thermal
relaxation and stress confinement temporal windows. The thermal gradient generated by HPPM
produces a pressure gradient in the fluid that propagates radially away from the thermal insult. The
acoustic frequency generated is proportional to the spatial profile of the induced heating. As the

611
volume heated increases the wavelength of the acoustic wave generated decreases. This can be
seen in photoacoustic tomography where the size of the absorber can be determined from the
wavelength of the photoacoustic wave measured [5]. The acoustic wave measured with PBDT from
fig. 3a has a fundamental frequency of 112 KHz corresponding to a wavelength in saline solution of
1.3 cm. The pressure generated was measured to be peak to peak 2.4 KPa. These measurements
validate a calculated heating diameter of 3 cm and 1.44 MW power emitted from a 1.8 GHz HPPM
transmitter.
Conclusion
The thermoelastic signals measured utilizing PBDT are in close agreement with HPPM models.
PBDT has been shown to be an EMI immune sensing methodology for real time RF thermoelastic
response quantification. The short duration pulses generated by the HPPM transmitter used in these
experiments are within the stress and thermal confinement conditions, thereby simplifying the
inverse calculations needed to predict the initial pressure generated. The fast rise time (<1ns) of the
PBDT sensor system used in this work allows for high fidelity signal processing of acoustic
phenomena resulting from RF exposure. This PBDT system can be expanded into a PBDT imaging
system as has been demonstrated in the past [4].
References
[1] R. G. Olsen and J. C. Lin, "Microwave-Induced Pressure Waves in Mammalian Brains," inIEEE
Transactions on Biomedical Engineering, vol. BME-30, no. 5, pp. 289-294, May 1983.
[2] Y. Watanabe, T. Tanaka, M. Taki and S. Watanabe, "FDTD analysis of microwave hearing effect,"
inIEEE Transactions on Microwave Theory and Techniques, vol. 48, no. 11, pp. 2126-2132, Nov
2000.
[3] R. A. Barnes, S. Maswadi, R. Glickman, and M. Shadaram, "Probe beam deflection technique as
acoustic emission directionality sensor with photoacoustic emission source," Appl. Opt. 53, 511-519
(2014).
[4] R. A. Barnes, C. C. Roth, H. T. Beier, G. Noojin, C. Valdez, J. Bixler, E. Moen, M. Shadaram, and
B. L. Ibey, "Probe beam deflection optical imaging of thermal and mechanical phenomena resulting
from nanosecond electric pulse (nsEP) exposure in-vitro," Opt. Express 25, 6621-6643 (2017).
[5] Wang LV. Tutorial on Photoacoustic Microscopy and Computed Tomography. Selected Topics in
Quantum Electronics, IEEE Journal of. 2008; 14:171-179.
[6] V Wang, Lihong. (2009). Multiscale photoacoustic microscopy and computed tomography. Nature
photonics. 3. 503-509. 10.1038/nphoton.2009.157.

Figures

612
Figure 1. PBDT acoustic measurement mechanism.

Figure 2. Signals acquired from PBDT and ONDA hydrophone for Calibration

613
 Figure 3. a. Thermoelastic acoustic wave generated from HPPM radio frequency exposure. b.
Normalized fourier transform of PBDT signal

PB-34 [14:30]
Electromagnetic field in the environment - measurements and monitoring in Poland
Pawel Bienkowski1 & Joanna Podlaska2
1Wroclaw University of Technology, Wrocław, Poland, 50-370
2Regional Inspectorate for Environmental Protection, Lodz, Poland, 90-374
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (published)
Presented by: Pawel Bienkowski
The brief overview of legal bases and EMF measurements in the environment presented in the work
shows that in Poland this issue has been known for years, research is conducted on a large scale.
The their results are widely available. The authors assail with questions about the impact of base
stations on the environment, the legal basis of measurements or even the duty to take
measurements.Analysis of the results of research presented or referred in the article indicates that
the level of electromagnetic background is well below the EMF acceptable levels in the environment.
This does not mean, that in places accessible to the public, there is no EMF intensity close to or
even exceeding the admissible values.
Intruduction
The emission of the electromagnetic field to the environment is the result of operation of any
radiocommunication systems in which this is a deliberate action. At the opposite exist power
systems, where the electromagnetic field is a side effect of electricity emission. The intensive
development of radiocommunication, in particular mobile radiocommunication, results in the building
of an increasing number of transceiver stations popularly called base stations (BS). The desire to
obtain good transmission parameters for as many users as possible leads to a high density of base
stations, but these stations can work with reduced power. In the initial period of development of
cellular systems, the priority was to cover as large area as possible hence base stations were
located relatively far away from each other and the antennas were suspended at height of 30 to 50
meters,
Fig.1. An example of a GSM base station installed on a high voltage pole

614
Over time, it has become more important than coverage to provide the largest network capacity -
that is, the ability to simultaneously serve the growing number of cutomers. At that time,
telecommunications operators began to compact the network - base stations were built at a smaller
distance from each other, new systems were introduced and this process is still ongoing. As a result,
the number of base stations increased significantly, antennas are mounted at lower altitudes due to
the specificity of modern systems and to ensure optimal network operation and internal compatibility.
The second important group are broadcasting systems - radio and television. From July 23, 2013,
when the last analog transmitter was turned off in Poland, only DVBT digital terrestrial television
working in. Radiophony is the only AM transmitter located near Solec Kujawski and broadcasting on
the long 225kHz freqecny Program of the first Polish Radio, a large network of FM transmitters and
the growing network of DAB + - terrestrial digital radio.

The first regulations regulating the permissible levels of electromagnetic fields in the environment in
Poland come from the 1980s. They were successively supplemented and ordered, and the
Regulation of the Minister of the Environment regarding the permissible levels of electromagnetic
fields in the environment and ways to check compliance with these levels , which is one of the
executive acts to the 2001 Environmental Protection Act . The Law on environmental protection also
includes other regulations regulating, among other things, the principles of qualifying, reporting or
determining the potential impact on the environment and monitoring measurements - Regulation of
the Minister of Environment on the scope and manner of conducting periodic tests of
electromagnetic fields levels in the environment. Regardless of the above legal requirements, other
measurements are carried out in the country, eg performed by Office of Electronic Communication
as part of electromagnetic spectrum monitoring or measurements carried out as part of research.
EMF measurements i enivoroment
Within the framework of the requirements specified in the Environmental Protection Law, two types
of measurements are performed - control measurements in accordance with the requirements of the
regulation and monitoring measurements, which are regulated by the secound regulation Control
measurements in accordance with the requirements of the Regulation apply to all
radiocommunication installations with an EIRP capacity above 15W and they are performed
immediately after the first start-up of the installation (BS) and after changes in installation conditions
that may affect the EMF levels that originate from this installation. It should be noted that changes in
the operating conditions of the installation may result from changes in the installation itself and its
surroundings. The aim of the measurements is to show that in places accessible to the public in the
surroundings of the installation there are no EMF levels higher than the permissible ones. Such
measurements are performed by research laboratories accredited by the Polish Center
Accreditation (PCA). The measurement methodology is based on the requirements set out in
Annexes 2 and 3 of the Regulation and from this year on the PCA DAB-18 document: Accreditation
program for testing laboratories performing electromagnetic field measurements in the environment.
Accreditation in this area has about 50 research laboratories, and such measurements are counted
in thousands during the year. The ordinance of the Minister of Environment of November 12, 2007
(Journal of Laws No. 221, item 1645), on the scope and manner of conducting periodic tests of
electromagnetic fields in the environment is the basis for monitoring measurements coordinated by
the Chief Inspectorate for Environmental Protection, and implemented by Voivodship Inspectorates
for Environmental Protection (16 accredited laboratories). Monitoring measurements are significantly
different from control measurements, the specificity consists in the method of selecting
measurement points and the measurement procedure itself. Pursuant to the Regulation, the
measurements include the intensity of the electrical component of the electromagnetic field in the
frequency range of at least 3 MHz to 3000 MHz. The measurement itself is carried out continuously
for two hours with the sampling frequency of at least one sample every 10 seconds, between 10 am
and 4 pm on working days. The result of the measurements is the arithmetic mean of the measured
effective values of the electric field intensity from the duration of the measurement - two hours. As a
result, it is an arithmetic average of 720 measurements. An example of the course of field E
variability during measurements in a small town is shown in Figure 2. The reports also include a
615
series of data, including, but not limited to, the geographical coordinates of the measurement points
and information on radio, radar, radio-navigation installations emitting electromagnetic radiation in
the frequency range from 3 MHz to 3000 MHz, located at a distance no more than 300 m from
measuring point.
Fig.2. Variability of the EMF intensity during the monitoring measurement (measurement September
7, 2017)
In turn, the measuring points are chosen so that they are not closer than 100 m from the transmit
antennas. In each province, 45 points were chosen - 15 in central districts or in settlements of cities
with a population of over 50,000, other cities and rural areas. In each of the 45 points,
measurements are made every 3 years, and the results are available on the websites. Due to the
method of selection of measuring points - monitoring measurements can be treated as
measurements of the so-called "Electromagnetic background", and their regular repetition allows
you to analyze the trends of change. An example of this is the analysis of monitoring measurements
in Lodz. On the map from Fig. 3, measurement points are marked out in three subsequent years -
they form a complete set of points, in which measurements are repeated every 3 years.
Fig.3 . Map with the location of EMF monitoring points in Lodz
In the chart from Figure 4, the results from the same measurement points were collected every 3
years - 2009 - 2012 -2015.
Fig 4 The results of monitoring measurements in Lodz, 2009-2012-2015
The analysis of the results of the statement from Figure 4 allows the assessment of changes in field
strength for 15 measuring points over a period of 9 years. At some points, a gradual increase in the
field intensity can be seen - it can be assumed that new base stations have been created or existing
ones have been extended.
Pursuant to the Regulation on Monitoring, the Chief Inspectorate for Environmental Protection also
regularly organizes field intensity measurements in the centers of large cities in Poland. Such
measurements are divided into frequency sub-bands at selected city points. Recent such studies
were carried out in 2014-2015 in Warsaw, Łódź, Kraków, Szczecin, Poznań and Gdańsk. A detailed
report on the results is available on the CIEP's website It should also be mentioned that the
monitoring of the electromagnetic spectrum also leads, within the framework of the frequency
management, to the Electronic Control Office. It is true that these measurements are not directly
aimed at measuring the intensity of the electromagnetic field - more important issues are the
"correctness" of the radio signal, but there is also the possibility of analyzing even the trends of field
intensity changes over time. The latest initiative related to the measurement of electromagnetic field
intensity in the environment is the program of the Ministry of Digitization, under which the Institute of
Communications conducts analysis of documentation and performs specified EMF measurements
with spectrum analysis and the use of specialized software to assess the maximum field intenisty
regardless of the momentary load of base stations with telecommunications traffic . The results of
the pilot work were included in the "Report from pilot studies and analyzes regarding permissible
electromagnetic field levels" dated March 2017 and available on the website. From a technical point
of view, the choice of the measurement method or the time of EMF measurement in the environment
is very important for the majority of currently operating radiocommunication systems. This is due to
the previously mentioned variability of field strength in time associated with the change power
transmitted by base stations, which are located relatively close to the places available to the public.
It is not always possible to set the base stations to the service mode with the highest power,
especially when there are fields from many stations of different operators at the measurement
location, and it is always possible to reconcile measurements with all operators. However, expect
some statistical dependencies describing the variability of telecommunications traffic or emitted
station power. Such studies have been conducted several years ago, among others in order to
determine the so-called "Conditions of normal operation", determining when they are present and
how it is possible to take such measurements to the mode of operation with the highest power
attained by the stations during typical operation. With the emergence of new systems (eg UMTS or

616
LTE), such tests have been repeated for many years and it can be stated that statistically regardless
of the system or location of the base station (excluding special locations - eg stations near stadiums,
marketplaces, etc. , where the variability of movement is directly related to the activity of such
places), the field intensity variation between the highest and the smallest values does not exceed
two times, and if it compares the variability over time of the most active station activity, these
differences do not exceed 20-30%. This information is important because in the case of
measurements of normally operating systems at any time, if the measurement result does not
exceed the half of the acceptable value, it can be assumed that this value will not be exceeded
regardless of the system's operating conditions. If such measurements are made during the
statistically most extensive activity, even if the measured value reaches 65-70% of the admissible
value - it can be assumed that in this case, the limit values will not be exceeded (in the environment
for frequencies above 3MHz, the limit value is 7V/m ). In the publications representative results of
long-term measurements were presented, and to confirm the above thesis in Figure 5 presents the
results of three-day monitoring of a multi-system suburban base station (LTE800, GSM / UMTS 900,
GSM / UMTS 1800, LTE 2600).
Figure 5. Monitoring results of field intensity variation from cellular telephony base station
(measurements September 2017)
Figures

Figure 1. Fig.3 . Map with the location of EMF monitoring points in Lodz

617
 Figure 2. Fig 4 The results of monitoring measurements in Lodz, 2009-2012-2015

Figure 3. Figure 5. Monitoring results of field intensity variation from cellular telephony base
station (measurements September 2017)

PB-36 [14:30]
Comparison between numerically and experimentally assessed skin temperature elevations
for localized RF exposure at frequencies above 6 GHz
Davide Colombi1, Bo Xu1, Christer Tornevik1, Björn Thors1, Andreas Christ2, Marvin Ziskin3, Kenneth Foster4
& Quirino Balzano5
1Ericsson Research, Ericsson , Stockholm, Sweden, 11222
2Research Consultant, Cabo Frio, Brazil
3Center for Biomedical Physics, Temple University Medical School, Philadelphia, PA, USA
4University of Pennsylvania, College Park, PA, USA
5University of Maryland, College Park, MD, USA

618
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Davide Colombi
Frequency bands above 10 GHz are key components of the next generation mobile networks, 5G,
which are expected to be commercially available before 2020. Such frequencies have been the
target of the ongoing revision of the IEEE C95.1 exposure standard. In this study the draft exposure
limits are evaluated in terms of skin temperature elevation both numerically and experimentally. The
results indicate that simple thermal models can be successfully used to evaluate temperature
elevation in the investigated frequency range although they might overestimate the skin temperature
value for longer exposure time.
Introduction
Frequency bands above 10 GHz, are key components of the next generation mobile networks, 5G
[1], which are expected to be commercially available before 2020. For this reason, the development
of thermal models useful to predict skin temperature increase due to exposure in the millimeter
wave frequency range has been the subject of some recent studies [2]-[7]. On the other hand, when
compared to frequency bands below 10 GHz, i.e. the spectrum where the vast majority of current
radio communication sources operate, the availability of experimental data addressing skin
temperature increase is more limited [8]-[9].
In this study, skin temperature elevation due to localized RF exposure when predicted by the
bioheat thermal equation is compared with measurements conducted on three persons belonging to
the EMF Lab staff at Ericsson Research, by means of an infrared camera.
Method
Numerical simulations based on the bioheat thermal equation were conducted using CST
Multiphysics Studio (CST AG, Darmstadt, Germany). The skin model was based on the flat
multilayer structure described in [6] consisting of a skin layer, a fat layer, and a muscle layer. The
dielectric properties were those provided in [10]. The temperatures of the room, deep muscle, and
the blood were set to 20 °C, 37 °C, and 37 °C, respectively, and the convective heat transfer
coefficient was equal to 10 W/m2/°C.
Thermographic measurements were conducted by means of a Flir A6700sc infrared (IR) camera.
This camera makes use of a cryo-cooled (about 73 K) infrared detector to allow for high-sensitivity
measurements (< 20 mK). Three persons belonging to the EMF Lab staff at Ericsson Research
were exposed by means of horn antennas (PE9850/2F-10 and PE9854/SF-10,Pasternack, Irvine,
USA) at 15 GHz, 28 GHz and 39 GHz (of which the last two frequencies have been selected as a
candidate for 5G [11]). The antenna was placed so that the horn aperture would be at 5 mm and 15
mm from the skin surface. The transmit power was set to comply with the applicable exposure limits
for workers in Sweden (where the study was conducted) and EU [12] which are based on the
guidelines provided by ICNIRP [13].
The measured and simulated temperature increases were scaled to the general public exposure
reference levels (ERL) recently proposed in the draft revision of IEEE C95.1 (currently in a balloting
phase). Such values are provided in terms of incident power density (IPD) and are reported in
Figure 1.
In order to be able to scale the measured temperature elevation for a specific ERL value, near-field
measurements of the power density for the horn antennas were conducted by means of a
miniaturized mmW probe (Schmid and Partner Engineering AG, Zurich, Switzerland). Power density
was measured in free-space according to the procedure specified by IEC 63170 [14].
Results
A sample of the evaluated configurations is provided in the figures below. Figure 2 shows the
maximum skin temperature increase as a function of time (up to 400 seconds) for the forearm
exposed by means of a horn antenna at 28 GHz and for a separation distance of 5 mm. The
transmit power is chosen to give an IPD level as specified in Figure 1. The corresponding values at

619
39 GHz is provided in Figure 3. The green curve was determined by means of numerical simulations
while the others correspond to the measurement data for the three investigated subjects.
The comparison between experimental and numerical data for tissue elevations of other tissue
areas (like the forehead and the trunk) will be presented during the BioEM 2018 conference.
Conclusions
A comparison of the skin temperature increase estimated by the bioheat thermal equation with
experimental data obtained using an infrared camera indicates that thermal models can be
successfully used to predict skin temperature elevation due to RF energy absorption.
The study indicated that, for the frequencies investigated (15 GHz, 28 GHz and 39 GHz) the
localized peak temperature elevation corresponding to the exposure reference levels (ERL) recently
proposed in the draft revision of IEEE C95.1 is less than 1 °C. This is well within the normal variation
of surface skin temperature when not exposed to any RF source, which for the forearm was found to
be between 4 °C to 5 °C.
Additional numerical evaluations are ongoing in order to quantify the impact of the layering structure
of skin tissue and the uncertainty of the thermal tissue parameters. These studies are carried out
using the Pennes solver of Sim4Life 3.4 (ZMT, Zurich, Switzerland).
References
[1] E. Dahlman, G. Mildh, S. Parkvall, J. Peisa, J. Sachs, and Y. Selén, ‘‘5G radio access,’’ Ericsson
Rev., vol. 6, pp. 2–7, 2014.
[2] K. R. Foster, M. C. Ziskin, and Q. Balzano, ‘‘Thermal response of human skin to microwave
energy: A critical review,’’ Health Phys., vol. 111, no. 6, pp. 528–541, 2016.
[3] K. R. Foster, M. C. Ziskin, and Q. Balzano, ‘‘Thermal modeling for the next generation of
radiofrequency exposure limits: Commentary,’’ Health Phys., vol. 113, no. 1, pp. 41–53, 2017.
[4] Y. Hashimoto et al., ‘‘On the averaging area for incident power density for human exposure limits
at frequencies over 6 GHz,’’ Phys. Med. Biol., vol. 62, no. 8, pp. 3124–3138, 2017.
[5] K. R. Foster and D. Colombi, ‘‘Thermal response of tissue to RF exposure from canonical dipoles
at frequencies for future mobile communication systems,’’ Electron. Lett., vol. 53, no. 5, pp.
360–362, 2017
[6] K. Sasaki, M. Mizuno, K. Wake and S. Watanabe, “Monte Carlo simulations of skin exposure to
electromagnetic field from 10 GHz to 1 THz”, Phys. Med. Biol., vol. 62, pp. 6993–7010, Aug. 2017.
[7] M. Ziskin, S. Alekseev, K. R. Foster and Q. Balzano, ‘‘Tissue models for RF exposure evaluation
at frequencies above 6 GHz,” Bioelectromagnetics, 2018.
[8] S. Alekseev, A. Radzievsky, I. Szabo and M. Ziskin, “Local heating of human skin by millimeter
waves: effect of blood flow,” Bioelectromagnetics, vol. 26, no. 6, pp. 489–50, 2015.
[9] S. Alekseev and M. Ziskin, “Local heating of human skin by millimeter waves: a kinetics study,”
Bioelectromagnetics, vol. 24, n0. 8, pp 571-8, 2003.
[10] D. Andreuccetti, R. Fossi and C. Petrucci, “An Internet resource for the calculation of the
dielectric properties of body tissues in the frequency range 10 Hz - 100 GHz”, IFAC- CNR, Florence
(Italy), 1997. Based on data published by C. Gabriel et al. in 1996. [Online]. Available:
http://niremf.ifac.cnr.it/tissprop/.
[11] Federal Communications Commission, “FCC 16-89 report and order and further notice of
proposed rulemaking”, July 14, 2016
[12] Directive 2013/35/EU of the European Parliament and of the Council of 26 June 2013 on the
minimum health and safety requirements regarding the exposure of workers to the risks arising from
physical agents (electromagnetic fields)
[13] The International Commission on Non-Ionizing Radiation Protection, ‘‘Guidelines for limiting
exposure to time-varying electric, magnetic, and electromagnetic fields (up to 300 GHz),’’ Health

620
Phys., vol. 74, no. 4, pp. 494–522, Apr. 1998.
[14] International Electrotechnical Commission, “TR 63170 ED1, measurement procedure for the
evaluation of power density related to human exposure to radio frequency fields from wireless
communication devices operating between 6 GHz and 100 GHz”, Jan, 2018.
Figures

Figure 1. Incident power density (IPD) exposure reference levels (ERL) for the general public
as specified in the IEEE C95.1 draft (yet to be approved). f (> 6 GHz) is the frequency in GHz.
The ERL is intended to be averaged over an area of 4 cm2.

Figure 2. Comparison of the skin temperature increase as a function of time (up to 400
seconds) obtained with measurements and simulations. The forearm was exposed by means
of a horn antenna at 28 GHz and for a separation distance of 5 mm. The transmit power is
chosen to give an IPD level as specified in Figure 1.

621
 Figure 3. Comparison of the skin temperature increase as a function of time (up to 400
seconds) obtained with measurements and simulations. The forearm was exposed by means
of a horn antenna at 39 GHz and for a separation distance of 5 mm. The transmit power is
chosen to give an IPD level as specified in Figure 1.

PB-38 [14:30]
EMPIR European project Vector SAR on SAR measurement using vector probes
Allal Djamel1 & Joe Wiart2
1Laboratoire National de Métrologie et d’Essais, Trappes, France, 78197
2Chaire C2M, Telecom Paristech, Paris, France, 75013
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Joe Wiart
Specific absorption rate (SAR) must be evaluated during the production of mobile
telecommunication devices. Vector SAR project will provide the methods, software tools and
datasets required for traceable calibration and uncertainty analysis of vector probe array systems
that automatically determine the 3D electromagnetic field and SAR mapping using amplitude and
phase information through a 3D reconstruction algorithm. This work will also contribute to the
international standard IEC 62209-3 and future standardisation of fifth generation (5G) devices within
IEC Technical Committee TC 106.
Introduction
In the context of an ever-increasing development of mobile phones and the number of
telecommunication protocols that need to be tested to assess SAR during the production of such
smartphones, the methods included in the international standards IEEE 1528, IEC 62209-1 and IEC
62209-2 now require excessively long testing times to assess compliance with SAR restrictions,
using the diode probe and robot as specified in these standards. Furthermore, new communications
standards will incorporate complex multiple-input multiple-output (MIMO) antennas that cannot be
efficiently assessed using the systems specified in current published standards, as they do not
measure phase. New SAR measurement systems which use arrays of vector probes [1] allow for

622
drastically reduced time for testing but they need methods for traceable calibration and well
quantified uncertainty estimates before they can be adopted into international standards.
Objectives
The aim of Vector SAR project is to provide the essential methods, software, data and validation
required for the successful completion of the international standard IEC 62209-3 related to the
measurement of SAR from handheld wireless telecommunications devices using vector based
systems. Specifically, traceable methods are being developed for the calibration of vector probes
and probe arrays up to 6 GHz. Methods for uncertainty propagation through multivariate models will
be established using the principles given in the ‘Guide to the expression of uncertainty in
measurement’ (GUM) [2]. The reliability of measurement systems for a wide range of transmitter
types will be verified and the measurement of telecommunication signals and SAR for a wide range
of device types will be improved. Test protocols for MIMO devices using vector probe arrays will be
developed to determine the maximum SAR value (worst case) by combining MIMO signal figures.
This will in whole contribute to the standards development work of the technical committee IEC - TC
106 on the successful adoption of IEC 62209-3 standard vector-based SAR measurement systems.
Conclusion
Vector SAR project results will enable manufacturers of vector-based SAR measurement equipment
to have better quality control of their products and will provide greater confidence in their
measurement accuracy. A wide impact to industry will be achieved by incorporation of the vector
array systems into IEC standards, which will enable the use of these systems by testing laboratories
for testing full compliance of mobile telecoms devices against exposure limits.
References
[1] Cozza A., Merckel O., and Bolomey J.-Ch. “A new probe-array approach for fast SAR
measurements,” International Workshop on Antenna Technology 2007 (iWAT07), Cambridge, March
2007.
[2] BIPM, IEC, IFCC, ISO, IUPAC, IUPAP, and OIML, Evaluation of measurement data Supplement
2 to the ‘Guide to the expression of uncertainty in measurement’ Extension to any number of output
quantities JCGM 102:2011, 1st ed. Paris, Sèvres: BIPM Joint Committee for Guides in Metrology,
2011.

PB-40 [14:30]
WITHDRAWN

PB-42 [14:30]
Attempt for determination the power absorption of nanomagnetic materials by reflectometry
Ivan Gresits1, 2, Ferenc Simon2 & György Thuroczy1
1Department of Non-Ionizing Radiation, National Public Health Institute, Budapest, Hungary, 1221
2Department of Physics, Budapest University of Technology and Economics, Budapest, Hungary, 1111
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: György Thuroczy
Nanomagnetic hyperthermia is intensively studied with the prospect of cancer therapy. A major
challenge is to determine the heat absorbed during in vivo conditions and conventional methods are
either invasive or inaccurate. We present a method which yields directly the heat absorbed during
hyperthermia: it is based on accurately measuring the quality factor change of a resonant
radiofrequency (RF) circuit which is employed for the irradiation.
Methods

623
The present method is based on the monitoring of the quality factor of resonant coils filled with
magnetic nanoparticles. We studied two kinds of resonant circuits: a conventional tank circuit which
consists of a solenoid for the test-tube studies and a so-called birdcage coil which is employed in
magnetic resonance imaging. Quality factors of resonant circuits were determined from
reflectometry. The method was validated by a conventional calorimetric measurement in a test tube
with high RF input power.
Introduction
Cancer is one of the major diseases worldwide. Despite the recent technical advancements in
cancer treatment such hyperthermia, the need for better efficient method for cancer therapy
remains. One of the recent methods in the cancer treatment is called nanomagnetic (nanofluid)
hyperthermia (NMH). The NMH is intensively studied in the last decade as a new prospect of cancer
therapy. It involves raising the temperature locally in the tumour with the intact of normal tissues. A
major challenge is to determine the heat absorbed during in vivo conditions and conventional
methods are either invasive or inaccurate. Herein, we present a method which allows determining
the power absorbed during nanomagnetic hyperthermia. Furthermore, our method needs low RF
power with less robust coils via reflection measurements. We validated this approach by a
comparison of the absorbed power with the theoretically expected values obtained from the result
with well-known calorimetric approach. This method allows determining the specific absorption rate
(SAR) for commercial magnetite samples.
Methods
The present method is based on the monitoring of the quality factor (Q) of resonant coils in empty
case and even if it is filled with nanomagnetic material. The absorbed power can be obtained from
these quality factors using Eq. (1).

(1)
We studied two kinds of resonant circuits: a conventional tank circuit which consists of a solenoid for
the test-tube studies and a so-called birdcage coil which is employed in magnetic resonance
imaging (Fig. 1. shows its block diagram). Quality factors of resonant circuits were determined from
reflectometry. We validate the method was validated by comparing it with a conventional calorimetric
measurement in for a test tube with high RF input power.
Results
Fig. 2. shows the reflection curves of the solenoid. The more magnetite is contained by the sample
the wider is the curve and the lower is the resonance frequency.
Fig. 3. shows the rise time of temperature in the sample warming multiplied by unit mass,

as a function of the absorbed power, Pabsorbed. The earlier quantity is obtained by monitoring the
sample temperature whereas the absorbed power is obtained using Eq. (1) from the measurement
of the resonator quality factors. In principle, the two quantities are related by:

(2)
where csample denotes the specific heat of the sample. The solid line in Fig. 3. is calculated
assuming

References
1. J. Beik, Z. Abed, F. S. Ghoreishi, S. Hosseini-Nami, S. Mehrzadi, A. Shakeri-Zadeh, and S. K.

624
 Kamrava, Journal of Controlled Release235, 205 (2016), ISSN 0168-3659
2. C. S. Kumar and F. Mohammad, Advanced Drug Delivery Reviews, 63, 789 (2011), ISSN
0169-409X
3. S. Y. Wang, S. Huang, and D. A. Borca-Tasciuc, IEEE Transactions on Magnetics 49, 255
(2013), ISSN 0018-9464.B.
4. Gyüre, B. G. Márkus, B. Bernáth, F. Murányi, and F. Simon, Review of Scientific Instruments 86,
094702 (2015)
Figures

Figure 1. The measurement setup of the tank circuit (a), and the birdcage coil (b). The hybrid
junction divides the incoming RF power into 2 equal parts. One half absorbes on the 50 Ω
resistance, the other half goes to the resonant circuit. By screwing the tuning (CT) and
matching (CM) capacitors one can minimize the reflection on a certain frequency. The reflected
power is also divided into 2 halves and the spectrum analyzer measures one of them. The
birdcage coil has fixed value chip capacitors with Q>1000.

625
Figure 2. The reflection curves of the unloaded (empty) and the filled circuits with different
diluted samples in log scale (upper panel) and linear one (lower panel). Critical coupling was
achieved for all measurements.

Figure 3. The rise time of warming per unit mass as determined from calorimetry for various
sample concentrations as a function of the absorbed power. Solid curve is a calculation using
the specific heat of water, i.e. the line is not fitted to the data. Note the excellent agreement
between the data points of measurements and calculation.

626
PB-44 [14:30]
Improving the accuracy of personal radiofrequency measurements using a novel body-worn
measurement device and comparison with measurements using exposimeters
Anke Huss1, Stefan Dongus2, Reza Aminzadeh3, 4, Arno Thielens3, 5, Sam Agneessens3, 4, Patrick Van
Torre3, 4, Rene De Seze6, Elisabeth Cardis7, Marloes Eeftens2, Wout Joseph3, 4, Roel Vermeulen1 & Martin
Röösli2
1University of Utrecht, Utrecht, the Netherlands
2University of Basel, Basel, Switzerland
3University of Ghent, Ghent, Belgium
4IMEC, Ghent, Belgium
5Berkeley Wireless Research Center, Berkeley, CA, USA
6National Institute for Industrial Environment and Risks, Verneuil-en-Halatte, France
7Barcelona Institute for Global Health, Barcelona, Spain
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Anke Huss
The aim of this study is to compare measurement results of a newly developed prototype of a multi-
band body-worn distributed-exposimeter (BWDM) with two commercially available personal
exposimeters (EXPOM and EME SPY 200).We are currently conducting field measurements in
various indoor and outdoor microenvironments in Belgium, Spain, France, Netherlands and
Switzerland.Our data will allow the comparison of the measurements of the three different
exposimeters and will be informativebetter understanding and interpretation of existing
epidemiological research results.
Human exposure to radiofrequency electromagnetic fields (RF-EMF) is often measured by personal
exposimeters. However, accuracy of measurements using the available portable devices is reduced
due to the presence of the human body and due to large measurement uncertainty. The aim of this
study is to compare measurement results of a newly developed prototype of a multi-band body-worn
distributed-exposimeter (BWDM) with two commercially available personal exposimeters (EXPOM
and EME SPY 200). The BWDM prototype has been developed for simultaneous measurements of
the incident power density in 11 frequency bands (LTE 800 and 2600 MHz, 900 MHz, 1800 MHz,
2100 MHz, DECT, Wi-Fi 2 GHz and 5 GHz, including uplink and downlink bands). The BWDM
consists of 22 separate antennas integrated in a garment (a vest), distributed in an optimal way on
the front and back of the human torso as well as right and left hips. For all frequency bands, antenna
pairs are placed on diametrically opposite locations on body, with the aim of minimizing body-
shielding. The BWDM was calibrated on the body under laboratory conditions for different body
sizes and shapes. We are currently conducting field measurements in various indoor and outdoor
microenvironments in Belgium, Spain, France, Netherlands and Switzerland. In each country, a
trained research assistant is using the BWDM in parallel with EXPOM and EME SPY 200
exposimeters by walking along pre-defined measurement routes, comparing different characteristic
microenvironments such as urban, suburban and rural areas, public transport infrastructure, and
public areas such as universities, parks and shopping centres. Our data will allow the comparison of
the measurements of the three different exposimeters. Results will be presented at the conference.
The results of the device comparisons will enable a better understanding and interpretation of
existing epidemiological research results, as well as improved risk assessment and communication
strategies.

PB-46 [14:30]
A comparison of the RF-EMF in educational facilities and its surrounding areas in South
Korea
Taewook Hwang1, Geojung Kim1 & Kyuhyeon Kim1

627
1RadioWave Environment Team, Korea Communications Agency, Naju, Korea, 58324
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Taewook Hwang
This paper is described the RF-EMF levels in/around kindergartens and elementary schools.The
RF-EMF of kindergarten and elementary school was average 0.09 V/m, which was much lower than
the ICNIRP reference levels in South Korea. In addition, the RF-EMF of roads approaching these
facilities was average 0.19 V/m, higher than the RF-EMF of kindergartens and elementary schools,
but below the ICNRIP guidelines. This result is expected to be due to the high density of mobile
communication base stations in South Korea.This paper is considered very important with its high
reliability to use in epidemiological researches and prediction model for EMF exposure including
researches on EMF health effects.
Introduction
As the number of base stations for mobile communication services rapidly grows with development
of mobile communications and increase of radio-use devices, there are vague concerns raised on
the possible health effects of RF-EMF radiated from base station antennas for mobile
communications services. This paper is described the RF-EMF levels in/around kindergartens and
elementary schools. This paper is considered very important with its high reliability to use in
epidemiological researches and prediction model for EMF exposure including researches on EMF
health effects.
Material and Method
In this paper, we were measured the RF-EMF level in/around 353 education facilities, 282
kindergartens and 71 elementary schools, in Republic of Korea. We have chosen to measure points
around classrooms, hallway and playground for educational facilities and access roads for schooling
within the radius of 50m about kindergarten and elementary school. The measurement frequency
bands are the mobile communications services, the base stations for CDMA, WCDMA, LTE and wifi
band in South Korea. We were used the EMF measurement systems, Narda SRM-3000 and self-
developed vehicle EMF measurement system, included 3-axis electric filed antenna. The Figure 1
shows the measurement outline of this paper.
Results
The RF-EMF of classroom and hallway were similar, and the RF-EMF of playground was higher
than the classroom/hallway except the wifi band. In the playground, the average RF-EMF of
WCDMA was the highest at 0.09 V/m, but was 0.15% lower than the reference levels for general
public exposure of ICNIRP. On the other hand, the EMF of wifi band in the classroom/corridor was
higher than in the playground, and the mean EMF was average 0.08 V/m, which was 0.12% of the
ICNIRP reference levels. The reason for this result is that most of the wifi devices are installed
indoors, so the wifi band is estimated to be high in the classrooms/hallways. The Table 1 shows the
RF-EMF results of mobile communication base stations and wifi band measured at kindergartens
and elementary schools.
We compared the RF-EMF of the surrounding roads within a radius of 50m to the education facilities
to see how high the RF-EMF level of the kindergarten and elementary school. The Table 2 and
Figure 2 show these results. The RF-EMF of the surrounding roads in three mobile communication
bands, CDMA, WCDAM and LTE, was average 0.19 V/m higher than the RF-EMF of kindergarten
and elementary school, and the wifi band was the opposite. The LTE band was the highest at 0.23
V/m in the surrounding roads of the educational facilities and 0.4% of the ICNIRP reference levels.
Conclusions
The RF-EMF of kindergarten and elementary school was average 0.09 V/m, which was much lower
than the ICNIRP reference levels in South Korea. In addition, the RF-EMF of roads approaching
these facilities was average 0.19 V/m, higher than the RF-EMF of kindergartens and elementary
schools, but below the ICNRIP guidelines. This result is expected to be due to the high density of

628
mobile communication base stations in South Korea. Nevertheless, it is important to note that the
RF-EMF of mobile communications band is lower than the ICNIRP reference levels.
References
[1] ICNIRP, “Guidelines for Limiting Exposure to Time-Varying Electric, Magnetic, and
Electromagnetic Fields (Up to 300 GHz)”, 1998
[2] TTA Standard TTAK.KO-06.0391, “Measurement Method of Human Exposure Levels for EMF
Vulnerable Area”, Dec. 2015.
[3] European standard EN50400, “Basic standard to demonstrate the compliance of fixed equipment
for radio transmission intended for use in wireless telecommunication networks with the basic
restrictions or the reference levels related to radio frequency electromagnetic fields, when put into
service”, June 2006.
Table 1. The RF-EMF in Educational Facilities

ICNIRP Electric Field Strength [V/m]

Frequency Band Reference
Level [V/m] Classroom Hallway Playground

869 ~ 874
CDMA(BS) 40.5 0.0259 0.0221 0.0632
MHz

1.8㎓, 2.1
WCDMA(BS) 59.3 ~ 61 0.0326 0.0295 0.0897
GHz

800 ~ 900
MHz
LTE(BS) 40.6 ~ 61 0.0334 0.0287 0.0800
1.8/2.1/2.6
GHz

Wifi 2.4 GHz 61 0.0607 0.0777 0.037

Table 2. The Comparison of the RF-EMF in Educational Facilities and Its Surroundings

Electric Field Strength [V/m]

ICNIRP
Frequency Band Reference Level
[V/m] Educational
Around Roads
Facilities

CDMA(BS) 869 ~ 874 MHz 40.5 0.0371 0.1160

WCDMA(BS) 1.8㎓, 2.1 GHz 59.3 ~ 61 0.0506 0.1975

800 ~ 900 MHz

LTE(BS) 40.6 ~ 61 0.0474 0.2399
1.8/2.1/2.6 GHz

Wifi 2.4 GHz 61 0.0585 0.0220

629
Figures

Figure 1. The RF-EMF Measurement Concepts in/around Educational Facilities

Figure 2. The RF-EMF in Educational Facilities

630
 Figure 3. The Comparison of the RF-EMF in Educational Facilities and Its Surrounding Areas

PB-48 [14:30]
STUDENT PAPER
Investigation of E-field strength from mobile phone base stations and transmitted power
from 4th generation mobile phones
Shota Kurosaki1, Masaki Hagiwara1, Masao Taki1, Atsuko Aimoto1, Miwa Ikuyo1, Kaoru Esaki1 & Kanako
Wake2
1Department of Electrical and Engineering, 1 - 1 Minami-osawa, Hachioji-shi, Japan, 192-0397
2Electromagnetic Compatibility Laboratory, National Institute of Information and Communications Technology,
Tokyo, Japan, Japan, 184-8795
Keywords: Dosimetry (measurements), RF/Microwaves, Review, Commentary, Recommendation,
Evaluation
Presented by: Shota Kurosaki
We measured 4G Tx power and the total downlink E-field strengths during drive around Tokyo and
Chiba area in Japan for about 8 hours. The association between 4G Tx power and the downlink
E-field strengths was investigated. The total downlink E-field strength was high and 4G Tx power is
low in the Center of the Capital while the total downlink E-field strength was low and the 4G Tx
power is high in residential/rural area.
Introduction
The number of mobile phones is increasing rapidly. Exposure from mobile phones has been a
matter of concern [1]. In previous studies, it is reported that transmitted power from 3rd generation
mobile phones (3G Tx power) [2] is much lower than 2nd generation (2G) Tx power [3]. Recently 4G
Tx power has been reported [4]. It is not reported, however, how the 4G Tx power is associated with
the E-field strengths from mobile phone base stations (i.e. downlink E-field strength). The objective
of this study is to investigate association between 4G Tx power and the downlink E-field strengths.

631
In this study, we measured 4G Tx power and the total downlink E-field strengths during drive around
Tokyo and Chiba area in Japan for about 8 hours.
The association between 4G Tx power and the downlink E-field strengths was investigated. We also
investigated the association between 4G Tx power and the downlink E-field strengths at fixed
locations.
Method
Measurement during drive
We drove around in Tokyo and Chiba, Japan for about 8 hours. We measured 4G Tx power and the
downlink E-field strengths during the drive. A smartphones application “Sigma-ML” (Meritech Co.,
Ltd., Japan) was used to record instantaneous Tx power from the 4G smartphone. The
instantaneous Tx power was averaged over every 1 second as the root mean square value. The
E-field strengths in the environment were measured by a three axes E-field probe “3502/01” with a
handheld spectrum analyzer “SRM-3006” (Narda S.T.S. GmbH Germany). The E-field probe was
located near a window of the car. SRM-3006 measured E-field strengths for the frequency band
from700 MHz to 2200 MHz. Resolution band width (RBW) of SRM-3006 was defined as 500
kHz.The frequency spectrum of E-field strengths was recorded every 6 seconds by SRM-3006.We
calculated the total downlink E-field strength for 4 downlink frequency bands for the relevant mobile
phone operator by integrating spectrums of downlink E-field strengths for each band. Table 1 shows
those frequency bands. We prepared two smartphones and they were in a state of talking with each
other.
Table 1: 4 downlink frequency bands for the relevant mobile phone operator

Band Name Frequency range

800 MHz 875 MHz – 890 MHz

1400 MHz 1495.9 MHz – 1510.9 MHz

1700 MHz 1859.9 MHz – 1879.9 MHz

2 GHz 2130 MHz – 2150 MHz

Measurement at fixed locations
We prepared personal exposure meter “ExpoM-RF” (Fields at Work GmbH Zurich) to measure the
total downlink E-field strength in fixed locations in Hakone (rural) and in Tokyo (Center of the Capital
region). We set up ExpoM-RF and a Sigma-ML installed smartphone in a car or in a building. The
car was parked in an outdoor parking lot during the measurement. ExpoM-RF measured E-field
strengths of 16 frequency bands between 87.5 MHz and 5,875 MHz. The 16 bands of our ExpoM-
RF were not the original ones but adjusted to Japanese frequency use. The frequencies for the 4
bands used are listed in Table 2. The downlink E-field strengths were recorded every 3 second by
ExpoM-RF.
Table 2: The frequencies for the 4 bands used
Band Name Frequency range

800 MHz 860 MHz – 890 MHz

1400 MHz 1475.9 MHz – 1510.9 MHz

1700 MHz 1844.9 MHz – 1879.9 MHz

2 GHz 2110 MHz – 2170 MHz

632
Result and Discussion
Measurement during drive
Figure 1 shows the histograms of the 4G Tx power (a) and the total downlink E-field strengths (b) of
800 MHz band and those of the 4G Tx power (c) and the total downlink E-field strengths (d) of 2
GHz band during the whole drive including residential/rural and Center of the Capital region. The
statistics are shown in Table 3.
Table 3: The statistics of (a) , (b) , (c) and (d)
90 %ile Average Median 10 %ile

(a) 12.7 dBm 13.2 dBm 2.2 dBm -10.4 dBm

(b) 106.3 dBuV/m 105.3 dBuV/m 104.8 dBuV/m 103.9 dBuV/m

(c) 15.5 dBm 14.6 dBm 4.5 dBm -6.1 dBm

(d) 103.8 dBuV/m 102.6 dBuV/m 99.8 dBuV/m 98.7 dBuV/m

It should be noted that the 4G Tx power is the peak value of transmitted waveform during call but
without speech. The real r.m.s. value is estimated in consideration of duration factor, which is the
temporal factor of the duration of packet transmission. The duration factor is roughly estimated 1 – 5
% for 4G-LTE systems according to our preliminary measurement. It should also be noted that the
Tx power increases several times larger when the speech exists than without speech.
The 4G Tx power and the total downlink E-field strengths of 800 MHz and 2 GHz band were plotted
on the route of drive in Google Earth (Figure 2). The total downlink E-field strength was high and 4G
Tx power is low in the Center of the Capital while the total downlink E-field strength was low and the
4G Tx power is high in residential/rural area.
Measurement at fixed locations
Figure 3 shows 4G Tx power and the total downlink E-field strengths of 800 MHz in the fixed
locations in the Center of the Capital region and a residential/rural area.The total downlink E-field
strength in the Center of the Capital (average 110 dBuV/m, median 109 dBuV/m) is higher than that
in residential/rural area (average 79 dBuV/m ,median 78 dBuV/m). The 4G Tx power in the Center
of the Capital (average 0.3 dBm, median -6.3 dBm) is lower than that in the residential/rural area
(average 6.6 dBm, median 6 dBm). The results confirm the observation from the measurement
during drive.
Conclusion
We measured 4G Tx power and the total downlink E-field strengths. The measured data suggested
the association between 4G Tx power and downlink E-field strength as follows.
In residential/rural area, the total downlink E-field strength is low while the 4G Tx power is high. In
contrast the total downlink E-field strength in the Center of the Capital is high while the 4G Tx power
is low.
Reference
[1] IARC, ”IARC classifies radiofrequency electromagnetic fields as possibly carcinogenic to
humans, ” press release, No.208, 31 May ,2011.
[2] A. Gati, A. Hadjem, M. F. Wong, and J. Wiart, “Exposure induced by WCDMA mobiles phones in
operating networks,” IEEE Trans. Wireless Comm. vol. 8, no. 12, pp. 5723-5727, Dec. 2009.
[3] Vrijheid M, Mann S, Vecchia P, et al. Determinants of mobile phone output power in a
multinational study: implications for exposure assessment. Occup Environ Med 2009;66:664–71.
[4] P. Joshi, D. Colombi, B. Thors, L. E. Larsson, and C. T¨ornevik, “Output power levels of 4G user

633
equipment and implications on realistic RF EMF exposure assessments,” IEEE Access, vol. 5, pp.
4545–4550, 2017.
Acknowledgment
This study is supported by Research on Biological Effect Project by Ministry of Internal Affairs and
Communications (Japan).
Figures

Figure 1. The histogram of the total downlink E-field strength of 800 MHz (a), the histogram of
4G Tx power of 800 MHz (b) and the histogram of the total downlink E-field strength of 2 GHz
(c), the histogram of 4G Tx power of 2 GHz (d) in the measurement during the whole drive
including residential/rural and Center of the Capital region.

634
Figure 2. The total downlink E-field strength of 800 MHz band (a), 4G Tx power of 800 MHz
(b), the total downlink E-field strength of 2 GHz band (c), and 4G Tx power of 2 GHz (d)

Figure 3. The histograms of the total downlink E-field strength of 800 MHz in the Center of the
Capital (a) and in residential/rural area (c). The histogram of 4G Tx power of 800 MHz(b) in
the Center of the Capital (b) and in residential/rural area (d).

635
PB-50 [14:30]
A method to analyze isotropy of electric field probe consisting of identical small dipoles
Byeongyoon Lee1 & Sang-Hwa Yi1
1Electrical Environment Research Center, Korea Electrotechnology Research Institute, Changwon, Korea,
51543
Keywords: Dosimetry (measurements), RF/Microwaves, Review, Commentary, Recommendation,
Evaluation
Presented by: Sang-Hwa Yi
A method is presented to analyze the isotropy of electric field probe consisting of a number of
identical small dipoles whose radiation patterns are torus shapes and which are arbitrarily positioned
and oriented.
Introduction
Electric field probe is one of the key components of electric field measurement instruments for the
assessment of human exposure to electromagnetic fields. Three main issues related to the probe
are sensitivity, compactness and isotropy. In this paper, a method to analyze the isotropy of electric
field probe consisting of a number of small dipoles which are arbitrarily positioned and oriented is
presented.
Methods
Fig. 1(Left) shows a cross-section of a torus shape of the radiation pattern of a small dipole aligned
with the z-axis on the yz plane. If a=b=0.5, the equation for this torus is as follows:

If a point represented with spherical coordinates in 3D space is on the surface of this torus, it must
satisfy Eq.(1). With this condition, we can obtain Eq.(2).

Therefore, we can get the contribution to the directivity of a small dipole arbitrarily positioned and
oriented in 3D space as shown in Fig.1(Right) if we can calculate Eq.(2). This can be obtained from
the following inner product.

where D is the center of the small dipole, P0 is any point on the surface of the sphere, Q is the
intersection point of surface S and the line passing through P0 and parallel to normal vector of
surface S. With this result, the isotropy of electric field probe consisting of a number of small dipoles
arbitrarily positioned and oriented can be analyzed. If the number of dipoles is N, the final directivity
F is expressed as Eq.(4).

Results
The presented method was applied to analyze two configurations of the electric field probe
consisting of 3 small dipoles. In Fig.2(Left), the axis of each dipole is rotated through angle 54.7o
about x-axis and then rotated through 0, 120 and 240o about z-axis. In Fig.2(Right), the axis of each
dipole is rotated through angle 54.7o about y-axis and then rotated about z-axis in the same was as
in Fig.2(Left). The direction of rotation is counterclockwise when viewed from the positive side of
636
each axis. Figure 3 shows the calculated final directivities for the left configuration of Fig.2 when the
distances from the origin to the center of the dipoles are 0 and 10 mm, respectively. It is confirmed
that directivity is almost uniform in all directions and isotropy can be achieved for the same
configuration as in references [1,2]. However, the larger the distance from the origin to the center of
the dipoles, the worse the isotropy.
Conclusions
The presented method is relatively easy to apply to the analysis of the isotropy of electric field probe
consisting of a number of small dipoles arbitrarily positioned and oriented. It also could be applied to
analyze the isotropy of electric field probe consisting of identical radiators if the radiation pattern of a
single radiator is known.
Acknowlegements
This research was supported by the KERI Primary research program of MSIT/NST(No. 18-12-
N0101-61)
References
[1] Bittera, M., Kovac, K. "One Way to Provision of HF E-field Sensor Isotropy, 3rd International
Conference on Measurement" - Measurement 2001, pp.163-166, Smolenice, Solvak Republic 14-17
May 2001
[2] Hartansky R. " Electromagnetic field sensors in EMC applications", Bratislave:FEI STU, 1999, 68
p. Doctoral Thesis.

Figures

Figure 1. Cross-section of radiation pattern of a small dipole aligned with the z-axis on yz
plane(left) and a small dipole positioned and oriented arbitrarily in 3D space(right)

637
 Figure 2. Two configurations of the electric field probe consisting of 3 small dipoles

Figure 3. Calculated directivities for the left configuration of Fig. 2 when distances from the
origin to the center of the dipoles are 0(left) and 10 mm(right), respectively.

PB-52 [14:30]
Reduction methods of ELF-MF radiated by high-voltage underground transmission lines
Seungwoo Lee1, Geun Teak Yeo1, Hong Lae Kim1 & Ho Sung An2
1Transmission & Substation Construction Dept., Korea Electric Power Corporation, Naju-si, Korea, 58217
2Transmission & Substation Dept., Korea Electric Power Corporation Research Institute, Daejeon-si, Korea,
34056
Keywords: Dosimetry (measurements), ELF/LF, Work in Progress
Presented by: Seungwoo Lee
We suggested the reduction methods of ELF-MF exposed by high-voltage underground
transmission lines in this research. To avoid the ELF-MF exposure from the underground cables, we

638
analyzed the effect according to the reduction methods which are the cable phase array, shielding
materials, shunt effect or eddy current effect, etc.
The electromagnetic fields are radiated by the electric power transmission lines, the base stations,
electric devices, etc. People are afraid of the electromagnetic fields, and some of them are
complained the symptom of electromagnetic hypersensitivity. However, the researching reports did
not show the defined results according to the biological study because of anxiety of electromagnetic
exposure. Therefore, some researchers developed the shielding or reducing methods.
The principles of shielding effects are widely classified by the shunt effect and the eddy current
effect. The shunt effect is normally applied to the ferromagnetic substance material which has a
characteristic of highly magnetic permeability, and the eddy current effect is applied to the metals
which have a characteristic of highly electric conductivity as shown in Figure 1. The shunt effect is a
phenomenon by reducing the fields in the magnetic materials, but the eddy current effect is a
phenomenon by canceling the magnetic fields between the original fields and the eddy current
fields.
Another shielding method is a cable phase array. In case of 3 cables, the 3-phase array (A, B, and
C) shows lower magnetic field strength than the horizontal array. When the cables are horizontally
arrayed, each cable is radiated the magnetic field respectively. Each field is overlapped, then the
magnetic field would be decreased. When the cables are arrayed by the triangle’s structure which
has 120 degree phase, the reducing rate of magnetic field strength is more effective.
In this study, we suggested the reduction methods of ELF-MF exposed by high-voltage underground
transmission lines.We compared the shielding effect due to the magnetic materials which are the
permalloy (PC) and the silicon steel Si-Fe (GO). The maximum reducing rate is almost 70% in the
laboratory environment.In the future, we are calculating and measuring the magnetic field strength
by the distance, shielding structure types, etc using materials and cable array.
Figures

Figure 1. The configurations of fields occurred by the shunt effect and the eddy current effect.
Figure 1(a) shows that the magnetic fields are horizontally interacted with the materials
composed with the copper and the ferrite. Figure 1(b) shows that the magnetic fields are
vertically interacted with the materials composed with the copper and the ferrite.

639
 Figure 2. The configurations of cable arrayed structure. (a) Cables are composed by the
horizontally arrayed structure in the tunnel (b) Cables are composed by the triangle arrayed
structure in the tunnel. The tunnel has the concrete and soil structures.

PB-54 [14:30]
Near-field power density for MPE evaluation of 5G wireless devices
Maryna Nesterova1
1R&D 5G solutions, Aprel Inc., Ottawa, Ontario, Canada, K2K 3J1
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Maryna Nesterova
The evaluation of a new generation of 5G wireless devices requires advanced methods in the near-
field measurement technologies. The high-speed and high- intensity modulated signal have to be
measured in the close proximity to a device to evaluate power density for safety matters. Because
the near-field zone processes have complex formation the testing methodologies should consider
the electric E and magnetic H field vector orientations and the beam-forming factors. The authors
present a new two-probe method as a solution for the 5G power density evaluation.
1. Introduction
The Maximum Permitted Emission (MPE) evaluation of the 5G mobile devices is based on the
measured power density averaged over some area. The options for a size of averaging area are
currently under discussion. The power density is derived through the Pointing vector theorem:

where B is magnetic flux density (T) and μ0 is the magnetic constant (H/m). Unless specified
otherwise, the terms magnetic field and magnetic field vector are used to refer to H.
2. Near-field zone for 5G devices
For the 5G wireless devices the complexity of processes in the near-field zone is magnified by the
beam-forming processes. The E and H fields in the close proximity to a transmitting device are
anisotropic and their oscillations are not in phase. In this conditions E and H waves should be
considered as unattached entities [1]. The two fields evolve differently in space to become

640
synchronised in phase in the far field zone. Because the 5G antennas are developed on the
massive MIMO technology the signal beam is dynamic, designed to maintain data transfer while
changing its direction. The field vector orientation is a key for the beam bearings. Thus, for the
complete evaluation of power density both E and H components should be quantified for amplitude
and vector orientation.
3. Two-probe method
The two-probe method is developed for the measurement of
electric and magnetic fields in the reactive near-field zone [2].
The E and H directional probes are rotated 360 degrees at
each point of a scan grid to evaluate the “footprint” of the field
anisotropy.The maximum amplitude Emax and Hmax vectors
and their orientation on the measurement plane are used to
calculate the Poynting vector at every measurement point.
4. Field vectors and their
projections
For the 5G devices the beam is
not necessarily perpendicular to
a device surface. In cases when
the evaluation plane/surface is
not orthogonal to orientation of the beam, the projections of field
amplitude vectors are evaluated. When both E and H field maximum
vectors are not on evaluation plane, their projections are not orthogonal
and the angle between them creates the phase shift effect.
5. Resonance phase shift
Another type of the phase shift between E and H waves is cause by the resonance of antenna. In
the reactive near-zone this shift defines the gain and efficiency of an antenna. For the 5G devices
the phase shift between elements is controlling the beam-forming processes and can vary in time.
This phase shift is a part of detected angle between measured maximum amplitude vectors.
6. Power density
The power density is evaluated through the cross product of E and H
vectors (Poynting vector theorem) and it accounts for the non-
orthogonal vector orientation via
sine of the angle between them.
For the power density evaluation
the source of the phase shift is
not relevant. Thus, calculated
Poynting vector (purple arrow),
which by definition is always perpendicular to the evaluation
plane, is only a projection of the beam Poynting vector (grey
arrow). Hence, only the partial power density is evaluated in this
case. the azimuth should be considered. In the near-field zone
the beam orientation changes at different distances from a device surface. As shown on the
diagram, the E vector is gradually adjusting its direction when the signal is progressing in space:
rotating 15 degrees from λ/4 to λ/2, another 15 degrees from λ/2 to λ, no changes in orientation for λ
and 2λ. For the full power density evaluation the averaged power density value should be multiplied
by cosine of the angle of signal direction offset from the azimuth.

641
7. Conclusions
For the 5G devices the near-field multi-layer measurement of both fields provides essential
information for the power density evaluation.
- The distances of evaluation planes
should be close to the device surface to
catch the beam,
- Phase shift between E and H field
vector projections on the evaluation
plane should be accounted in the power
density calculations,
- 360-degree scanning provides
information on the anisotropy of electromagnetic field of high frequency.
References
1. M. Nesterova and S. Nicol, “Analytical Study of 5G Beamforming in the Reactive Near-Field
Zone”, EuCAP Conference, London, April 2018.
2. M. Nesterova, S. Nicol, Y. Nesterova, “Two-probe methodfor advanced EM wave propagation
analysis,” IEEE RADIO 2017 International Conference, September 2017.

PB-56 [14:30]
Vortical motion of liquid medium in small, shallow, containers with different end wall
temperatures
Alessandra Paffi1, Francesco Feo1, Micaela Liberti1, Francesca Apollonio1, Asher Sheppard2 & Quirino
Balzano3
1Department of Information Engineering, Electronics and Telecommunications (DIET), Sapienza University of
Rome, Rome, Italy, 00184
2Asher Sheppard Consulting, Santa Rosa, CA, USA
3Department of Electrical and Computer Engineering (ECE), University of Maryland, College Park, MD, USA,
20742
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Alessandra Paffi
Using mica microparticles to detect liquid flow, curvilinear mass flow was observed in a shallow
liquid medium using a rectangular chamber with end walls held at a 2 K temperature difference,
resembling conditions in some in vitro RF biological studies. Thermal and fluid dynamics simulations
of the velocity field gave a maximum velocity of about 6×10-4 m/s. These observations indicate that
vortices could occur in some in vitro RF experiments. Prolonged non-Brownian motion in the cell
microenvironment might bias biological outcomes because nutrient flow in RF-exposed medium
cannot be replicated in controls. Future work for detection of liquid motion in Petri dishes is planned.

642
Introduction
In previous papers [1-2], we have shown that the medium of biological studies in Petri dishes
absorbs the bulk of radiofrequency (RF) energy in a non-uniform manner that causes an uneven
distribution of temperature within the liquid.
It is well known that there is convection in containers with shallow liquid levels and differentially
heated end walls [3-6]. We report observations of convective flow in a rectangular container with a
liquid level of less than a centimeter and wall temperature differences up to 2 K that are comparable
to temperature gradients produced under some RF exposure conditions [2]. We are particularly
interested in visualizing whether the liquid medium flows in closed curvilinear paths (loops), that is,
vortices. Such vortical motion could be understood as a consequence of increased entropy acquired
from non-uniform thermal energy absorption in the medium.
Experimental Method
Due to practical difficulties in making observations inside a closed RF apparatus such as a
waveguide or TEM cell, we built a 37.6 x 8.7 x 38.7 mm rectangular plexiglass container with one
end wall heated by an aluminum plate in contact with a resistor carrying current from a 50-Hz
current source (10 A max) that determined the hot wall temperature (Figure 1A).
The container was filled with a 0.8 g/L saline solution having the same conductivity (1.59 S/m) as
the Dulbecco Modified Eagle Medium (DMEM) used previously [2]. The temperature of the liquid
was observed using an infrared (IR) thermocamera (FLIR A325, FLIR Systems, Wilsonville, OR,
USA) (Figure 1B).
Liquid motion was detected by light reflected from mica microparticles (density 0.85 g/cm3, similar to
the density of water) gently placed on the surface from a dropper without stirring. Strong light
reflectivity of the mica particles makes them clearly visible while they slowly drift down from the
surface to the bottom of the liquid medium, thereby displaying vertical and horizontal liquid motions
within the container.
Experimental Results
During experiments, temperature distribution on the liquid surface was monitored using the IR
thermocamera and temperature variation was recorded at specific points labeled with markers.
Figure 2 shows the stability of the temperature difference obtained on the surface of the 1 cm deep
liquid between the “hot” and “cold” walls for a steady state difference of about 2 K (corresponding to
a resistor voltage drop of 1.8 V).
In this condition, the mica microparticles showed convective motion within the liquid clearly visible to
the naked eye. In the absence of a temperature gradient, mica particles placed on the liquid surface
reached the bottom along a vertical trajectory (Figure 3A). Conversely, when the temperature
gradient was created, mica particles followed a curved trajectory from the hot wall to the cold wall
near the liquid surface and in the opposite direction near the bottom of the container (Figure 3B).
This velocity field, typical of laminar convective motion, was confirmed by thermal and fluid
dynamics simulations performed using Comsol Multiphysics for the experimental conditions
described above. The simulation gave a maximum velocity of about 6×10-4 m/s. A similar velocity
distribution was predicted for a liquid depth of 5 mm and a temperature difference of 1 K, but the
maximum velocity was much lower, 1×10-4 m/s. In this latter case, the deflection of mica particles
from the vertical trajectory was difficult to detect with the available mica suspension. However, even
if small, a constant 6 mm/minute flow of the medium might be consequential in biological
experiments with a prolonged RF exposure.
Conclusions
Vortical motion exists in shallow liquids with a small temperature difference (<2 K) on end walls
spaced by about 35 mm. These results with mica microparticles suggest future experiments using
finely ground fish scales in order to detect liquid motion in Petri dishes even for the small rotational

643
velocities estimated by computer simulations for wall temperature differences <1 K.
A small, constant, non-Brownian motion of the medium during long term RF exposure of cell
preparations may have biological consequences due to a bias in the cell microenvironment created
by the flow of nutrients carried by the medium. This dynamic effect within the medium, although
small, cannot be replicated in the isoentropic, isothermal, non-exposed medium of the controls. The
existence of convective currents in liquid media inserts doubt about whether RF energy alone was
the cause of biological effects reported from some in vitro experimental studies.
References
[1] A. Paffi et al., Culture Medium Geometry: The Dominant Factor Affecting In Vitro RF Exposure
Dosimetry, International Journal of Antennas and Propagation, Hindawi Publishing Corporation,
International Journal of Antennas and Propagation Volume 2015, Article ID 438962.
[2] A. Paffi et al., In Vitro Exposure: Linear and Non-Linear Thermodynamic Events in Petri Dishes,
Bioelectromagnetics 36:527-537 (2015).
[3] B. Boehrert, Convection in a long cavity with differentially heated end walls, Int. J. Heat and Mass
Transfer. Vol. 40, No. 17, pp. 4105-4114, 1997.
[4] A. Bejan and L.C. Tien, Laminar Natural Convection Heat Transfer in a Horizontal Cavity with
Different End Temperatures, Journal of Heat Transfer, Nov. 1978 Vol. 100. pp 641-647.
[5] D. E. Cormack et al., Natural convection in a shallow cavity with differentially heated end walls. J.
Fluid Mechanics, (1974), vol. 65, part 1, pp. 209-229.
[6] J. Imberger, Natural convection in a shallow cavity with differentially heated end walls. J. Fluid
Mechanics, (1974), vol. 65, part 3, pp. 247-260.
Figures

Figure 1. Infrared camera (panel B) and plexiglass container with the heating plate (panel A).

644
Figure 2. A) Top view of the steady state temperature distribution of the container with the
liquid inside, 20 minutes after starting current in the resistor; Sp1 and Sp2 mark locations at
opposite sides of the liquid surface (Sp1 near hot wall and Sp2 near cold wall). B) IR camera
record of temperature at Sp1 and Sp2 shows steady state temperature difference of
approximately 2K.

Figure 3. Mica particles motion within the 1 cm deep liquid medium photographed from the
side in the absence of temperature gradient (A) and with a 2 K wall temperature difference
between the left (“hot”) and right (“cold”) walls (B). Superimposed arrows indicate top-to-
bottom straight line flow (gray cloud) without a temperature gradient and vortical flow creating
a widely dispersed mica cloud in the presence of a horizontal temperature gradient.

645
PB-58 [14:30]
Development of 28 GHz band exposure equipment for studies on thermal perception
thresholds of biological effects exposed to millimeter-wave at 5th generation wireless
systems candidate frequency band
Keita Sakakibara1, Tetsuya Sekiguchi1 & Takashi Hikage1
1Hokkaido University, Sapporo, Japan, 0600814
Keywords: Dosimetry (measurements), RF/Microwaves, Work in Progress
Presented by: Keita Sakakibara
Authors developed a radio-wave exposure equipment for studies on thermal perception thresholds
of biological effects exposed to millimeter-wave at 28 GHz band that is 5th generation wireless
systems (5G) candidate frequency band. To achieve high duty RF exposure on living body at 28
GHz required, we newly designed a lens antenna that can irradiate focused beam. In order to obtain
basic exposure performance of the developed equipment, we used physical phantom that simulated
human’s skin and achieved exposure experiments of temperature rise measurements during at 28
GHz band millimeter-wave exposure.
Introduction
While expanding use of millimeter-wave, such as 5th generation wireless systems (5G) and WiGig
(IEEE 802.11ad), is progressing, experimental studies for establishing safety guideline considering
the action on living organisms due to millimeter-wave exposure is required [1]-[3].
This paper presents development of 28 GHz band exposure equipment to achieve experimental
studies on thermal perception thresholds of biological effects exposed to millimeter-wave that is 5G
candidate frequency band. Newly developed millimeter-wave exposure equipment that uses beam
focusing lens antenna is shown. To achieve high duty RF exposure on living body at 28 GHz, we
designed a dielectric lens of antenna that can irradiate focused beam on required exposure area. To
obtain basic exposure performance of the developed equipment, the authors developed physical
phantom that simulates human’s skin and achieve exposure experiments.
Methods
An example of experimental system for studies on thermal perception thresholds of biological effects
exposed to millimeter-wave is illustrated in figure 1. Since the system requires to achieve high duty
millimeter-wave exposure on living body, the authors developed 28 GHz band exposure equipment
consists of a signal generator, 20 Watts transistor amplifier, and lens antenna. In manufacturing the
WR-28 rectangular waveguides are mainly used. As shown in figure 2, the antenna input power are
always monitored and control input power under safety level required by examiner.
Results and Conclusions
Figure 3 (a) shows an overview of developed exposure equipment. The equipment was constructed
in an anechoic chamber and tested to obtain basic exposure performance. We used physical
phantom that simulated human’s skin [4] and achieved exposure experiments of temperature rise
measurements during at 28 GHz band millimeter-wave exposure, as shown in figure 3 (b). After 60
sec. exposure, the temperature of the phantom surface increased 25 degree Celsius. In summary,
developed equipment can achieve required high exposure on the sample.
Acknowledgement
This work was partly supported by a Grant from the Ministry of Internal Affairs andCommunications,
Japan.
References
[1] Masaki Kouzai, Atsuhiro Nishikata, Taiji Sakai, Soichi Watanabe, ”Characterization of 60GHz
Millimeter-Wave Focusing Beam for Living- Body Exposure Experiments,” EMC’09/Kyoto, 22S1-1,
Kyoto, Japan, Jul. 2009.

646
[2] Kanako Wake, Watanabe Soichi, Atsuhiro Nishikata, Hiroyuki Enomoto, Yoshikazu Ugawa
"Dependence on Spot Size and Exposure Duration to Thermal Sensation by Millimeter-Wave
Exposure at 60 GHz" URSI commission B EMTS, Hiroshima, Japan, May 2013.
[3] Hiroki Sakiyama, Atsuhiro Nishikata, Kensuke Sasaki, and Yoshikazu Ugawa, “Analysis for Skin
Surface Temperature Rise Under Convergence Beam Exposure of 40GHz and 95GHz Millimeter
Wave,” IEICE Technical Report EMCJ2017-64, 2017 (In Japanese).
[4] C.Gabriel, “Compilation of the dielectric properties of body tissues at RF and microwave
frequencies,” Brooks Air Force Technical Report AL/OE-TR-1996-0037, 1996.
Figures

Figure 1. Experimental system for studies on thermal perception thresholds of biological

effects exposed to RF wave at 28 GHz band that is 5th generation wireless systems frequency
band.

Figure 2. 28 GHz band exposure equipment.

647
 Figure 3. Overview of developed 28 GHz band exposure equipment and measurement of
temperature rise.

PB-60 [14:30]
Non-sinusoidal magnetic field exposure: Comparison of time domain assessment methods
proposed in the Non-binding guide for practical implementation of EU directive 2013/35/EU
Gernot Schmid1, 2, Rene Hirtl1 & Theodoros Samaras2
1EMC & Optics, Seibersdorf Laboratories, Seibersdorf, Austria, A-2444
2Department of Physics, Aristotle University of Thessaloniki, Thessaloniki, Greece, GR-54124
Keywords: Dosimetry (measurements), ELF/LF, Work in Progress
Presented by: Gernot Schmid
Non-sinusoidal magnetic fields at common workplaces were measured and exposure was
comparatively assessed by the Weighted Peak Method in time domain (WPM-TD) and an alternative
time domain assessment method (TDA method) published in the non-binding guide for practical
implementation of the “workers directive” 2013/35/EU. It could be demonstrated that the TDA
method systematically and substantially (up to 11.7 dB) underestimates exposure compared to the
WPM-TD (indicated as the reference method in 2013/35/EU). We therefore recommend to remove
this method from the non-binding guide as soon as possible.

648
1. Introduction
The European directive 2013/35/EU [1], transposed into national laws of member states since
summer 2016, aims at an EU-wide harmonized protection level for workers exposed to electric,
magnetic and electromagnetic fields. In the frequency range 1 Hz – 10 MHz, exposure limit values
(ELV) in terms of the in situ electric field strength Ei induced inside the tissues by external electric
and magnetic fields have been defined in order to prevent relevant effects on the cell membrane
potential. Moreover, the directive also specifies action levels (AL) in terms of the (easily measurable)
external electric and magnetic field strengths, derived from the ELV under conservative
assumptions. Particularly for magnetic fields low action levels (lowAL), high action levels (highAL)
and high action levels for limb exposure (limbAL) are specified in 2013/35/EU in terms of frequency
dependent rms values of sinusoidal waveforms. It is known that large overestimations of exposure
may occur when non-sinusoidal or pulsed waveforms are assessed by the so called multiple
frequency rule (MFR), i.e., computing the exposure index (ExpInd) as the simple sum of weighted
(with the corresponding AL) spectral components of the rms or amplitude spectrum. In order to
avoid such overestimations, the so called “weighted peak method in time domain” (WPM-TD) is
specified by 2013/35/EU as the reference method for the assessment of non-sinusoidal waveforms.
However, it is noted in 2013/35/EU that also “other scientifically proven and validated exposure
evaluation procedures can be applied, provided that they lead to approximately equivalent and
comparable results.” Unfortunately, no indication is given what “approximately equivalent” means.
In the Non-binding guide [2] for practical implementation of directive 2013/35/EU there has been
published (beside the WPM-TD) another time domain assessment method called “Alternative
method: Simple Assessment on physiological basis”. This method is a modified version of the so
called “BGV B11” method, which was formerly widely used in Germany [3], and will be referred to as
“TDA method” in the following of this paper.
While some comparisons based on example waveforms are given in [2] between the WPM-TD and
other established assessment methods in frequency domain, there is no such comparison given
between the WPM-TD and the TDA method.
In this paper we use magnetic field waveforms measured at workplaces close to different devices
and compare the assessment results obtained with the TDA method with those obtained by the
reference method, i.e. the WPM-TD.
2. Materials and Methods
Magnetic field time domain waveforms were measured and recorded close to 7 different devices on
workplaces during routine operation using the ELT 400 field probe (Narda STS) and recording its
real time analog output signal with an A/D frontend (PXIe 6368, National Instruments). Time domain
characteristics, sampling rate and origin of the considered waveforms are given in figure 1. In all
considered cases the magnetic field occurs as pulsed waveform, and under typical work conditions
the pulse repetition period can be assumed 1 s or higher.
We implemented the WPM-TD method for the assessment based on the magnetic field highAL
according to 2013/35/EU by digital filters using LabView (National Instruments). In this case the
weighting function is represented by a simple first order high pass filter ([2],[4]) with a 3 kHz corner
frequency and a gain factor 1/(sqrt(2)*100) which can be applied to the recorded waveforms during
post processing.
As the definitions of the TDA method do not allow a reasonable automatic assessment, it was
applied “manually” after visual inspection of the time domain characteristics of the considered
waveforms. The relevant parameters of the TDA method are defined for sinusoidal, triangular,
trapezoidal and exponential waveforms only (Figure 2). Nevertheless it is stated in [2] and [3] that
the method can be used for all kind of pulsed waveforms. Figure 2 summarizes the parameter
definitions of the TDA method according to [2]. Basically, after determination of the parameters
τp,min (from which the equivalent frequency fp=1/(4*τp,min) is derived) and τD (sum of all time
intervals in which the waveform shows dB/dt ≥ 0), a weighting factor V is determined which is further

649
used to relax the maximum dB/dt values obtainable from the original AL given in 2013/35/EU (table
1).

Maximum values of (dB/dt)max and

HighAL according to 2013/35/EU and
corresponding maximum dB/dt for
sinusoidal waveforms
(dB/dt)mean according to the TDA-
method for the assessment based on
the HighAL

Brms [T] (dB/dt)max [Ts-1] (dB/dt)max [Ts-1] (dB/dt)mean [Ts-1]

1 Hz ≤ fp¹ < 3
0.3/f 2.67 2.7 * V 1.7 * V
kHz

3 kHz ≤ fp¹ < 10

0.0001 0.00089*f 0.001*fp*V 0.0006*fp*V
MHz

¹For sinusoidal waveforms the frequency fp corresponds to the actual frequency of the sinusoidal
waveform. For the TDA assessment of non-sinusoidal waveforms, fp corresponds to the
equivalent frequency (see definitions of the TDA method)

Table 1. High AL according to 2013/35/EU for the external (rms) magnetic induction and
corresponding (dB/dt)max values for sinusoidal waveforms, as well as maximum allowed values of
(dB/dt)max and (dB/dt)mean according to the TDA-method [2].
It must be noted, that according to the definitions of the TDA method, the parameter τp,min is
specified as the shortest duration of all “field changes” of the waveform, without any relation to the
corresponding extent of the field change (dB/dt). For sinusoidal, triangular or trapezoidal signal
curves, “field change” is defined from zero to the positive or negative peak value or from the positive
or negative peak value to zero, respectively (slightly different definitions apply for exponential field
changes (figure 2)). Moreover, it is noteworthy that the integration time TI is set to 1s for all cases in
which the pulse repetition period is ≥ 1s. Finally, from the definitions of the TDA method in [2], it
remains widely unclear what has to be understood by (dB/dt)mean in case of arbitrary waveforms.
Within this paper the term (dB/dt)mean was interpreted in its most common technical meaning, i.e.,
from the time when the pulse starts to leave the zero line until the pulse reaches its maximum
amplitude
3. Results
Figure 3 shows the ExpInd for the highAL calculated by the TDA method for the considered
magnetic field waveforms M1-M7 as depicted in figure 1, i.e., amplitude of the waveform scaled to a
value which yields an ExpInd = 1 when assessed by the reference method (WPM-TD). This means,
ExpInd > 1 in figure 3 would mean that the TDA method overestimates exposure compared to the
reference method, ExpInd < 1 mean that the TDA method underestimates exposure. In fact two
different ExpInd (one regarding (dB/dt)max and one regarding (dB/dt)mean) have been calculated and
the maximum of them is displayed in figure 3. As can be seen from figure 3 the TDA method
substantially underestimates exposure for all considered waveforms. The TDA method-based
exposure indices range from 0.62 to 0.26, i.e. the extent of underestimation ranges between a factor
of 1.6 (4.2 dB) and a factor of 3.9 (11.7 dB), depending on the waveform characteristics (highest
underestimation for waveform M7).
4. Discussion

650
As evident from the results presented above, the TDA method significantly and systematically
underestimates magnetic field exposure for all considered waveforms compared to the WPM-TD,
i.e., the reference method according to 2013/35/EU. Basically, two reasons for this behaviour of the
TDA method can be identified. The first one is related to the definition of the integration time TI
which equals 1s for all considered waveforms (having pulse repetition intervals of ≥ 1s). As the
pulse duration (and therefore the parameter τD) is significantly less than the pulse repetition time,
this leads to weighting factors in the range between 1.8 and 2.6, i.e., the TDA method “allows” dB/dt
values which are 1.8 – 2.6 fold above the maximum dB/dt values specified by the definition of the
highAL in 2013/35/EU. For example, waveform M1c is essentially a short (a few periods lasting)
burst of a 50 Hz sinusoidal waveform, and therefore the highAL (and corresponding dB/dt values)
for 50 Hz should apply from the perspective of evoking action potentials in peripheral nerves.
However the TDA method-specific weighting factor V=2.6 allows 2.6 fold higher dB/dt values, which
must be seen inconsistent with the underlying principles of 2013/35/EU. In the initial publication of
the TDA method [3] the specification which allows up to 1 s integration time was based on “practical
reasons”. However, as demonstrated in [5], this assumption is obviously not consistent with
biophysical basis and the aim of respecting the underlying ELV with respect to avoid action
potentials. The second reason, which at the end may lead underestimation factors substantially
higher than 2.6 is the fact that in the TDA method the parameter τp,min is specified completely
independent from the extent of field change (dB/dt) to which it corresponds. This means that minor
(small dB/dt) but short field changes determine the τp,min, and consequently the equivalent
frequency fp. As the allowable dB/dt increases with frequency (table 1), this lead to an additional
underestimation factor, which can be even higher than the factor of 2.6 coming from the above
mentioned integration time. An illustrative example for this effect is waveform M4, in which a short
but minor field change yields a τp,min corresponding to an equivalent frequency fp in the order of
4.2kHz although the actually relevant frequency (caused by the inverter) is around 2 kHz.
Overall, the observed underestimations for the considered (non-exhaustive) sample of waveforms
reached values of up to 3.9 (11.7 dB), and it is hard to believe that such an underestimation is still in
the range which would be covered by the criterion for alternative assessment methods “provided
that they lead to approximately equivalent and comparable results” given in 2013/35/EU.
5. Conclusion
The investigated TDA method, published as an alternative time domain assessment method in the
Non-binding guide to good practice for implementing Directive 2013/35/EU, systematically and
substantially underestimates exposure in case of short pulsed magnetic fields, when compared to
the WPM-TD method (reference method according to 2013/35/EU). Underestimation factors of up to
3.9 (11.7 dB) were found based on a (non-exhaustive) sample of magnetic field waveforms
measured at different workplaces. In view of the underlying goal of 2013/35/EU, i.e. an EU-wide
harmonized protection level for workers, we therefore strongly recommend to eliminate the TDA
method from the non-binding guide as soon as possible.
References
[1] EU Directive 2013/35/EU. Official Journal of the European Union L179/1-L179/21
[2] Non-binding guide to good practice for implementing Directive 2013/35/EU – Electromagnetic
Fields. ISBN 978-92-79-45869
[3] Heinrich H. 2007. Health Physics, Vol. 92:541-546
[4] Keller H. 2017. Health Physics, Vol. 113:54-65
[5] Schmid G., Hirtl R. 2017. Journal of Radiological Protection, Vol. 36:N77-N88
Figures

651
Figure 1. Considered waveforms M1 – M7 of magnetic induction, measured and recorded
(with sampling frequency fs) at different workplaces close to 7 different devices. For better
comparison the waveforms are linearly scaled to an amplitude value such that an assessment
with the WPM-TD gives an exposure index = 1 (for high AL).

652
Figure 2. Definition of the relevant parameters of the TDA method according to [2]

653
 Figure 3. Assessment results in terms of the exposure index for normalized waveforms M1-M7
obtained with the TDA method with respect to the highAL. Waveforms M1-M7 were normalized
to an amplitude which yields a WPM-TD-based exposure index = 1 before the TDA method
was applied.

PB-62 [14:30]
Radio frequency electromagnetic fields exposure assessment in different environments in
Slovenia
Tomaž Trček1, Blaž Valič1 & Peter Gajšek1
1Institute of Non-Ionizing Radiation (INIS), 1000 Ljubljana, Slovenia
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Blaž Valič
In this study, we investigated exposure to Radio frequency (RF) electromagnetic fields (EMF) in
different environments in Slovenia (urban areas, smaller towns, rural areas). In total, we performed
60 measurements, 20 for each environment. Measurements were band selective, this means 11
most important frequency bands, where most of the systems operate were measured (FM, DVB-T,
GSM ...). The measured values were compared with the reference values according to ICNIRP
guidelines for general public. The results showed that the measured values are quite low and
generally comparable with the results of similar studies [2]. The highest total RF-EMF exposure was
below 0.7 percent according to ICNIRP guidelines for general public.
Introduction
Radio-frequency electromagnetic fields (RF EMF) are nowadays the main channel for data, video
and voice transmission. The sources of RF-EMF are therefore essential and are the main
component of modern telecommunication systems. Most widely used sources are base stations,
radio and television transmitters, WiFi access points and similar devices.
In order to determine the actual exposure to RF-EMF in envrironment, we have performed a total of
60 RF-EMF measurements. 20 measurement locations were selected in urban areas, 20 in smaller
towns and 20 in rural areas.

654
Similar surveys have shown that exposure to RF-EMF due to base stations and other EMF sources
in human accessible locations in the European Union are well below ICNIRP guidelines for general
public [1]. At a height of 1m above the ground the maximum exposure to EMF according to ICNIRP
guidelines for general public reached up to 10 %, on average less than one percent [2].
Measurement methods and equipment
Measurements were performed from January to July 2017 in Slovenia, mainly in locations where
base stations are located nearby. In total, measurements in 60 locations were performed. The
measurement locations were divided into three groups according to type of environment, i.e. urban
areas, smaller towns and rural areas. For each group 20 measurements were performed. Locations
of measurements are shown in Figure 1. The measurements were performed in human accessible
locations 1m above the ground. At each measurement location, measurements were performed
according to the 'sweep' protocol, which provides an overview of area approximately 1 x 1m. The
maximum measured value is stored as the measurement result. The measurements were carried
out between 9am and 4pm, when exposure RF-EMF is usually the highest since the base station
traffic is usually the highest.
The measurements were performed according to the standard SIST EN 50492 [7]. Band selective
measurements were performed at all measured locations for the most important frequency bands
(FM radio, DAB radio, TETRA, DVB-T, 800MHz, GSM-R, 900MHz, 1800MHz, 2100MHz, WiFi,
2600MHz). At each location electric field E [V/m] for each frequency band was measured.
Measurement equipment
Measurements were carried out with state-of-the-art measurement system, the code selective
spectrum analyser Narda SRM-3006 with 3-axis probe for measurement electric field in the
frequency range from 27 MHz to 3 GHz. The extended measurement uncertainty for entire
measurement system is ± 2.9 dB. The measurement system is specially adapted for measuring
base stations of mobile systems of all three generations (GSM, UMTS and LTE).
Evaluation of exposure to electromagnetic radiation
The measured electric field were evaluated according to ICNIRP guidelines for general public.
Exposure index (EI) was calculated according to equation E 1.
E 1: EI=∑(Ei/LE,i)2 where
f- frequency of measured signal
Ei- electric field for i-source or i-frequency
LE,i- ICNIRP reference level for i-source or i-frequency
Measurement results
The measurement results showed that electric fields for all environments were low. Maximum and
average values of measured electric field for all frequency bands and environments are given in
Table 1.
Table 1 Maximum and average values of electric field in V/m for all environments (urban areas,
smaller towns, rural areas).
Urban areas Smaller towns Rural areas
Frequency
band Eavg Emax Eavg Emax Eavg Emax

FM radio 0.120 0.545 0.052 0.154 0.079 0.694

DAB radio 0.023 0.068 0.018 0.035 0.018 0.035

TETRA 0.042 0.344 0.014 0.024 0.015 0.027

655
 DVB-T 0.034 0.083 0.051 0.226 0.061 0.220

800 0.464 2.265 0.376 1.764 0.451 1.115

GSM-R 0.062 0.352 0.022 0.152 0.020 0.074

900 0.914 1.700 0.744 2.231 0.584 1.372

1800 0.994 2.346 0.391 1.016 0.259 1.649

2100 0.807 1.944 0.278 0.737 0.121 0.766

WiFi 0.092 0.474 0.029 0.136 0.019 0.036

2600 0.058 0.681 0.022 0.023 0.024 0.047

Total 1.796 3.610 1.076 3.022 0.901 2.201

From Table 1 it is evident that the maximum total electric fields were lower than 4 V/m for all
environments. Maximum value was 3.61 V/m for urban areas, 3.02 V/m for smaller towns and 2.20
V/m for rural areas, respectively.
Similar to maximum value, average value of electric field were the highest for urban areas. It was
1.80 V/m for urban areas, 1.08 V/m for smaller towns and 0.90 V/m for rural areas, respectively.
From Table 2 it is evident that the maximum total exposure index according to ICNIRP did not
exceed 0.7 %. Maximum exposure index was 0.64 % for urban areas, 0.52 % for smaller towns and
0.21 % for rural areas, respectively.
Table 2 Maximum and average exposure index (EI) according to ICNIRP for all environments.
Urban areas Smaller towns Rural areas
Frequency
band %avg %max %avg %max %avg %max

FM radio 0.0033 0.0379 0.0005 0.0030 0.0035 0.0615

DAB radio 0.0001 0.0006 0.0000 0.0002 0.0000 0.0002

TETRA 0.0009 0.0151 0.0000 0.0001 0.0000 0.0001

DVB-T 0.0002 0.0008 0.0006 0.0057 0.0008 0.0055

800 0.0333 0.3434 0.0266 0.2083 0.0196 0.0832

GSM-R 0.0006 0.0071 0.0001 0.0013 0.0000 0.0003

900 0.0638 0.1652 0.0501 0.2846 0.0268 0.1076

1800 0.0370 0.1613 0.0074 0.0302 0.0066 0.0797

2100 0.0229 0.1016 0.0031 0.0146 0.0013 0.0158

WiFi 0.0007 0.0060 0.0000 0.0005 0.0000 0.0000

2600 0.0006 0.0125 0.0000 0.0000 0.0000 0.0001

656
 Total 0.1634 0.6363 0.0885 0.5207 0.0587 0.2071
From Table 2 it is evident that average total exposure index did not exceed 0.17 % of the reference
levels. Average total exposure index for urban areas was 0.16 %, 0.09 % for smaller towns and
0.06% for rural areas. Average total exposure index for all environments was 0.10 %
In terms of exposure index, average values are more important than maximum values as they better
represent actual exposure. From Table 2 it is evident that base stations in frequency band 900 MHz
have the greatest contribution to total exposure index. In urban areas, base stations 900 contributed
0.06 %, in smaller towns 0.05 % and in rural areas 0.03 %, respectively. The average contribution of
frequency band 900 for all 60 measurements for all types of environment was 0.05 % according to
ICNIRP guidelines. In urban areas, the band 900 contributed 39 % to total exposure, 57 % in
smaller towns and 46 % in rural areas. Base stations 900 were still the most widespread system,
although it is the oldest system of base stations still operating.
Second greatest contributor to total exposure index is frequency band 800. This frequency band
began to be used for mobile telephony when DVB-T was introduced and some UHF frequency
spectrum was released. Due to good coverage, base stations 800 are now located on most of all
base station locations. The average contribution of band 800 for all 60 measurements for all types of
environment was 0.03 %. In urban areas and smaller towns band 800 contributed 0.03 %, while in
rural areas 0.02 %, respectively. In urban areas, band 800 contributed 20 % to total exposure index,
30 % in smaller towns and 33 % in rural areas.
One of the reason that exposure index was much higher in band 800 and 900 than in band 1800,
2100 and 2600 MHz are also significantly lower reference levels. At lower frequencies, reference
levels are lower. According to E 1, exposure index is calculated as square of measured values
divided by reference levels. Therefore the same electric field at lower frequency contributes much
more to total exposure than at higher frequency. However, the exposure in band 1800 and 2100 was
higher in urban areas than in smaller towns and rural areas. In urban areas, the need for these base
stations is greater, as the traffic capacity is more important than the coverage.
The measured values of electric field of other sources were much lower. Nevertheless, some trends
are noticeable. The Figure 2 shows the average values of electric field for all frequency bands
excluding base stations.
For WiFi, it is noticeable that the average value was significantly higher in urban areas than in
smaller towns and rural areas, although values are extremely small. Average exposure for urban
areas was 0.0007% according to ICNIRP guidelines, 0.00004% for smaller towns and 0.00001% for
rural areas. This was expected due to public Wi-Fi networks usually installed in urban areas, rather
than in smaller towns and rural areas. WiFi networks inside buildings were usually negligible. DVB-T
caused higher exposure in rural areas than in urban areas and smaller towns, as DVB-T
transmitters were usually located outside urban areas. The average exposure from band DVB-T for
urban areas was 0.0002% according to ICNIRP guidelines, 0.0006% for smaller towns and 0.0008%
for rural areas, respectively. Average exposure index for all environments was 0.0005%.
Conclusion
Measurements of RF-EMF in different environments in Slovenia have shown that exposure index is
quite low at all measurement locations and well below ICNIRP guidelines for general public [1]. The
highest total exposure index was 0.64% for urban areas, 0.52% for smaller towns and 0.2% for rural
areas. The average exposure index for all environments was below 0.21%.
References
[1] ICNIRP. Guidelines for limiting exposure to time-varying electric, magnetic, and electromagnetic
fields (up to 300 GHz). Health Physics 74: 494-522, 1998.
[2] Gajšek P, Ravazzani P, Wiart J, Grellier J, Samaras T, Thuróczy G. Electromagnetic field
exposure assessment in Europe radiofrequency fields (10 MHz–6 GHz). J Expo Sci Environ
Epidemiol. 25: 37-44, 2015.
657
[3] Trček T, Valič B, Kotnik T, Gajšek P. Elektromagnetna sevanja v okolici baznih postaj LTE.
Elektroteh. Vestn. 81: 39-44, 2014.
[4] Joseph W, Verloock L, Goeminne F, Vermeeren G, Martens L. Assessment of general public
exposure to LTE and RF sources present in an urban environment. Bioelectromagnetics 31:
576–579, 2010.
[5] Joseph W, Verloock L, Goeminne F, Vermeeren G, Martens L. In situ LTE exposure of the
general public: Characterization and extrapolation. Bioelectromagnetics 33: 466 – 475, 2012.
[6] IMST. Confidence by evidence – Putting LTE to the test. IZMF, Berlin, 2013.
[7] SIST EN 50492:2009 – Basic standard for the in-situ measurement of electromagnetic field
strength related to human exposure in the vicinity of base stations.
Figures

Figure 1. Measurement locations: black (urban areas), blue (smaller towns) and cyan (rural
areas).

658
 Figure 2. Average values of electric field in V/m for all environments (urban areas, smaller
towns, rural areas) for frequency bands excluding base stations.

PB-64 [14:30]
Measurement of EMF exposure around small cell base station installations
Marthinus Van Wyk1, Christo Visser1 & Christiaan le Roux1
1Alphawave Mobile Network Services, Stellenbosch, South Africa, 7600
Keywords: Dosimetry (measurements), RF/Microwaves, Completed (unpublished)
Presented by: Marthinus Van Wyk
Small cell base stations are used frequently as a mobile network deployment method. A
measurement program was conducted to perform EMF exposure measurements around small cell
base station installations. Results are compared to available data around base stations in general.
As mobile network technologies and usage change, cells have become smaller to meet the demand
for data throughput. Thus, small cells are deployed to augment the network of macro base stations.
These small cells transmit less power than a macro base station and cover a smaller geographical
area. The small cells are predominantly installed to service areas with a high density of people or to
cover a small geographical area.
Various measurement programs of the electromagnetic exposure have been conducted around
base stations and these results have been published. There is, however, little data available on the
electromagnetic exposure levels around small cell installations. This scope of this project was to
measure the electromagnetic exposure around small cell installations, and also to compare the
results to other results available.
Measurements were conducted around 100 sites across 3 South African and 2 European cities.
Sites were classified into groups: “Indoor”, “External 2-4 m” and “External >4 m”. The “indoor” group
is used for installations where antennas are installed indoors, typically a distributed antenna system
to provide coverage inside a building. Installations where antennas are mounted outdoors at a
height of between 2 and 4 metres are classified as “External 2-4 m”. A small cell installation at a bus
stop is an example of a site in this group. “External >4m” is used for scenarios where the antenna is
mounted outdoors, at a height of at least 4m. A small cell installed on a lamp pole is an example of

659
such an installation.
The measurements were performed based on IEC 62232:2017. At each installation 3 positions
accessible to the general public were identified and measurements performed at these locations. An
active download from the site being surveyed was maintained for the duration of the measurements.
This ensures that the small cell being surveyed was live and carrying traffic at the time of the survey.
The measured results are compared for the different site groups described above. As can be seen
from Figure 1, the difference between the measured exposure level at the three site types is small.
Further, the measured results are compared to that of other projects performed around mobile
phone base stations in general. As an example, the comparison to a dataset of general
measurements around base stations in South Africa between 2006 and 2016 is shown in Figure 2.
Figures

Figure 1. Measured small cell exposure, also shown for different site groups.

Figure 2. Measured small cell exposure, compared to base stations in general (South Africa
2006 to 2016 dataset).

660
PB-66 [14:30]
Cancellation effect is present in high-frequency bipolar pulse treatments and depends on
electroporation buffer composition
Janja Dermol-Cerne1, Tamara Pezić1, Matej Reberšek1 & Damijan Miklavčič1
1University of Ljubljana, Faculty of Electrical Engineering, Ljubljana, Slovenia, SI-1000 Ljubljan
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Janja Dermol-Cerne
In electrochemotherapy and irreversible electroporation as ablation technique, 100 μs long pulses
are applied at 1 Hz which causes muscle contractions and pain. Recently, bursts of short bipolar
pulses (H-FIRE) were suggested as an ablation technique with no muscle contractions.
Independently of H-FIRE studies, cancellation effect was observed with nanosecond pulses, where
the effect of the first pulse is partially canceled by the second pulse of reverse polarity. In our study,
we focused on cancellation effect in the range of pulses, usually used in HFIRE treatments. We
varied pulse duration and delay between pulses. Our results show that cancellation effect is also
present with HFIRE pulses, and is electroporation buffer dependent.
1 Introduction
Electroporation is a phenomenon in which application of short high-voltage electric pulses increases
cell membrane permeability to molecules for which the membrane is usually impermeant [1]. If cell
recovers, it is considered reversible electroporation. When damage is too extensive, and cell dies, it
is considered irreversible electroporation. Electroporation is used in biotechnology, food processing,
and medicine [2], e.g., electrochemotherapy (ECT), gene therapy, irreversible electroporation as an
ablation technique (IRE), DNA vaccination, and transdermal drug delivery.
In ECT and IRE, square 50-100 µs long monopolar pulses are usually applied at repetition
frequency 1 Hz. The main drawbacks to the treatment of tissues with these monopolar pulses are
discomfort and pain [3], the need to administer muscle relaxants and anesthesia [4], need for
synchronization of the pulse delivery with the electrocardiogram [5] and inhomogeneous tissue
impedance [6].
One of the possibilities to overcome these obstacles was suggested to be high-frequency
irreversible electroporation (H-FIRE). In H-FIRE, long monopolar pulses are replaced with bursts of
a few µs long bipolar pulses [7], [8] which mitigate muscle contractions, however at the expense of
delivering higher energy. Independently of the H-FIRE research, several reports on so-called
cancellation effect have been published [9]. Cancellation effect is observed when the effect of the
first pulse is canceled by the second pulse of reversed polarity. Cancellation effect was observed for
a pulse duration of few ns to 900 ns [10] which is already in the range of H-FIRE pulses. The reason
for cancellation effect has not yet been determined; suggested mechanisms are the assisted
membrane discharge, reduced pore expansion, two-step oxidation of membrane phospholipids,
reverse transmembrane ion transport and localized charging and discharging across the membrane
[9].
In our study, we focused on the cancellation effect in the range of H-FIRE pulses, i.e., 1-10 μs long
pulses. We evaluated cell membrane permeability and cell survival after H-FIRE pulses while
changing the pulse duration, and the pause between pulses in two different commonly used
electroporation buffers.
2 Materials and methods
Chinese hamster ovary cells were grown in full HAM F-12 culture medium at 37°C in a humidified
5% CO2. After reaching 80% confluency, cell suspension was prepared (2x106 cells/ml) in two
electroporation buffers [11] (Table1). Cells were transferred between 2 mm stainless steel parallel
plate electrodes, pulses were delivered, and the sample was incubated at 22°C for 3 min. In

661
permeability experiments, the cell suspension was supplemented with 100µg/ml of propidium iodide
(PI).
Table 1: Composition of electroporation buffers

Electroporation buffer Composition Electrical conductivity [mS/cm]

10 mM K2HPO4/KH2PO4
Low-conductivity buffer 1 mM MgCl2 1.76
250 mM sucrose

10 mM K2HPO4/KH2PO4
High-conductivity buffer 1 mM MgCl2 19.12
150 mM NaCl

Cell permeability was determined on a flow cytometer (Attune NxT, Life Technologies, USA), where
10,000 events were measured. Percentage of permeabilized cells was determined from PI
fluorescence intensity histogram by gating the permeabilized from non-permeabilized cells.
Cell viability was determined by the MTS assay (CellTiter 96® AQueous One Solution Cell
Proliferation Assay, Promega, USA). MTS tetrazolium compound is reduced by viable cells into a
colored formazan product. After pulse application, 2x104 cells were transferred to 96-well and grown
for 24h. 20 µl of MTS test was added to each well and incubated for 2.5 hours. Absorbance at 490
nm was measured with a spectrofluorometer (Tecan Infinite M200, Tecan, Austria). Survival was
calculated by normalizing the absorbance of the samples to the control.
For the application of pulses, we used a laboratory prototype pulse generator (University of
Ljubljana) based on H-bridge digital amplifier with 1 kV MOSFETs (DE275-102N06A, IXYS, USA)
[8]. We measured the delivered voltage and current by oscilloscope Wavesurfer 422, 200 MHz,
differential probe ADP305 and a current probe CP030 (LeCroy, USA).
We applied monopolar and bipolar pulses. Monopolar pulse treatment was composed of 8x100 µs
pulses, 1 Hz repetition frequency and served as a positive control, and bipolar pulses treatment of 8
bursts of high-frequency bipolar pulses of different durations and delays between positive and
negative pulses and between bipolar pulses, 1 Hz repetition frequency (Fig. 1 and Table 2). The on-
time of pulses in the monopolar and bipolar treatment was 800μs.
Table 2: Description of HFIRE pulse parameters.

The delay between

positive and negative
Pulse length (µs) Number of bipolar
pulse and between Number of bursts (-)
ΔT bipolar pulses (µs) pulses in one burst (-)
Δdelay

1 0.5 50 8

1 1 50 8

1 5 50 8

662
 1 10 50 8

1 100 50 8

1 1000 50 8

1 10000 50 8

5 0.5 10 8

5 1 10 8

5 5 10 8

5 10 10 8

5 100 10 8

5 1000 10 8

5 10000 10 8

10 0.5 5 8

10 1 5 8

10 5 5 8

10 10 5 8

10 100 5 8

10 1000 5 8

10 10000 5 8

3 Results
We determined permeability and survival curves as a function of the applied electric field in two
different electroporation buffers (Fig. 2) and as a function of the delay between pulses (Fig. 3).
Monopolar pulses were more efficient than bipolar pulses (Fig. 2a vs. 2b). The effect of the bipolar
pulses depended on the electroporation buffer (Figs. 2b and 2c), the delay between the pulses (Fig.
3) and the duration of the pulse (Fig. 3).
4 Discussion
663
Traditionally, it was believed that bipolar pulses cause symmetric and higher cell membrane
permeability than monopolar pulses while avoiding high cell mortality [12] with reduced electrolytic
contamination [13] and muscle contractions [7]. However, this was challenged in [14], where lower
efficiency of bipolar nanosecond pulses in comparison with monopolar nanosecond pulses was
observed. Similar results were also found in μs domain [8] which could have an important effect on
H-FIRE treatments. We explored the cancellation effect in the range of the H-FIRE pulses.
Longer monopolar pulses were more efficient than the H-FIRE pulses of comparable energy in both
permeability and survival experiments (Fig. 2a vs. 2b) as already observed [8]. In the low-
conductivity buffer (Fig. 3b), in the permeability experiments with ΔT = 1 and 5 μs long pulses, the
opposite of cancellation effect was present, i.e., longer delay between pulses caused lower
efficiency of treatment. In survival experiments, there was no obvious cancellation effect present. In
the high-conductivity buffer (Fig. 3c), there was cancellation effect present in permeability as well as
in survival experiments. The cancellation effect decreased with increasing pulse length but was still
present with 10 μs long pulses. The reason for discrepancy could be different electroporation buffers
used in experiments. It was already shown that the composition of electroporation buffer influenced
cell-sensitization [11]. We have shown that a similar conclusion can also be extended to cancellation
effect.
Three main conclusions can be observed from our work. 1) Monopolar pulses are more efficient
than short bipolar pulses. 2) Cancellation effect is present in the H-FIRE treatments and could be
responsible for decreased treatment efficiency in comparison with the longer monopolar pulses. 3)
The cancellation effect is electroporation-buffer dependent.
5 Acknowledgments
This work was supported by the Slovenian Research Agency [research core funding No. P2-0249,
P3-0003, and funding for Junior Researchers to JDČ and TP]. Experiments were performed within
the Infrastructure Programme: Network of research infrastructure centers at University of Ljubljana
(MRIC UL IP-0510), and in the scope of the LEA-EBAM.
6 References
[1] T. Kotnik, P. Kramar, G. Pucihar, D. Miklavčič, and M. Tarek, “Cell membrane electroporation-
Part 1: The phenomenon,” IEEE Electr. Insul. Mag., vol. 28, no. 5, pp. 14–23, 2012.
[2] S. Haberl, D. Miklavčič, G. Serša, W. Frey, and B. Rubinsky, “Cell membrane electroporation –
Part 2: The applications,” Electr. Insul. Mag. IEEE, vol. 29, no. 1, pp. 29–37, 2013.
[3] C. B. Arena and R. V. Davalos, “Advances in Therapeutic Electroporation to Mitigate Muscle
Contractions,” J. Membr. Sci. Technol., vol. 02, no. 01, 2012.
[4] C. Ball, K. R. Thomson, and H. Kavnoudias, “Irreversible Electroporation: A New Challenge in
‘Out of Operating Theater’ Anesthesia,” Anesth. Analg., vol. 110, no. 5, pp. 1305–1309, May 2010.
[5] B. Mali et al., “Electrochemotherapy of colorectal liver metastases - an observational study of its
effects on the electrocardiogram,” Biomed. Eng. OnLine, vol. 14, p. S5, 2015.
[6] S. P. Bhonsle, C. B. Arena, D. C. Sweeney, and R. V. Davalos, “Mitigation of impedance changes
due to electroporation therapy using bursts of high-frequency bipolar pulses,” Biomed. Eng. OnLine,
vol. 14, no. Suppl 3, p. S3, 2015.
[7] C. B. Arena et al., “High-frequency irreversible electroporation (H-FIRE) for non-thermal ablation
without muscle contraction,” Biomed. Eng. OnLine, vol. 10, no. 1, p. 102, 2011.
[8] D. C. Sweeney, M. Reberšek, J. Dermol, L. Rems, D. Miklavčič, and R. V. Davalos,
“Quantification of cell membrane permeability induced by monopolar and high-frequency bipolar
bursts of electrical pulses,” Biochim. Biophys. Acta BBA - Biomembr., vol. 1858, no. 11, pp.
2689–2698, Nov. 2016.
[9] A. G. Pakhomov et al., “Cancellation of cellular responses to nanoelectroporation by reversing
the stimulus polarity,” Cell. Mol. Life Sci., vol. 71, no. 22, pp. 4431–4441, Nov. 2014.

664
[10] C. M. Valdez, R. A. Barnes, C. C. Roth, E. K. Moen, G. A. Throckmorton, and B. L. Ibey,
“Asymmetrical bipolar nanosecond electric pulse widths modify bipolar cancellation,” Sci. Rep., vol.
7, no. 1, Dec. 2017.
[11] J. Dermol, O. N. Pakhomova, A. G. Pakhomov, and D. Miklavčič, “Cell Electrosensitization
Exists Only in Certain Electroporation Buffers,” PLOS ONE, vol. 11, no. 7, p. e0159434, Jul. 2016.
[12] T. Kotnik, L. M. Mir, K. Flisar, M. Puc, and D. Miklavčič, “Cell membrane electropermeabilization
by symmetrical bipolar rectangular pulses. Part I. Increased efficiency of permeabilization,”
Bioelectrochemistry Amst. Neth., vol. 54, no. 1, pp. 83–90, Aug. 2001.
[13] T. Kotnik, D. Miklavcic, and L. M. Mir, “Cell membrane electropermeabilization by symmetrical
bipolar rectangular pulses. Part II. Reduced electrolytic contamination,” Bioelectrochemistry Amst.
Neth., vol. 54, no. 1, pp. 91–95, Aug. 2001.
[14] B. L. Ibey et al., “Bipolar nanosecond electric pulses are less efficient at electropermeabilization
and killing cells than monopolar pulses,” Biochem. Biophys. Res. Commun., vol. 443, no. 2, pp.
568–573, Jan. 2014.
Figures

Figure 1. Waveform of one pulse in HFIRE treatment. The pulse consisted of the positive and
negative pulse and the delay between them. In the experiments, we changed the voltage (ΔU),
the delay between positive and negative pulse and between pulses (Δdelay) and the duration
of a positive or negative pulse (ΔT). The delay between positive and negative pulse and
between bipolar pulses was always the same. Pulses were delivered in eight bursts delivered
at repetition frequency 1Hz.

665
Figure 2. Permeability and survival as a function of the ratio voltage to distance between the
electrodes. A) Permeability and survival after 8x100μs monopolar pulses in the low-
conductivity buffer, which served as a positive control. B) Permeability and survival after
application of ΔT =1μs long HFIRE pulses with varied delay (Δdelay = 1μs or 10ms) between
the positive and negative pulse and between bipolar pulses in the low-conductivity buffer, and
C) in the high-conductivity buffer. The monopolar pulse/burst repetition frequency was 1Hz
and 8 monopolar pulses/bursts were delivered.

666
Figure 3. Permeability and survival as a function of the delay between pulses. The Δdelay was
the same between the positive and negative pulse and between bipolar pulses and was varied
from 0.5μs do 10 ms. Three pulse lengths ΔT (1μs, 5μs or 10μs) were tested. In the control
sample, the survival was assumed to be 100% and permeability 0%. The permeability
experiments were done at 1000 V/cm and the survival experiments at 3000 V/cm. The burst
repetition frequency was 1Hz and 8 bursts were delivered. A) The low-conductivity buffer, B)
the high-conductivity buffer.

667
PB-68 [14:30]
STUDENT PAPER
Flexible conductive polymer microelectrode arrays for electropulsation, neurostimulation
and electroporation in vitro and in vivo
Gerwin Dijk1, Hermanus Ruigrok2 & Rodney P. O'Connor2
1Bioelectronics, Ecole des Mines de Saint-Etiennes - Panaxium, Gardanne, France, 13120
2Bioelectronics, Ecole des Mines de Saint-Etiennes, Gardanne, France, 13120
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Gerwin Dijk
Organic conductive polymers are highly versatile materials that have been widely adopted in
commercial electronics as display technology, transistors, LEDs and solar cells. Our research is
exploring their use for interfacing biology with electronics. We have developed microelectrode array
devices using the conductive polymer PEDOT:PSS for electropulsation, neurostimulation and
electroporation investigations in vitro and in vivo. Here we will present our work using plastic
bioelectronics to develop multiwell electrode arrays for in vitro studies using live cell imaging and
flexible, implantable electrodes that are suitable for studying electropulsation effects with in vivo
imaging and electrophysiology.
Introduction
Microfabricated multielectrode array (MEA) devices are shown that consist of arrays of interdigitated
fingers of gold that act as electrodes to generate an electric field (Figure 1). Microfabrication with
standard photolithography equipment allows for small resolution patterning (micron scale) and a
flexibility in electrode design and spacing1,2.
Results
A key aspect of the present work will be the demonstration of the utility of coating gold electrodes
with PEDOT:PSS, forming a conductive polymer3 layer that exhibits volumetric capacitance. This
enhanced capacitance allows for 3D coupling with the biological environment and enhances the
charge injection capacity. Moreover, the properties of PEDOT:PSS makes it a highly biocompatible,
soft material well suited to microscale devices for biological applications and may obviate some
issues associated with sizing electrodes in the scale of cells4 and high electric field at edges. The
low absorbance in the visible spectrum of PEDOT:PSS is another significant property, since it allows
for the construction of transparent microelectrode arrays without gold traces (Figure 1) that are
highly convenient for imaging with light microscopy.
Finally, the patterning of organic materials like PEDOT:PSS on parylene C allows for the design of
flexible, conformable devices for implantation and in vivo imaging assays using fluorescence
microscopy, as shown in Figure 1.
Cells can be cultured on interdigitated finger MEAs designed in a multiwell platform that can be used
for medium content screening of electropulsation effects. Figure 2 shows a cell culture slide
containing an array of interdigitated electrodes that can be used to electropulse cells. This platform
allows for multiple parallel experiments (and nonstimulated controls) using live cell imaging with
fluorescent microscopy.
To evaluate our in vitro multiwell devices, human glioblastoma cells (U87-MG) were seeded in each
well and loaded with fluo4-AM and propidium iodide for real-time live-dead staining that also reveals
cellular calcium responses. An example is shown in Figure 2, where 7 monophasic pulses of 300 µs
at 60 V were applied over 500 μm distances, showing the temporal cytosolic calcium responses due
to electropulsation as well as eventual cell permeabilization.
Material and methods
Multielectrode Array Fabrication: The fabrication process includes the deposition and patterning of

668
gold, parylene C and PEDOT:PSS, as previously described3. Photoresist is spin coated, exposed
and developed followed by the deposition of a chromium seed layer, gold and then lift-off. A
sacrificial parylene C layer is deposited, photoresist is spin coated, exposed and developed followed
by etching of the parylene C. PEDOT:PSS is mixed with a crosslinker, spin coated and hard baked
in the oven. Finally, the sacrificial parylene layer is peeled-off leaving behind the PEDOT:PSS on the
electrodes.
Cell culture: 100 µL at 8.5.105 cells/mL of U87-MG glioblastoma cells where seeded in each well of
the exposure device 24h before the experiment and kept at 37 °C, 5 % CO2 in DMEM media
supplemented with 10 % FBS, Glutamax and 100 U/mL penicillin and 100 µg/mL streptomycin. One
hour before the experiment, cells where incubated with 5 µM of Fluo4-AM in a HEPES buffered
physiological solution (Live Cell Imaging Solution) supplemented with 5.5 mM of glucose. After
incubation, cells where washed, solution was discarded and replaced with physiological solution
containing 10 µg/mL of propidium iodide. Fluorescence microscopy was performed on a Zeiss LSM
800 inverted confocal microscope using FITC and Texas Red excitation, appropriate dichroic and
filter sets, and analyzed using Image Analyst MKII.
References
1) Khodagholy, D. et al. Highly Conformable Conducting Polymer Electrodes for In Vivo Recordings.
Adv. Mater. 2011, 23 (36), H268–H272.
2) Sessolo, M et al. Easy-to-Fabricate Conducting Polymer Microelectrode Arrays. Adv. Mater. 2013.
25 (15), 2135−2139
3) Someya, T. et al. The rise of plastic bioelectronics. Nature. 2016. 540, 379-385.
4) Li, Y.et al.Electroporation on microchips: the harmful effects of pH changes and scaling down.Sci.
Rep.2015. 5, 17817.
Figures

Figure 1. Devices with interdigitated electrodes. Electrodes are fabricated with different
dimensions and different materials including conductive polymers. In vitro electrodes are
deposited on glass substrates, whereas flexible in vivo devices are parylene based.

669
 Figure 2. Electropulsation of U87 glioblastoma cells in a multiwell device. Top left corner,
multiwell device with interdigited gold electrodes with a 500 µm spacing. Top right corner, U87
glioblastoma cells measured in with Fluo 4-AM and Propidium Iodide seeded in each well.
Bottom, measurement of Fluo 4-AM and Propidium Iodide fluorescence during 30 min. 7
pulses of 300 µs, 60 V where applied at 700 sec.

PB-70 [14:30]
STUDENT PAPER
Cell electroporation enhanced by conductive nanoparticles
Amina Ghorbel1, Lluis M. Mir1 & Tomás García-Sánchez1
1UMR 8203, CNRS, Gustave Roussy, Villejuif, France, 94805
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Amina Ghorbel
In order to optimize the outcome of Electroporation , a novel approach based on the use of
conductive nanoparticles, is proposed to enhance locally the efficacy of the electric field intensity in
the targeted biological sample.
Introduction
Electroporation, defined as the permeabilization of the cell membrane induced by the exposure to
short and intense electric pulses, has been developed and applied in a number of clinical,
preclinical, and medical research applications against cancer. This approach uses an electric field in
order to increase the permeability of the cell membrane, allowing chemicals, drugs, or DNA to be
introduced into the cell. Electro-chemotherapy (ECT) represents the most successful clinical
application of electroporation in cancer treatment, consisting of a combination of high voltage
electric pulses and non-permeant (or a low permeant) anticancer drugs. In order to optimize this
technique, a novel approach based on the use of conductive nanoparticles, is proposed to enhance
locally the efficacy of the electric field intensity in the targeted biological sample. This approach
would lead to a drastic reduction in the external electric field intensity required to achieve effective
cell permebilization and thus its therapeutic effect.

670
Materials and methods
In vitro experiments were performed to confirm the increase of electropermeabilization effect in the
presence of conductive nanoparticles. Electropermeabilization was studied by cloning efficacy tests.
This test was performed using bleomycin, a non-permeant cytotoxic drug extremely active when it
enters the cells by electroporation. Cloning efficacy tests permitted us to detect the differences in
cell permeabilization with high sensitivity.
The first set of experiments was performed on attached DC3F (Chinese Hamster lung fibroblasts).
Cells were cultured in a special chamber fabricated in PDMS (Polydimethylsiloxane). Specially
adapted electrodes were used to expose the entire cell monolayer to electric field pulses (Fig. 1).
Cells were subjected to electroporation with or without nanoparticles and the results were
compared. Different conductive nanoparticles of differents materials were used. Also, different
pulsed electric field exposure conditions were studied using 100 microsecond (µs) at a fixed low
voltage.
The effect of the electric field was studied in combination with bleomycin or alone.
Results
The permeabilization rates calculated as the percentage of non-surviving cells with respect to
controls were compared in treatments with and without conductive nanoparticles.
As preliminary results, it is observed that in the absence of bleomycin, the pulse condition used
produces a negligible cell death that is similar in the presence or absence of nanoparticles. This
result confirms that the contact of cells with the conductive nanoparticles does not produce any toxic
effect.
When blemycin is added, results show that the combination with nanoparticles reduces
systematically the cell survival rates comparing to the situation without conductive nanoparticles for
the different pulses condition assayed.
Conclusions
Cloning efficacy tests have confirmed how the combination of electroporation with conductive
nanoparticles increases the cell electropermeabilization using a low external electric field.
Moreover, experiments have shown that the used nanoparticles did not produce any additional cell
toxicity by themselves or when combined with electric field pulses and that cell death was increased
only when the anticancer drug bleomycin was added.
Figures

Figure 1. Chamber fabricated in PDMS for the exposure of a cell monolayers. Schematic
representation of cells subjected to conductive nanoparticles.

671
PB-72 [14:30]
Cells modeling and design characterization of electropulsation microchambers for linear and
non-linear optical microspectroscopy
Caterina Merla1, 2, Micaela Liberti3, Paolo Marracino3, Adeline Muscat2, Francesca Apollonio3 & Lluis M. Mir2
1ENEA SSPT Division of Health Protection Technologies, Rome, Italy, 00123
2CNRS UMR 8203, Gustave Roussy, Univ. Paris-Sud, Université Paris Saclay, Villejuif, France, 94805
3Department of Information Engineering Electronic and Telecommunications, Sapienza University of Rome ,
Rome, Italy, 00184
Keywords: Electroporation, Pulsed, Work in Progress
Presented by: Caterina Merla
Mechanisms underlining the cell membrane permeabilization and consequent cellular and
intracellular effects are still debated. To shed light on the involvement of phospholipids and water
molecules rearrangements in this phenomenon, linear and non-linear optical imaging and
microspectroscopy seem particularly appropriate. To combine optical detection with electric pulses
used for cell permeabilization, an optimal delivery setup is presented in this paper. The dosimetry of
the fabricated microchamber was analyzed in time domain. Mesodosimetry was also performed to
study electric field distribution at cell level. These data give interesting information for interpreting
cellular microspectroscopic results.
Introduction
Electropulsation technologies, implying the delivery of extremely short (down to hundreds of ps),
high amplitude (up to tens of MV/m) electric pulses to cells and tissues, may represent one of the
most important use of electromagnetic fields in biology and medicine [1]. The primary effects of
these electric pulses lead to the permeabilization of the cell membrane followed by a series of
secondary events that are still under analysis and comprehension. Cell membrane permeabilization
is itself a complex phenomenon. The increase of the cell transmembrane potential caused by the
application of the electric pulses is responsible for phospholipids and water molecules
rearrangements as predicted by molecular dynamic simulations [2]. Therefore, a first hypothesized
model for cell membrane permeabilization implies phospholipids and water molecules mechanical
rearrangement under the action of the electric field.
Nonetheless, other mechanisms can concur to cell permeabilization as chemical reactions
happening at the membrane interface as hypothesized in [3]. Indeed, lipid peroxidation and water
molecule rearrangements can drive permeabilization of cell membrane explaining even long lasting
effects after the ending of electric pulses.
The label free analysis of the vibrational modes of specific chemical bounds could be very useful to
study cell membrane permeabilization. To this end, linear and non-linear optical microspectroscopic
techniques such as Raman and coherent anti-stokes Raman scattering (CARS) are of particular
interest.
Therefore, to perform such analyses during or immediately after the electric pulse delivery
microchamber has been designed and presented in this paper. It fulfills specific requirements in term
of reduced emission to limit electromagnetic compatibility problems in the proximity of the
microscope electronics and acquisition cameras. It is a transparent path to assure the lasers
passage through cells, as well as a 50-ohm matching on a wide band of frequencies, and high
efficiency.
Numerical and experimental dosimetry of this microchamber is detailed in this abstract as well as
microdosimetry to define the electric field distribution around a single cell in a real experimental
configuration. This modeling can give interesting information to assist and interpret single cell
microspectroscopic outcomes.
Design
672
The analyzed setup was based on a coplanar wave guide. To further decrease electromagnetic
radiation, the designed setup presents a connection between the lateral and bottom ground
electrodes as depicted in Fig. 1. This new coplanar configuration has been named ground closed
coplanar waveguide (GCCPW). Analytical dimensioning, as suggested in [4], of the setup was
performed and is reported in Fig. 1.
To contain the cell solution, bio-compatible reservoirs in polydimethylsiloxane (PDMS) have been
designed as sketched in Fig. 1 with dimensions optimized using numerical simulations. The
designed devices use Indium Thin Oxide (ITO) layer to assure electric continuity of the ground
plane, limiting as much as possible radiations while maintaining the transparency of the setup
required for optical signal acquisition.
Dosimetry frequency and time domain approaches
To assess numerical characterization of the proposed microchambers and to choose the optimal
structure for optical imaging, simulations in frequency and time domains were carried out using
HFSS v. 15.
Simulations in the frequency domain were carried out between 100 MHz up to 5 GHz in a sweep
mode. Time domain analyses were conducted using a Lorentzian shaped electric pulse of 200 ps as
a worst case. In frequency domain, radiated electric fields on different planes out (above and below)
of the devices were computed. Maximum electric field levels below and above the device along a
vertical line passing through its center were found to be compliant with European regulation for
protection of laboratory equipment, with maximum radiated field of 2x10-3 V/m at 1 GHz for 1 W of
input power which is one of the lowest level reached so far so such a type of miniaturized exposure
system.
In time domain, the electric field waveforms obtained in the biological solution were analyzed to
verify the avoidance of electric pulse deformations. Looking at the electric field levels reached within
the biological solution and specifically into the visibility windows, the GCCPW resulted to be a very
efficient setup. An efficiency of around 1500 V/mV for the GCCPW is achieved. The efficiency was
evaluated as the ratio between the induced field level at 100 ps and the input voltage at the same
instant. This result means that to reach high field intensities required for cell permeabilization, the
use of the GCCPW setup is much more convenient with respect to other systems like slot lines or
edge coupled microstrip lines. Therefore, the GCCPW represents the best compromise in term of
matching behavior, safety, and exposure efficiency.
For these reasons, this setup was fabricated and experimentally characterized in time domain to
effectively assess transmission behaviors using a time domain reflectometry module (SDA 100 G Le
Croy). The fabricated GCCPW was connected in a two port configuration. At port one, a voltage
ramp signals with rise times of 30, 100, and 200 ps were applied and the transmitted signals trough
the GCCPW were acquired at port two.
Delays in rise times of the transmitted pulses were measured as well as the attenuation in the
transmitted voltage to quantify signal distortion and losses through the fabricated GCCPW.
Measurements were performed with PDMS reservoir filled with a conductive cell buffer named
DMEM (measured volume 30µl) and non-conductive distilled water. Measured signals for DMEM are
presented in Fig. 2(A), showing limited delay of the time domain signal only at 200 ps in comparison
with ideal thru connections. Attenuation in the mentioned condition is rather low and it never
exceeds 20%, also for water (data not shown). These results demonstrate a good transmission
behavior of the fabricated setup at 200 ps, even in the presence of the connectors not accounted in
the simulations.
In Fig. 2(b) the GCCPW transfer functions, for DMEM and water filling the PDMS reservoir, are
shown indicating a GCCPW pass band of few GHz.
Time domain mesodosimetry
To assess electric field distribution down to the single cell level, a microdosimetric analysis was

673
carried out. A realistic cell topology was obtained staining human mesenchymal stem cells (HaMSC)
with the fluorescent dye YOPRO-1 after cell permeabilization using the GCCPW electrodes.
Fluorescence was acquired using an inverted microscope AxioVert 100, with a 20× objective.
Fluorescence images were made binary using Image J software and traced using the “find edge”
command following a procedure similar to that detailed in [5]. Transformed images were imported in
COMSOL v. 5 for simulations. Simulations were performed in time domain using a 10 ns long
electric pulse with amplitude of 10 MV/m. Cell electrical parameters were set as in [6], [7].
Electric field distribution and current densities at the middle of the pulse are shown in Fig. 3. Indeed,
inhomogeneity of electric field is rather evident leading to variations in the exposure of a single cell.
Clearly, electric field distribution depends on the specific cell topology making extremely pertinent
this characterization in the microchambers when single cell measurements (e.g. fluorescence,
Raman and CARS spectra and imaging) have to be performed to study electropermeabilization.
Conclusions
In this abstract, the numerical and experimental characterization of a GCCPW for real-time optical
spectroscopy of cells to study permabilization is reported. An optimal device in terms of emission,
efficiency, and matching band was defined and realized. Numerical results fit well with experimental
measurements. To evaluate electric field distribution down to the single cell, a first microdosimetric
analysis was performed on a realistic cell topology. These electric distributions are helpful to assist
real-time single cell analyses using sophisticated optical micro-spectroscopic techniques. Future
extension of the model will include quantification of electropermeabilization to obtain a one to one
matching with experiments.
Acknowledgments
This project has received funding from European Union’s Horizon 2020 Research and Innovation
Program under the Marie Sklodowska-Curie Action grant agreement No. 661041 OPTIC BIOEM.
Work performed also in the frame of the LEA (European Associated Laboratory) EBAM (Pulsed
Electric Field Applications in Biology and Medicine).
References
[1] D. Miklavcic, Handbook of Electroporation, Springer, 2017.
[2] M. Tarek, “Membrane electroporation: a molecular dynamics simulation”, Biophys. J., vol. 88, no.
6, pp. 4045-4053, 2005.
[3] M. Breton and L. M. Mir, “Investigation of the chemical mechanisms involved in the
electropulsation of membranes at molecular level”, Bioelectrochemistry, vol. 119, pp. 76-83, 2018.
[4] A. Balanis Antenna Theory: analysis and design, 4th Ed. Wiley, 2015.
[5] A. Denzi et al., “A microdosimetry study of electropulsation on multiple realistically shaped cells:
effect of neighbours”, J. Memb. Biol.,vol. 249, no. 5, pp. 691-701, 2016.
[6] C. Merla, A. G. Pachomov, I. Semenov, P. T. Vernier, “Frequency spectrum of induced
transmembrane potential and permeabilization efficacy of bipolar electric pulses”, BBA
Biomembranes, vol. 1859, no. 7, pp. 1282-1290, 2017.
[7] Denzi et al., “Assessment of cytoplasm conductivity by nanosecond pulsed electric fields”, IEEE
Trans. Biomed. Eng., vol. 62, no. 6, pp. 1595-1603, 201
Figures

674
Figure 1. GCCPW geometry and dimensioning on a frontal plane (xz). Sketched device is not
in scale.

Figure 2. Transmission measurements for GCCPW loaded with DMEM (a) and the
corresponding computed transfer function for DMEM and water (b).

675
 Figure 3. Microdosimetry of MSC exposed to 10 ns electric pulse at 6 ns.

PB-74 [14:30]
Time-domain finite-element modeling for analysis of in vivo electroporation-based treatments
Matevz Pintar1, Janez Langus1, Matej Kranjc2, Erik Brecelj3, Ibrahim Edhemovic3, Gregor Serša3, Tomaz
Rodic1, Tomaz Sustar1, Damijan Miklavčič2, Tadej Kotnik2 & Bor Kos2
1c3m d.o.o., Ljubljana, Slovenia, 1000
2University of Ljubljana, Ljubljana, Slovenia, 1000
3Institute of Oncology Ljubljana, Ljubljana, Slovenia, 1000
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Bor Kos
Electrochemotherapy and irreversible electroporation are two methods which rely on the mechanism
of electroporation to achieve their clinical goal. For successful treatment, coverage of the target
tissue with sufficiently strong electric field is the most important prerequisite. Here we present a
time-dependent model of electroporation, which takes into account the complex time course of
conductivity change due to tissue electroporation and apply the model to a case of in vivo
electroporation treatment.

676
Introduction
Electroporation is a physical method of enhancing transport of molecules across the cellular
membranes, which occurs when high voltage electric pulses are delivered to cells [1], [2]. The most
common explanation of the mechanism of electroporation is the formation of pores in the cell
membrane lipid bilayer [3]. Electrochemotherapy is a method of cancer treatment which involves
using poorly permeant cytotoxic drugs, such as bleomycin or cis-platin combined with
electroporation to achieve a large potentiation of their effect [4], [5].
We have previously described a time-dependent model of electroporation, which was initially tested
on ex vivo samples of beef liver. This model is able to predict the time evolution of an electric pulse
current within a 5% error for a simple, well defined geometric case of two parallel needle electrodes.
Here, we apply this model on one of the first cases of electrochemotherapy treatments of colorectal
carcinoma metastasis in the liver treated at the Institute of Oncology Ljubljana [6].
Methods
In vivo procedure
The treated patient was recruited in the scope of a phase I/II clinical study (EudraCT number
2008-008290-54; ClinicalTrials.gov (NCT01264952)), approved by the Institutional Medical Board
and the Medical Ethics Committee of the Republic of Slovenia. The patient was a 55 year old
female, presenting with a single non-resectable metastasis in contact with the inferior vena cava,
middle hepatic vein, and right hepatic vein.
Before the treatment, numerical treatment planning was performed using Comsol Multiphysics
version 3.5a (Comsol AB, Stockholm, Sweden). Pre-treatment MRI images were segmented into
tumor, vessels and liver and reconstructed into a 3D geometry in Comsol Multiphysics (Comsol AB,
Stockholm, Sweden). A six electrode configuration was chosen during the treatment planning step
as shown in Figure 1. During the treatment ECG was recorded using Biopac MP150 with two
ECG100C modules (BIOPAC Systems Inc, Goleta, CA) to assess any potential adverse effects on
the functioning of the heart.
Numerical model
The computational mesh for simulation by finite element method (FEM) was created in Comsol
Multiphysics software and exported to a text file. It contained 57464 second-order tetrahedral
elements. The conductivity model of the time-dependent model is based on three parameters: : -
level of poration increases as a function of the amplitude of electric field and decreases
exponentially during the pulses; - poration damage indicator models the growth of pores during the
pulse application and - thermal damage indicator models damage effects due to Ohmic tissue
heating, as shown in equation 1.

E 1
The model was implemented in Wolfram Mathematica (ver. 11.2, Wolfram Research, IL) using the
AceGen and AceFEM packages [7] (ver. 6.819).
For the boundary conditions, the actual measurements of delivered electrical pulses as delivered by
the pulse generator Cliniporator Vitage (IGEA, Carpi, Italy) were used. A total of 11 sequences of
pulses were delivered as shown in Table 1. Some voltages were lower than the original treatment
plan due to excessive current being delivered, which resulted in the pulse generator’s current
protection triggering. The precise timing of the pulses used in the simulations was determined from
the ECG recording, to the precision better than 50 ms. Parameter values for the electroporation
model and tissue properties were taken from the literature [8], [9].
Table 1. Sequences of electric pulses

677
 Sequence Electrode pair No. of pulses Voltage [V]

1 2-6 8 2104

2 1-6 8 2109

3 3-5 8 2104

4 4-5 8 2104

5 5-6 8 1704

6 3-6 8 1896

7 2-5 8 1704

8 4-6 8 1702

9 2-5 15 1512

10 4-6 8 1519

11 1-5 16 1329

Results
The total duration of the treatment was 852 seconds, due to the reductions of voltage to certain
electrode pairs, which introduced some delay between the pulses. Figure 2 shows three selected
sequences. In panel a), the different mechanisms of current increase can be seen, since the current
gradually increases from 46 A at the end of the first simulated pulse, to 47 A at the end of the last
pulse. Panel b) shows an example where the measured current is decreasing with time, possibly
due to electric-pulse induced movement of the patient. In panel c), an example of the measured
current being capped at 49.5 A can be seen, while the simulated current goes up to 54 A. The
RMSE values were determined for each sequence separately, and were between 1.9 A and 11.5 A.
Table 2: Comparison of simulated and measured electric currents

Average
Average Stationary
Sequence Electrode pair measurements RMSE [A]
simulation [A] simulation [A]
[A]

1 2-6 48.6 46.7 1.9 48.3

2 1-6 46.2 50.5 4.3 52.3

3 3-5 48.8 41.8 7 44.3

678
 4 4-5 48.3 40 8.4 42.0

5 5-6 48.8 59.2 10.4 62.1

6 3-6 49.1 53.2 4.1 53.7

7 2-5 49.1 53 3.9 52.5

8 4-6 49.1 38.1 11.1 39.8

9 2-5 45.5 46 3.7 45.6

10 4-6 41.3 32.8 8.5 34.1

11 1-5 32.6 44 11.5 42.3

Discussion
The presented application of the time-domain model of electroporation to an in vivo case shows
promising results. Tissue conductivity values and other model parameters were taken directly from
previously published literature and have not been adjusted in order to achieve better agreement of
simulated and measured current [8]–[12].
One limitation of the present study is the levelling-off of measured current seen Figure 2 c, which
can be explained by the maximum current capability of the pulse generator, which is 50 A. It is
possible that the actual current was above the total measurement range, but the overcurrent
protection was not triggered. Another limitation is the positioning of the electrodes during the
treatment. The distances were measured, but complete parallelism was difficult to achieve due to
limited space in the surgical opening.
Recent advances in theoretical understanding of time dependent tissue modeling have led to the
development of an alternative dynamic numerical model with great agreement between measured
and simulated current of a single electroporation pulse[13].
Conclusions
The model shows a good agreement with the measured currents and allows the accurate prediction
of currents delivered during electroporation treatments as well as the determination of the time
course of the electric field and the degree of electroporation in the tissue.
References
[1] E. Neumann and K. Rosenheck, “Permeability changes induced by electric impulses in vesicular
membranes,” J. Membr. Biol., vol. 10, pp. 279–290, 1972.
[2] T. Kotnik, P. Kramar, G. Pucihar, D. Miklavcic, and M. Tarek, “Cell Membrane Electroporation-Part
1: The Phenomenon,” Ieee Electr. Insul. Mag., vol. 28, no. 5, pp. 14–23, Oct. 2012.
[3] J. C. Weaver, “Electroporation in cells and tissues: A biophysical phenomenon due to
electromagnetic fields,” Radio Sci., vol. 30, no. 1, pp. 205–221, Jan. 1995.
[4] L. M. Mir, S. Orlowski, J. Belehradek, and C. Paoletti, “Electrochemotherapy Potentiation of
Antitumor Effect of Bleomycin by Local Electric Pulses,” Eur. J. Cancer, vol. 27, no. 1, pp. 68–72,
1991.
[5] G. Sersa, M. Cemazar, and D. Miklavcic, “Antitumor Effectiveness of Electrochemotherapy with

679
Cis-Diamminedichloroplatinum(ii) in Mice,” Cancer Res., vol. 55, no. 15, pp. 3450–3455, Aug. 1995.
[6] I. Edhemovic et al., “Electrochemotherapy: a new technological approach in treatment of
metastases in the liver,” Technol. Cancer Res. Treat., vol. 10, no. 5, pp. 475–485, Oct. 2011.
[7] J. Korelc and P. Wriggers, Automation of Finite Element Methods. Cham: Springer International
Publishing, 2016.
[8] B. Kos, P. Voigt, D. Miklavcic, and M. Moche, “Careful treatment planning enables safe ablation
of liver tumors adjacent to major blood vessels by percutaneous irreversible electroporation (IRE),”
Radiol. Oncol., vol. 49, no. 3, pp. 234–241, Sep. 2015.
[9] J. Langus, M. Kranjc, B. Kos, T. Suštar, and D. Miklavčič, “Dynamic finite-element model for
efficient modelling of electric currents in electroporated tissue,” Sci. Rep., vol. 6, p. 26409, May
2016.
[10] M. Marčan, B. Kos, and D. Miklavčič, “Effect of blood vessel segmentation on the outcome of
electroporation-based treatments of liver tumors,” PloS One, vol. 10, no. 5, p. e0125591, 2015.
[11] D. Sel, D. Cukjati, D. Batiuskaite, T. Slivnik, L. M. Mir, and D. Miklavcic, “Sequential finite
element model of tissue electropermeabilization,” IEEE Trans. Biomed. Eng., vol. 52, no. 5, pp.
816–827, 2005.
[12] A. Zupanic, S. Corovic, and D. Miklavcic, “Optimization of electrode position and electric pulse
amplitude in electrochemotherapy,” Radiol. Oncol., vol. 42, no. 2, pp. 93–101, Jun. 2008.
[13] D. Voyer, A. Silve, L. M. Mir, R. Scorretti, and C. Poignard, “Dynamical modeling of tissue
electroporation,” Bioelectrochemistry, vol. 119, no. Supplement C, pp. 98–110, Feb. 2018.

Figures

Figure 1. a) Original treatment plan showing the overlay of MR image and the Comsol
Multiphysics model. b) Positioning of the electrodes during the treatment.

680
Figure 2. Comparison of measured and simulated electric current for pulse sequences. The
time scale is not regular and does not show the long pauses between pulses. Only
approximate time at which the pulse was fired is indicated. a) Sequence 1 – a good agreement
between the measurement and simulation is seen. b) Sequence 2 – the measured current is
decreasing during the delivery, possibly due to patient movement. c) Sequence 6 – cut-off of
the measured current is seen.

681
PB-76 [14:30]
Calcium ions with nanosecond pulsed electric fields in human colon adenocarcinoma cells
with drug resistance
Nina Rembiałkowska1, 2, Anna Choromańska1, Anna Szewczyk3, Olga Michel1, Jolanta Saczko1, 2 & Julita
Kulbacka1, 2
1Department of Medical Biochemistry, Medical University, Chalubinskiego 10 St., 50-368 Wroclaw, Poland
2Department of Molecular and Cellular Biology, Medical University, Borowska 211 A, 50-556 Wroclaw, Poland
3Department of General Zoology, Zoological Institute, University of Wroclaw, Sienkiewicza 21 St., 50-335
Wroclaw, Poland
Keywords: Electroporation, Pulsed, Completed (unpublished)
Presented by: Nina Rembiałkowska
Calcium is one of the most important factors in cell proliferation, differentiation and cell death
(apoptosis or necrosis). Manipulation of calcium levels using electroporation can cause different
effect on normal and malignant cells. Efficiency and safety of electroporation combined with Ca2+
ions was examined in two human colon adenocarcinoma cell lines: sensitive- LoVo and drug
resistant-LoVo/DX. As normal control CHO-K1 cells were applied. The results show that nsPEF
supported by Ca2+ is cytotoxic in particular for resistant LoVo/DX cells, less cytotoxic in case of
LoVo cells and simultaneously safe for normal cells. These results may potentially be important for
treatment of colon adenocarcinoma using calcium nsPEF.
1. Introduction
Colon or colorectal cancers are mainly caused by the unhealthy lifestyle, aging and, in some cases,
genetic factors [1,2]. This type of cancer is the third-leading cause of cancer-related deaths in the
USA [3].
Electroporation supports various types of anticancer therapies by increasing the selective transport
of cytostatics [4]. The application of nanosecond pulsed electric fields (nsPEFs) is a quite novel
method to affect and inhibit cancer cells’ proliferation process. NsPEFs can directly act on the all cell
membranes and activate cell death pathways in various cell lines. The increase of intracellular
calcium level by nsPEF may be a new approach to the treatment of numerous types of cancer.
Calcium ions are necessary in numerous physiological processes like regulation of genes
transcription metabolism, induction of proliferation and differentiation. Calcium is also involved in the
initiation of cell death by apoptosis or necrosis [5].
A therapy based on electroporation and calcium ions could have either negative or positive impact
on normal and malignant cells. The effect of nsPEF and nsPEF combined with Ca2+ on cell survival
has been studied in vitro.
2. Material and methods
Two colon adenocarcinoma cell lines were used: LoVo – sensitive and LoVo/DX – doxorubicin-
resistant cell line. Additionally, ovarian fibroblasts (CHO-K1) were used as normal control.
Nanosecond pulses delivery protocol
Nanosecond pulses were generated by PPG-20 generator (FID Technology, Germany) and
measured with oscilloscope TDS3054 (DPO, 500MHz, 5 GS/S, 4 channels, Tektronix, Tespol,
Poland) with high-voltage probe (75 MHz, 40 kV, 1000X 10 FT; P6015A, Tektronix, Tespol, Poland).
The following electrical field parameters were used: electric field intensity – nsPEF=12.5 – 50
kV/cm, repetition frequency - 100 Hz, number of pulses – 200, pulse duration 10 ns with time rise of
2 ns. For viability experiments, cells were suspended in the electroporation buffer (STM buffer) with
low electrical conductivity of 0.12 S/m (10 mM KH2PO4/K2HPO4, 1 mM MgCl2, 250 mM sucrose, pH
7.4), SMEM (Sigma) with medium conductivity (1.5 S/m). Then cells were pulsed in a 4 mm cuvette
with two aluminum plate electrodes (BTX, Syngen, Poland). According to the manufacturer’s
description the field type in used cuvettes is homogeneous. After electroporation, the cells in

682
cuvettes were left for 10 min at 37˚C, centrifuged 4 min at 537×g, and resuspended for further
analysis in proper cell culture medium.
Viability assay
The3(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT, Sigma-Aldrich, Poland) test
was used for the assessment of cell viability. After experiments cells were seeded into 96-well
microtiter plates at the concentration of 5 x 103 cells/well. The MTT assay was performed according
to the manufacturer’s protocol after 24 and 72 h. The absorbance was determined using a multiwell
scanning spectrophotometer at 570 nm (EnSpire Perkin Elmer, Poland). Mitochondrial
oxidoreductive function was expressed as a percentage of viable treated cells in relation to the
untreated control.
Cytoskeleton organization
Cytoskeleton reorganization was determined by immunofluorescent cells’ labeling of F-actin with
Alexa Fluor®546 Phalloidin (Thermo Fisher) and pan-Cadherin (Abcam) with Atto®488 (Thermo
Fisher). Cells were subjected to the nanosecond pulses delivery protocol with/without calcium ions
and then immunoassayed after 24 and 72 h. For imaging, Olympus FluoView FV1000 confocal laser
scanning microscope (Olympus) was used.
GSH/GSSG assay
After experiments, cells were seeded into white 96-well microtiter plates at the concentration of 5 ×
103 cells/well. The level of reduced and oxidized glutathione was determined after 24 h via
luminescence-based assay using a multiwell scanning spectrophotometer at 570 nm (EnSpire
Perkin Elmer, Poland).
Proteasomal activity
Cells were seeded into white 96-well microtiter plates at the concentration of 5 × 103 cells/well. The
activity of chymotrypsin-like, trypsin-like or caspase-like protease associated with the proteasome
complex in cultured cells were determined after 24 h by a luminescence-based assay using a
multiwell scanning spectrophotometer at 570 nm (EnSpire Perkin Elmer, Poland). These three
activities are responsible for much of the protein degradation required to maintain cellular
homeostasis, including degradation of critical cell-cycle proteins, tumor suppressors, transcription
factors, inhibitory proteins and damaged cellular proteins [6].
3. Results
The research confirmed anti-cancer properties of combining the nsPEF with Ca2+. The most
cytotoxic effect was observed in case of LoVo/DX cells. The viability decreased ca. 70% after 24h
and 85% after 72h using 50 kV/cm and 1mM of CaCl2. The same parameters caused less cytotoxic
effect in normal CHO-K1 cells and minimal viability decrease in LoVo cells.
The visualization of cytoskeleton proteins revealed that LoVo/DX cells’ morphology was the most
affected after 24h (Fig.1) and 72 h. nsPEF-Ca therapy reorganized phalloidin fibers and inhibited
pan-Cadherin expression.
Of nsPEF method led to alterations in the cell morphology, redox potential of the GSSG/GSH, and
induced protein degradation by an imbalance of proteasomal activity. These results confirm that
calcium ions supported by EP can be an effective anti-cancer therapy for the colon
adenocarcinoma, whilst being relatively safe for normal cells. Our studies confirm previous reports
that calcium-nsPEF is preferentially toxic to malignant cells and paves the way for further
investigations of the use of calcium-EP in colon cancer treatment.
Concluding, the obtained results may suggest a promising anticancer potential of nsPEF combined
with calcium ions against colon cancer,in particular against resistant type.
Financing: The research was supported by the grant of National Science Centre, Poland 2016/22
/E/NZ5/00671 (SONB.A040.17.001; PI J.Kulbacka).

683
4. References
[1] Bibbins-Domingo, K. et al. Screening for Colorectal Cancer: US Preventive Services Task Force
Recommendation Statement. JAMA. 315, 2564–75 (2016).
[2] Astin, M., Griffin, T., Neal, R. D., Rose, P. & Hamilton, W. The diagnostic value of symptoms for
colorectal cancer in primary care: a systematic review. Br. J. Gen. Prac. 61, 231–43 (2011).
[3] Siegel, R., Desantis, C. & Jemal, A. Colorectal cancer statistics. C.A. Cancer J. Clin. 64, 104–117
(2014).
[4] S.Orlowski et.al., Transient electropermeabilization of cells in culture. Increase of the cytotoxicity
of anticancer drugs, Biochem Pharmacol, (1988), 4727-33.
[5] B. Zhivotovsky et.al., Calcium and cell death mechanisms: A perspective from the cell death
community, Cell Calcium, (2011), 211-21.
[6] J. Adams, Development of the proteasome inhibitor PS-341. Oncologist 7, (2002) 9–16.
Figures

Figure 1. Cytoskeleton protein evaluation after 24h in CHO-K1 cells (upper panel); LoVo
(middle panel) and LoVo/DX cells (bottom panel). F-actin-red; pan-cadherin – green, nuclei –
blue.

684
PB-78 [14:30]
Does RF exposure increase delivery of lead from maternal blood to fetus?
Yong-Han Lee1, Eunae Burm2, Jong Hyuk Choi1, Hee-Sun Yang1, Sanghyuk Bae1, Ho-Jang Kwon1, Eun-
Hee Ha3, Hyesook Park4, Yangho Kim5, Yun-Chul Hong6, Ae-kyoung Lee7, Hyung-Do Choi7 & Mina Ha8
1Department of Preventive Medicine, Dankook University, Cheonan, Korea
2Department of Nursing, Moonkyung College, Moonkyung, Korea
3Department of Occupational and Environmental Medicine, Ewha Womans University, Seoul, Korea
4Department of Preventive Medicine, Ewha Womans University, Seoul, Korea
5Department of Occupational and Environmental Medicine, Ulsan University, Ulsan, Korea
6Department of Preventive Medicine, Seoul National University, Seoul, Korea
7Radio Technology Research Department, Electronics and Telecommunications Research Institute, Daejeon,
Korea
8Environmental Health Policy Division, Ministry of Environment, Sejong, Korea
Keywords: Epidemiology, RF/Microwaves, Work in Progress
Presented by: Yong-Han Lee
We measured individually maternal RF exposure during pregnancy and lead concentration of
umbilical cord blood and maternal blood at delivery. The more RF exposure during pregnancy, the
greater likelihood that more lead was transferred from mother to fetus through the placenta. This
result suggest that RF affects function of placenta or causes conditions that more lead deliver to
fetus. In addition, RF exposure more affected pregnant women with null type of GSTT1 gene
involved in metabolism of toxic substances. This finding supports that RF exposure during
pregnancy can be hazardous in women vulnerable to toxic materials.
Introduction
Some hazardous environmental factors transfer placenta to fetus and affect pregnancy outcomes
while the effects of maternal RF exposure on the fetus are not well known. We aimed to examine
whether maternal RF exposure during pregnancy has an association with maternal-fetal transfer of
lead.
Method
We used the data of Mothers and Children’s Environmental Health (MOCEH), a multicenter
prospective birth cohort study. From 2011 through 2014, we enrolled 1751 pregnant women in South
Korea. Of those, we analyzed 131 women who voluntarily underwent RF exposure measurement
and blood samplings in pregnancy. RF exposure was measured via personal exposure meter (EME-
Spy100®) for 24 hours at enrollment within 20 weeks of pregnancy. RF exposure index of each
woman was calculated as sum of square of arithmetic mean for measured levels over LOD
(>0.05V/m) divided by guidance level of each frequency band. Concentrations of lead were
measured in maternal blood and umbilical cord blood at the end of pregnancy. We examined the
effect of RF in transfer of lead from mother to fetus by using a linear regression model with a
dependent variable of level of lead in cord blood (with maternal level of those as a covariate) in
association with RF exposure. We considered potential confounding factors or covariates such as
maternal age, BMI, smoking, drinking alcohol, income, education, past medical history, pregnancy
period, child's sex, gestational age, birth weight and maternal genotypes in GSTT1 and GSTM1. We
also examined the association after stratified by the maternal genotypes, which are known to be
related with the metabolism of toxic substances. The generalized additive model (GAM) to examine
a non-linear association was applied.
Result
When total RF exposure doubled, the concentration of umbilical cord blood lead (adjusted maternal
blood lead) increased by 12.2% in pregnant women who was null type of GSTT1. When RF
685
exposure by mobile phone doubled, the concentration of umbilical cord blood lead increased by
11.7% in them and the conditions. The associations were statistically significant. When RF exposure
by except mobile phone doubled, the concentration increased by 2.8% in them and the conditions.
But the association was not statistically significant. And the result of GAM (Figure 1) was also
showed a positive relationship between RF exposure and transfer of lead from mother to fetus in
women who was null type of GSTT1.
Discussion
In the women who was null type of GSTT1 gene Involved in metabolism of toxic materials, the
association was stronger between RF exposure during pregnancy and transfer of lead from mother
to fetus. It suggests that RF exposure during pregnancy is more hazardous to women vulnerable to
toxic materials and their children. But in women who was null type of GSTM1 gene playing a similar
role to GSTT1, no significant association has been found. With more sophisticated personal RF
exposure measurement, further research should be carried out on a larger number of pregnant
women. In addition, further studies are needed to elucidate mechanisms and replicate the present
findings.
Conclusion
Association between RF exposure by phone use during pregnancy and maternal-fetal transfer of
lead was significant. In the mothers vulnerable to toxic materials, RF exposure can be more
hazardous to their children.
Figures

Figure 1. The results of the generalized additive model (GAM) in women null type of GSTT1
gene

PB-80 [14:30]
Electroconvulsive therapy and amyotrophic lateral sclerosis
Gabor Mezei1, Edmund Lau1 & Leeka Kheifets2
1Health Sciences, Exponent, Menlo Park, California, USA, 94025
2Department of Epidemiology, University of California, Los Angeles, Los Angeles, California, USA, 90095
Keywords: Epidemiology, ELF/LF, Work in Progress

686
Presented by: Gabor Mezei
To evaluate a potential association between exposure to electric shocks and development of
amyotrophic lateral sclerosis (ALS), we conducted a cohort study using a systematic sample of
approximately one million beneficiaries enrolled in the U.S. Medicare health insurance program from
1997 to 2015. Using time-varying proportional hazards modelling, we compared ALS occurrence
among patients receiving electroconvulsive therapy (ECT), a source of exposure to electric currents
in the brain without significant exposure to magnetic fields, given as a treatment for some psychiatric
conditions, to ALS occurrence among patients receiving no ECT, but diagnosed with the same
psychiatric conditions.
Introduction
Amyotrophic lateral sclerosis (ALS) has been associated with electrical environments and electrical
occupations in occupational epidemiologic studies. Occupational exposure to extremely low
frequency magnetic fields (ELF-MF) does not appear to explain this association. It has been
suggested that exposure to electric shocks (ES) in these electrical environments and occupations
rather than exposure to ELF-MF may be the causal factor. In occupational settings, it is frequently
difficult to separate exposure scenarios for ES and ELF-MF as they often occur together, making an
independent assessment of the effects of the two exposures challenging. While several
occupational epidemiologic studies have attempted to use separate job-exposure matrices for ELF-
MF and ES occurring during work, the possibility of confounding by the other exposure remains in
the individual studies and may explain the inconsistencies in the observed results. Electroconvulsive
therapy (ECT), a relatively commonly used treatment for some psychiatric conditions, presents a
setting where a well-documented ES occurs in a medical setting without substantial exposure to
ELF-MF. This provides a unique opportunity to examine the potential role of ES in ALS development
in a controlled and well-documented clinical environment without the confounding effect of ELF-MF.
The purpose of our analyses was to assess whether the incidence of ALS among ECT recipients
differs from the incidence of ALS among those who did not receive ECT.
Methods
We conducted cohort analyses using a 5% systematic sample of beneficiaries in the U.S. Medicare
health insurance program from 1997 to 2015. The Medicare program has a nationwide and
practically universal coverage for both inpatient and outpatient services for elderly U.S. residents
over 65 years of age. People with certain disabilities and medical conditions under 65 may also be
eligible for Medicare. The Medicare program covers approximately 50 million beneficiaries annually.
For each year, the Medicare 5% systematic sample data files include de-identified information on
both inpatient and outpatient claims from hospitals and nursing facilities, as well as professional
service claims from physicians regardless of the setting where service is rendered. Because ECT is
primarily used as a treatment modality for certain psychiatric conditions, our cohort was restricted to
patients who were diagnosed with schizophrenic disorders, episodic mood disorders, or other
depressive disorders. Each subject entered the cohort at the of first mention of any of these
psychiatric diagnoses, and were followed up until diagnosis of ALS, death from any reason, end of
enrollment, or end of 2015, whichever occurred first. We used time-varying proportional hazards
modelling with yearly intervals to assess ALS risk in association with ECT with adjustments for
competing risk due to death. The following covariates were available for assessment and potential
inclusion in our models: age, sex, race, geographic region of residence, health status indicators
(Charlson Morbidity Index), year of birth, year of entry into the cohort, number of emergency room
and physician visits, number of hospital stays, Medicare-Medicaid dual eligibility (representing lower
socioeconomic status), and whether the patient was eligible for Medicare because of end-stage
renal disease. Age was the primary time axis in our proportional hazards models. At a given interval,
the cumulative number of ECT treatments was used as the measure of ES exposure. Members of
the cohort were considered unexposed prior to the receipt of their first ECT treatment, and were
considered exposed following their first ECT. For an evaluation of a potential exposure-response
relationship, the number of ECT treatments were categorized into three categories: no exposure (no

687
ECT), low exposure (1-10 ECT treatments), and high exposure (11 or more ECT treatments). For
our analyses, we used de-identified data files released and approved for research purposes by the
Centers for Medicare and Medicaid Services; no additional ethics approval was required.
Results
A total of 1,019,610 subjects with psychiatric conditions were included in our cohort representing
approximately 21% of the approximately 4.8 million beneficiaries ever included in the 5% systematic
sample of Medicare files from 1997 to 2015. A larger fraction of the study subjects (9.7%) entered
the cohort in 1997, the first year of the study, than in any subsequent years (range from 4.3% to
5.9%). About 84% of the subjects were white, 64% of the subjects were women, and 65% of the
subjects were age 65 or older. Of the included study subjects, 6% were diagnosed with
schizophrenic disorders, 33% were diagnosed with episodic mood disorders, and 61% were
diagnosed with other depressive disorders. Of the cohort members, 7,245 subjects (0.7%) received
ECT treatment, and 1,861 subjects (0.2%) were diagnosed with ALS. As anticipated, ALS incidence
increased with age, varied between geographic regions, and ALS was more common among men
(adjusted hazard ratio [aHR] for women vs. men, 0.69; 95% confidence interval [CI], 0.63-0.76) and
whites (aHR for blacks vs. whites, 0.70; 95% CI, 0.59-0.85; and aHR for others vs. whites, 0.90;
95% CI, 0.72-1.12). ECT specific aHRs varied slightly depending on the specific covariate
adjustments in the models, but confidence intervals included 1 in all analyses. A more detailed
assessment of the association between ECT and ALS occurrence, with due consideration of
potentially confounding variables, will be presented at the conference.
Discussion
To our knowledge, this is the first study to explore the potential relationship between ECT, a source
of electric currents in the brain without significant exposure to magnetic fields, and subsequent
development of ALS. The strengths of our study include its cohort design, the large sample size
(approximately one million study subjects), the nationwide coverage of the study, the length of follow
up (almost 20 years), the well-documented nature of a specific form of ES in a medical setting, and
the virtually complete ascertainment of ALS diagnoses among the cohort members. Limitations of
our study include the relatively small fraction of subjects with ECT (0.7%) despite the large cohort
size, the lack of information on other types of ES, including occupational exposure to ES,
experienced by the study subjects, the lack of information on treatment, including ECT, received
prior to Medicare enrollment by subjects with pre-existing psychiatric conditions at enrollment, and
the fact that ECT is employed in a highly selective population of psychiatric patients, which may limit
the generalizability of our findings.

PB-82 [14:30]
Modelling the exposure of children to extremely low frequency magnetic fields in the
ELFSTAT project
Marta Bonato1, Marta Parazzini1, Emma Chiaramello1, Serena Fiocchi1, Laurent Le Brusquet2, Isabelle
Magne3, Martine Souques3, Martin Röösli4 & Paolo Ravazzani1
1Istituto di Elettronica e di Ingegneria dell’Informazione e delle Telecomunicazioni , Consiglio Nazionale delle
Ricerche, Milan, Italy
2Département Signal et Statistiques, Centrale Supélec , Paris, France
3Medical Studies Department of EDF , Electricite de France, Paris, France
4Epidemiology and Public Health, Swiss Tropical and Public Health Institute, Basel, Switzerland
Keywords: Human, ELF/LF, Completed (published)
Presented by: Paolo Ravazzani
ELFSTAT project, founded by the French ANSES (2015-2019, Grant agreement n. 2015/1/202),
aims at characterizing children’s exposure to extremely low frequency magnetic fields (ELF-MF) in
real exposure scenarios using stochastic approaches. In this paper, a stochastic approach for

688
extracting information from dataset of recorded ELF-MF signals is presented. The aim is to obtain a
better characterization and description of the phenomenon and to investigate on possible
correlations or different features between various data subsets
Introduction
Children exposure to extremely low frequency magnetic fields (ELF-MF, 40-800 Hz) is a topic of high
interest, due to the possible correlation between the exposure and the onset of children leukaemia
[1] [2]. Measurement campaigns have been therefore carried out to investigate on the real level of
children exposure in real life, particularly in Europe [3] [4] [5]. The aim of this paper is to characterize
the children’s exposure to ELF-MF in real exposure scenarios using a stochastic approach based on
segmentation starting from personal measurements of children exposure [6].
Methods
In this study, two databases of ELF magnetic field children personal exposure measurements have
been considered and summed together: the first is the dataset of the ARIMMORA project [3] [4],
which includes a total of 331 children registrations in Italy and Switzerland during winter and
summer seasons and the second one is the dataset of the EXPERS project [5], for a total of 977
children registrations in France during cold season. For each subject, full days of personal
measurements were extracted (from 00:00 to 24:00), which means 2880 values per recording (the
time step is 30 s). From the two datasets, the complete days of recorded registrations are 682 for
ARIMMORA and 767 for EXPERS, for a total of 1449 full days.
ELF-MF time series were characterized by abrupt changes, like sudden jumps of the value of the
mean or in dispersion around the mean. Therefore, the first step of our approach consisted in
detecting these change-points and dividing the 24 h personal measurements in segments between
two consecutive jumps. There are different approaches to detect change-points, like Binary
Segmentation [7], Circular Binary Segmentation [8], Wild Binary Segmentation [9] and Pruned Exact
Linear Time method (PELT) [10] [11] [12]. In this work, the PELT method was used, due its ability to
estimate the number of change-points, their locations and the Auto-Regressive (AR) model
parameter of each segment. After the process of segmentation, the number of change-points
detected for the whole dataset was of 11510 points in total. The second step of our approach
consisted in modeling every obtained segment between 2 jumps with a stationary process, in
particular with an AR model. At the end, every segment was described by 4 parameters: i) the mean
of the stationary process; ii) the variance of the stationary process; iii) the Auto-Regressive
coefficient of the AR process of order 1; iv) the duration of the stationary process. To carry out a first
analysis, the whole dataset was divided in two different groups, the first one includes the points
regarding segments during the day time (from 07:00 to 21:00 h) and the second one the points
regarding the night time (from 21:01 to 06:59 h). In order to compare the two groups and to extract
information, the Kernel Density Estimation (KDE) of the 4 parameters was estimated. Kernel density
estimation is a nonparametric technique for probability density function estimation [13]. This study
reported the first results of the kernel density estimations of the 4 parameters divided in day and
night segments.
Results
The preliminary results on the analysis of the union of the two datasets (ARIMMORA and EXPERS)
are shown in the following figure. Fig. 1 represents, as example, the evaluated kernel density
estimation of the 4 parameters for the segments regarding the day time compared with the
segments regarding the night time.
Conclusion
The aim of this paper was to obtain a better description and a more accurate analysis about the
exposure of children to ELF magnetic field. Preliminary analysis seems to show that the mean value
of the exposure to ELF-MF is higher during the day than during the night. Moreover, the segments
during the night show a longer duration and less variance compared with the segments of the day.
This resulted also in differences in KDE of the coefficient of the AR process of order 1 of the two
689
groups. A more systematic search of the features that are the most influent will be conducted. This
will be performed splitting the dataset into several sub-datasets according to a feature: the features
that will lead to the biggest difference between models will be considered as the most discriminant.
Acknowledgement
This research is supported by the French National Research Program for Environmental and
Occupational Health of ANSES (2015/1/202): Project ELFSTAT - In depth evaluation of children’s
exposure to ELF (40 – 800 Hz) magnetic fields and implications for health risk of new technologies,
2015-2019. The French data come from the EXPERS study database, subsidized by the French
Ministry of Health, EDF and RTE, and carried out by Supélec, EDF and RTE.
References
[1] A. Ahlbom, N. Day, and M.Feytching, “A pooled analysis of magnetic fields and childhood
leukemia”, British Journal of Cancer, vol. 83(5), pp. 692-698, 2000.
[2] L. Kheifets, A. Ahlbom, C.M. Crespi et al., “Pooled analysis on recent studies on magnetic fields
and childhood leukemia”, British Journal of Cancer, vol. 103(7), pp. 1128-1135, 2010.
[3] I. Liorni, M. Parazzini, B. Struchen, S. Fiocchi, M. Röösli, and P. Ravazzani, “Children’s Personal
Exposure Measurements to Extremely Low Frequency Magnetic Fields in Italy”,International journal
of environmental research and public health,vol. 13(6), pp. 549, 2016.
[4] B. Struchen, I. Liorni, M. Parazzini, S. Gängler, P. Ravazzani, and M. Röösli, “Analysis of
personal and bedroom exposure to ELF-MFs in children in Italy and Switzerland”, Journal of
Exposure Science and Environmental Epidemiology,vol. 26(6), pp. 586-596, 2016.
[5] I. Magne, M. Souques, I. Bureau, A. Duburcq, and E. Remy, J. Lambrozo “Exposure of children
to extremely low frequency magnetic fields in France: Results of the EXPERS study.” JESEE, 2017,
27 (5) : 505-512.
[6] H. Tolba, L. Le Brusquet, M. Parazzini, S. Fiocchi, E. Chiaramello, P. Ravazzani, M. Röösli, I.
Magne, and M. Souques. "Modelling the Extremely Low Frequencies Magnetic Fields Times Series
Exposure by Segmentation." In11ème Congrès National de Radioprotection, 2017.
[7] A. J. Scott, M. Knott, “A cluster analysis method for grouping means in the analysis of variance”,
Biometrics, pp.507-512, 1974.
[8] A.B. Olshen, E. Venkatraman, R. Lucito, and M. Wigler, “Circular binary segmentation for the
analysis of array-based dna copy number data”, Biostatistics, vol. 5, pp. 557-572, 2004.
[9] P. Fryzlewicz et al. “Wild binary segmentation for multiple change-point detection”, The Annals of
Statistics, vol. 42, pp. 2243-2281, 2014.
[10] R. Killick, P. Fearnhead, I. Eckley, “ Optimal detection of changepoints with a linear
computational cost”, Journal of the Americn Statistical Association, vol. 107, pp. 1590-1598, 2012.
[11] B. Jackson, J. D. Scargle, D. Barnes, S. Arabhi, A. Alt, P. Gioumousis, E. Gwin, P.
Sangtrakulcharoen, L. Tan, and T. T. Tsai, “An algorithm for optimal partitioning of data on an
interval”, IEEE SignalProcessing Letters, vol. 12, pp. 105-108, 2005.
[12] R. Killick, I. Eckley, “Changepoint: An r package for changepoint analysis”, Journal of Statistical
Software, vol. 58, pp. 1-19, 2014.
[13] D.W. Scott, S.R. Sain. "9-Multidimensional Density Estimation", Handbook of statistics 24, pp.
229-261, 2005.
Figures

690
 Figure 1. Estimated kernel density function of the four variables of the segments of the day (in
red) comparing with estimated kernel density function of the four variables of the segments of
the night (in blue), from above to down, left to right: mean, variance, coefficient of the AR
process and duration

PB-84 [14:30]
STUDENT PAPER
Does Extremely Low Frequency Magnetic Fields stimulations of the vestibular system
modulate postural control in humans?
Nicolas Bouisset1, 2, Sebastien Villard1, 3, Daniel Goulet7, Michel Plante7, Martine Souques6, François
Deschamps5, Genevieve Ostiguy7, Jacques Lambrozo6 & Alexandre Legros1, 2, 3, 4, 8
1Human Threshold Research Group, Lawson Health Research Institute, London, ON, Canada
2Department of Kinesiology, Western University, London, ON, Canada
3Department of Medical Biophysics, Western University, London, ON, Canada
4Department of Medical Imaging, Western University, London, ON, Canada
5RTE, Département Concertation et Environnement, Paris-La Défense, France
6Service des Études médicales , EDF, Levallois-Perret Cedex, France
7Hydro-Québec,, Montréal, Québec, Canada
8EuroMov, Université de Montpellier, Montpellier, France
Keywords: Human, ELF/LF, Work in Progress
Presented by: Nicolas Bouisset
Exposure to extremely low frequency magnetic fields (ELF-MF) induces electric fields and currents
within the human body that modulate nervous system functions. Although still a matter of debate,
there are evidences that the vestibular system could respond to ELF-MF. It is well established
691
however that specific electrical stimulations of the vestibular system trigger specific postural
outcomes. This work investigates the vestibular response to ELF-MF and alternating electrical
currents through their impact on human postural sway.
Introduction
Workers and patients within the vicinity of MRI scanners often report transient sensory sensations
such as illusions of rotating, vertigo, dizziness and nausea [Rauschenberg et al., 2014]. Such
sensations are related to modulations of vestibular perceptions. Recent research [Laakso et al.,
2013] have underlined that these vestibular modulations could be due to induced electric fields and
currents resulting from magnetic induction caused by time-varying Magnetic Fields (MF) in the
Extremely Low Frequency (ELF< 300 Hz) range. For instance, ELF MFs have been reported to
possibly impact human postural sway [Legros et al., 2012] and produce vestibular like symptoms
[Schaap et al., 2015]. Interestingly, another source of direct and alternative currents called galvanic
vestibular stimulation (DCGVS and ACGVS) is known to specifically impact the vestibular system
and to increase postural instability [Fitzpatrick and Day, 2004].
ELF MFs generated by power lines, ubiquitous in our daily lives, oscillate at 60 Hz in North America.
Considering the important issue of protection of the public and workers in ELF-MF environments,
international exposure recommendations are issued. These recommendations are based on
knowledge relative to exposure induced phosphenes [ICNIRP, 2010; IEEE, 2002] Phosphenes are
described as flickering lights appearing in the visual field. They can be triggered both by ELF-MF
and electrical currents [Lövsund et al., 1980] and can be linked to the retinal photoreceptors’
membrane electrical potential modulation [Attwell, 2003]. The eyes’ photoreceptors share common
neurophysiological properties with the hair cells within the vestibular system, which function is to
transduce mechanical information into an electrical signal. Both types of cells use graded potential
for signal processing [Juusola et al., 1996] both releasing glutamate gradually from ribbon synapses
[Sadeghi et al., 2014]. Considering the close neurophysiological similarities between the two
sensory systems and the evidences of vestibular susceptibility to ELF MFs, our group developed an
exposure system and a protocol aiming to detect subtle changes in postural response to vestibular
MF exposures of up to 100 mT [Villard et al., under review]. This study showed postural modulations
due to DCGVS as expected, but it failed to detect changes in postural control due to electric and
magnetic stimuli given at frequency ranging from 20 to 160 Hz (up to 2 mA and 100 mT
respectively). This was mainly due to methodological limitations (mechanical clues available to the
participant helping them stabilize their posture) that have been addressed in this work.
The objective of this work is 1. to evaluate if human postural control can be modulated by high levels
of ELF MF (up to 100 mT) applied to the human vestibular system and 2. to evaluate if the response
is frequency dependent. Based on previous evidences, we hypothesize that electric and magnetic
vestibular stimuli will increase postural instability and that the effect will be stronger at lower
frequencies. The DCGVS will serve as a positive control and will induce a lateral loss of balance
towards the unstimulated ear [Fitzpatrick and Day, 2004].
Material and Methods
This is an ongoing study. Twelve participants have completed the study so far (40 participants total
planned) in the Human Threshold Research Facility at St. Joseph’s Hospital in London, Ontario,
Canada. Healthy volunteers aged between 18-55 without history of vestibular-related pathology or
dysfunction, chronic illnesses and any neurological diseases are eligible to the experiment.
A force plate (OR6-7-1000, AMTI, USA) is used to record participant’s body sway at 1 kHz, using an
A/D National Instrument card (NI SCB-68A, National Instruments, USA), driven by LabVIEW
(National Instruments, USA). MF vestibular stimulations are delivered to the subjects’ right vestibular
system via a customized head set coil, consisting of 570 turn-coils of 5.9 cm of mean diameter, with
a 2-cm diameter core of Permendur-49 (The Goodfellow Group, Coraopolis, PA, USA). This device
is attached with rods to a vest worn by the participants around their chest. Electrical vestibular
stimulations are delivered using a StarStim device (StarStim, Neuroelectrics, Spain). The cathode is
placed behind the right mastoid process and the return electrode behind the C7 process.

692
Participants are tested in periods of 20 seconds spaced with 30 seconds of rest in between. The test
consists in standing on the force plate, upon a 6-cm thick foam surface with the eyes closed, arms
resting at their side, feet together (to maximize vestibular contribution). Exposure conditions consist
in five seconds of MF or electrical stimulations (100 mT and 2 mA respectively) at different
frequencies (20 Hz, 60 Hz, 90 Hz, 120 Hz and 160 Hz). Three sham conditions (for DCGVS,
ACGVS and MF exposures) were also performed. This protocol was approved by the Health
Sciences Research Ethics Board (#106122) at Western University.
Results
Five sway characteristics were calculated off-line from the recorded center-of-pressure (COP) data.
For the purpose of this abstract, one representative characteristic has been selected to be reported
in the statistical analysis: the sway path length in the frontal plane (calculated in cm – represents the
total length of the center-of-pressure (COP) excursion).
A paired t-test was used to analyze the effect of our positive control (DCGVS vs no stimulation).
Results show a significantly bigger sway path with DCGVS as compared to no stimulation (t (11) =
6.1392, p < 0,001, R 2= 0.62) (Figure 2). So far, all subjects have lost their balance towards the left
side when stimulated by direct current.
A two-way ANOVA (2 stimulation modalities x 6 exposure conditions) for repeated measures was
used to explore the effect of frequency of the time-varying exposure types. The stimulation
modalities were ACGVS and MF and the exposure conditions compared the sham exposure and the
5 frequencies (20 Hz, 60 Hz, 90 Hz, 120 Hz and 160 Hz). There is no significant main effect of the
stimulation modality on the path length (F (1, 11) = 0.0134, p= 0.91, h2 = 0.03, power = 0.05). No
significant main effect of exposure conditions was found either (F (5, 110) = 1.48, p=0.20, h2 = 0.14,
power = 0.16). Finally, there is no significant interaction effect between the type of stimulation and
the type of frequency on the path length (F (5, 110) = 0.91, p=0.47, h2 =0.03, power = 0.05) (Figure
2).
Discussion
This study explores the effects of ELF MF and ACGVS exposures at matching frequencies. As
expected, our positive control results confirm an instantaneous effect of DCGVS. All participants,
lost their balance towards the opposite side of the cathodal stimulation [for review of this effect, see
Fitzpatrick and Day, 2004].
Cathodal currents are thought to depolarize vestibular’s afferents leading to an increase in its firing
rate, whereas anodal currents produce opposing effects [Goldberg et al., 1984].
The brain perceives head motion by integrating the difference between the firing rate of both
vestibular systems. When still with eyes are closed, GVS evokes sensations of head movement
[Fitzpatrick and Day, 2004]. The perception starts 500 ms after GVS is turned on [Barnett-Cowan,
2013]. With DCGVS, the time delays needed for perceived movement is enough to sense one is
falling towards the cathodal side which forces one to counterbalance towards the anodal side
leading to an actual fall. With ACGVS the best perceptions of movement are felt under 2 Hz
[Stephan et al., 2005], which corresponds to the aforementioned time perception threshold reported
by Barnett-Cowan (2013)..
At 20 Hz, the current’s period is 50 ms which is under the time perception threshold. This might
explain why our stimulations, between 20 hz and 160 Hz, appear ineffective in modulating postural
control.
However, stimuli up to 500 Hz are used clinically to assess vestibular function, highlighting the
sensitivity of vestibular hair cells to high frequencies [Curthoys, 2017]. Also, when one is standing,
ACGVS up to 75 Hz triggers muscles responses in the neck, while legs muscles only respond below
20 Hz [Forbes et al., 2013]. However, these myogenic responses do not destabilize posture and are
potentially due to existing biomechanical filtering mechanisms [Forbes et al., 2014].
Designing future studies focusing on vestibulo-ocular or vestibulo-cortical activity could avoid
693
perceptual and biomechanical filtering and better characterize vestibular responses to ELF MF.
It is important to keep in mind that this is an ongoing study and that only preliminary results are
presented here. Final conclusion can only be drawn upon study completion. This work will complete
the pool of human data aiming to contribute to the update of MF safety exposure recommendations
and guidelines for workers and the public.
References
Attwell D. 2003. Interaction of low frequency electric fields with the nervous system: the retina as a
model system. Radiat. Prot. Dosimetry 106:341–8. http://www.ncbi.nlm.nih.gov/pubmed/14690277.
Curthoys IS. 2017. The new vestibular stimuli: sound and vibration—anatomical, physiological and
clinical evidence. Exp. Brain Res. 235:957–972. http://link.springer.com/10.1007/s00221-017-4874-
y.
Fitzpatrick RC, Day BL. 2004. Probing the human vestibular system with galvanic stimulation. J.
Appl. Physiol. 96:2301–16. http://jap.physiology.org/content/96/6/2301.
Forbes PA, Siegmund GP, Schouten AC, Blouin J-S. 2014. Task, muscle and frequency dependent
vestibular control of posture. Front. Integr. Neurosci. 8:94. http://www.ncbi.nlm.nih.gov/pubmed
/25620919%5Cnhttp://www.pubmedcentral.nih.gov/articlerender.fcgi?artid=PMC4288134.
Goldberg JM, Smith CE, Fernández C. 1984. Relation between discharge regularity and responses
to externally applied galvanic currents in vestibular nerve afferents of the squirrel monkey. J.
Neurophysiol. 51:1236–56. http://www.ncbi.nlm.nih.gov/pubmed/6737029.
ICNIRP. 2010. Guidelines for limiting exposure to time-varying electric and magnetic fields (1 Hz to
100 kHz). Health Phys. 99:818–36.
IEEE. 2002. IEEE Standard for Safety Levels with Respect to Human Exposure to Electromagnetic
Fields, 0–3 kHz. Technology.
Juusola M, French AS, Uusitalo RO, Weckström and M. 1996. Information processing by graded-
potential transmission through tonically active synapses. Trends Neurosci.:292–297.
Laakso I, Kännälä S, Jokela K. 2013. Computational dosimetry of induced electric fields during
realistic movements in the vicinity of a 3 T MRI scanner. Phys. Med. Biol. 58:2625–40.
http://iopscience.iop.org/article/10.1088/0031-9155/58/8/2625.
Legros A, Corbacio M, Beuter A, Modolo J, Goulet D, Prato FS, Thomas AW. 2012.
Neurophysiological and behavioral effects of a 60 Hz, 1,800 lT magnetic field in humans. Eur. J.
Appl. Physiol. 112:1751–1762.
Lövsund P, Öberg PÅ, Nilsson SEG. 1980. Magneto- and electrophosphenes: A comparative study.
Med. Biol. Eng. Comput. 18:758–764.
Rauschenberg J, Nagel AM, Ladd SC, Theysohn JM, Ladd ME, Möller HE, Trampel R, Turner R,
Pohmann R, Scheffler K, Brechmann A, Stadler J, Felder J, Shah NJ, Semmler W. 2014. Multicenter
study of subjective acceptance during magnetic resonance imaging at 7 and 9.4 T. Invest. Radiol.
49:249–59. http://www.ncbi.nlm.nih.gov/pubmed/24637589.
Sadeghi SG, Pyott SJ, Yu Z, Glowatzki E. 2014. Glutamatergic Signaling at the Vestibular Hair Cell
Calyx Synapse 34:14536–14550.
Schaap K, Portengen L, Kromhout H. 2015. Exposure to MRI-related magnetic fields and vertigo in
MRI workers. Occup. Environ. Med.:161–166.
Stephan T, Deutschl??nder A, Nolte A, Schneider E, Wiesmann M, Brandt T, Dieterich M. 2005.
Functional MRI of galvanic vestibular stimulation with alternating currents at different frequencies.
Neuroimage 26:721–732.
Figures

694
Figure 1. Effect of DCGVS vs no stimulation on the path length in the frontal plane. Error bars
represent standard error of mean. The asterisk describes that the path length with DC current
is significantly higher than with the sham

Figure 2. Effect of the type of stimulation and the type of frequency on the path length in the
frontal plane. Error bars represent standard error of mean.

695
PB-86 [14:30]
Determination of external voltage level for causing heart fibrillation in human body
Stefan Cecil1, Richard Überbacher1, Michael Koch2 & Michael Bartonek2
1EMC&Optics, Seibersdorf Laboratories, Seibersdorf, Austria, 2444
2Eaton Industries Austria GmbH, Schrems, Austria, 3943
Keywords: Human, ELF/LF, Completed (unpublished)
Presented by: Stefan Cecil
In this project the current flow through the human body on two different paths was investigated.
Simulations with the body model "Duke" were done, to determine the flowing electric current through
the heart. The results of the simulations were compared to the results from the corresponding
measurements with rabbit hearts, where the fibrillation threshold was evaluated. With the results of
the simulations and measurments the external voltage level for heart fibrillation in the human body
could be determined.
Introduction
In case of an electric accident a current flows through the human body. A part of the current flows
through the human heart and can cause dangerous heart fibrillation, because of the impact of the
flowing current to the nerval stimulation of the heart. In conventional households the fault current,
which can flow through the human body is limited by a fault current protective device, to avoid the
dangerous situation of heart fibrillation. Because of new technologies of current sources with high
switching frequencies fault currents have to be limited up to 100kHz.
In the project “Kermit” investigations in simulation and measurement were done, to determine the
needed limits of the current flow through the human body up to 100kHz.
In previous investigations a model to simulate the current flow through the human body was
developed. Especially a simulation model of the human skin had to be developed, because the
impedance of the skin has a big impact and is voltage and frequency depended and cannot be
modelled by conventional available simulation models. With the developed simulation model the
current density in the human heart, which is caused by an external voltage on the skin of the human
body, can be simulated. In the measurement part of the project “Kermit” a fibrillation threshold for
the current density in the human heart was determined. With these measured values of current
density and the simulation model a threshold of fibrillation for an external voltage can be identified.
Methods
The current flow through the human body was numerical simulated with the FDTD-tool SEMCAD X.
Because of the low frequencies the method of frequency scaling was applied [2]. The LF-solver of
SEMCAD X was applied too and showed nearly the same results than the frequency scaling
(differences between the methods < 2%). The human body model "Duke" from the virtual family [3]
was applied in the simulations. The dielectric properties for the assignment to the tissues were taken
from the SEMCAD X tissue properties data base, which corresponds to the values published by C.
Gabriel [4].
The method and the verification of the model of the skin is described in [1]. Basically the skin at the
contact of electric circuit and the human body is replaced by a network of two capacitors and two
resistors. This network models the frequency dependency of the skin. For different voltage levels the
elements of the network have different values, so the voltage dependency of the skin is considered.
The results of the current density in the heart strongly depend on the model of the heart. So the
heart in the body model was improved by the application of anatomical data. Details of the heart
were added and the shape of the heart was smoothed. Also the blood vessels around the heart
were improved. The new model of the heart was positioned in the body model "Duke".
With the body model and the network for the skin two current paths through the human body were
simulated: From left hand to the right hand and from left hand to both feet. The two scenarios are
pictured in Figure 1. The simulations were done for 50Hz, 1kHz, 10kHz and 100kHz and for the
696
voltage levels of 10V, 100V, 230V and 1000V. This gives 16 different setups for each of the two
scenarios.
From the simulation results the impedance of the body and the current densities in the human heart
were extracted. The most important value is the current density in the right ventricle, because this
value was measured in the tests with the rabbit heart. So from the simulations the maximum current
density in the right ventricle and the maximum of the current density in the whole heart were
extracted.
The measurements to determine the fibrillation threshold were done with a rabbit heart in a pot filled
with blood. To analyze the current distribution in the measurement setup the setup was modelled in
SEMCAD too. So a model of the rabbit heart had to be developed. A rabbit heart, which was
formerly used at the measurements was radiographed in a computer tomography. With the resulting
sectional images a model of the rabbit heart was generated and simulations of the measurement
setup were performed.
Results
From the simulations the current densities in the heart and the impedance of the current path
through the body were extracted. The current distribution at the two different paths at 50Hz/230V
are pictured in Figure 2.
The simulations were done at the 4 different frequencies (50Hz, 1kHz, 10kHz, 100kHz) and for
some different voltage levels. At the same frequency with different voltage levels only the values of
the elements in the model of the skin are different. The impedance of this discrete elements can be
calculated analytically. So the impact of the different skin models to the impedance and the current
density in the heart can be calculated analytically. The values for a scenario at the same frequency
and different voltage level can be calculated from another simulation. So not each combination of
frequency and voltage level had to be simulated, only one simulation with each frequency at one
voltage level had to be done. The results of a directly calculated simulation and the results
calculated from another scenario at another voltage level were compared, the difference was <0,5%.
Results for the current density in the right ventricle are shown in Table 1.
Max J Venticle, Current path left hand-feet

U_eff [V] 10 100 230 1000

f [Hz] [A/m²] [A/m²] [A/m²] [A/m²]

50 0,1367 3,77 11,89 74,70

1000 0,4169 7,90 19,02 85,59

10000 0,7930 8,35 19,23 83,66

100000 0,7800 7,80 17,95 78,04

Table 1: Maximum of current density in the right ventricle at the 16 different combinations of voltage
level and frequency at the current path from left hand to both feet. The values for 230V are directly
from simulation, the other values are calculated analytically from the scenario with 230V.
From the measurements in the "Kermit"-project with the rabbit heart values of the fibrillation
threshold of the current density in the right ventricle of the human heart are known. The correlation
of the current densities in the human and the rabbit heart and the corresponding fibrillation
thresholds was derived by a physiologist. From the simulations the correlation of the current density
in the right ventricle and the external applied voltage is known. The current density was evaluated at
4 voltage levels, with these levels a continuous curve was interpolated. On this curve from the
simulations the measured fibrillation thresholds were applied and a corresponding external voltage

697
level for the fibrillation threshold can be determined.
This procedure could only be done for 50Hz and 1kHz, because the measurements showed, that at
10kHz and higher frequencies a fibrillation of the heart could only be induced at very high current
values, where significant thermal effects appeared. Because of the natural variation of the used
rabbit hearts the determined fibrillation thresholds are shown in a maximum, average and a
minimum value. The results for 50Hz are shown in a diagram in Figure 3. The results in numbers for
50Hz and 1kHz are shown in the following Table 2.
Path hand- hand at 50Hz Path hand- hand at 1kHz

Measurement Measurement
at voltage [V] at voltage [V]
[A/m²] [A/m²]

Min 10,61 238 Min 81,32 875

Avg 27,58 482 Avg 141,42 1550

Max 40,66 650 Max 254,66 3000

Path left hand- feet at 50Hz Path left hand- feet at 1kHz

Measurement Measurement
at voltage [V] at voltage [V]
[A/m²] [A/m²]

Min 10,61 210 Min 81,32 940

Avg 27,58 450 Avg 141,42 1550

Max 40,66 615 Max 254,66 2600

Table 2: Fibrillation threshold from measurements (Min, Max, Avg.) and the determined external
voltage levels on the human body at 50Hz and 1kHz with the two current paths
Conclusion and Outlook
The shown results are liable to several uncertainties especially the human body model. The present
results are only valid for the body model "Duke", a variation due to other body models could not be
investigated. The tissues of the body model are defined isotropic, in reality the electric conductivity
is anisotropic, especially the skeletal muscles. The body model shows a continuous muscle tissue
on the chest and the back. In reality there are several groups of muscles, which are separated by
small layers of connective tissue. Because of this it is imaginable, that too much of the electric
current is flowing through the muscles and not through the heart. If the percentage of the current
through the heart would be higher, the fibrillation threshold could be reached with lower voltage
levels.
So for further investigations the impact of the body model to the results shall be evaluated more in
detail and if needed, the body model shall be modified. Additionally more different body models shall
be applied to evaluated the impact of the anatomical variation.
Acknowledgement
The project “Kermit” is founded by Eaton Industries Austria GmbH and the Austrian Research
Promotion Agency (FFG).
Literature

698
[1] S. Cecil, G. Schmid, M. Koch, M. Bartonek; Modelling of ELF-current flow throw the human body;
BioEM2016, Ghent, Belgium, Jun 05 - 10, 2016
[2] Gandhi, Om P.; Chen, Jin-Yuan; Numerical Dosimetry at Power-Line Frequencies Using
Anatomically Based Models; Bioelectromagnetics Supplement 1:43-60 (1992)
[3] ITIS Human models “Virtual Family” http://www.itis.ethz.ch/index/index_humanmodels.html , ITIS
Foundation, Zeughausstrasse 43, CH-8004, Zürich, Switzerland
[4] Gabriel, C.; Gabriel, S. (1996): “Compilation of the Dielectric Properties of Body Tissues at RF
and Microwave Frequencies“
[5] Gottfried Biegelmeier: „Wirkungen des elektrischen Stromes auf Menschen und Nutztiere“, vde-
Verlag, ISBN 3-8007-1452-3
[6] J. Patrick Reilly: „Applied Bioelectricity – From Electrical Stimulation to Electropathology“ Kapitel
2, Seite 18; Springer-Verlag 1998
[7] „Bildgebende Verfahren der Medizintechnik“, Kapitel „Impedanztomographie“, Institut für
Biomedizinische Technik Universität Karlsruhe, Vorlesungsskriptum http://www-ibt.etec.uni-
karlsruhe.de/postscript/bvm_Impedanztomographie.pdf
Figures

Figure 1. Model for simulation of flowing current from one hand to the other and from left hand
to both feet with the human body model “Duke”. The properties of the skin are modelled by
lumped elements.

699
Figure 2. Distribution of current flow through the body model duke. Two different current paths
simulated: From left hand to right hand and from left hand to both feet (230V with 50Hz)

Figure 3. Determination of the needed voltage on the body for meeting the fibrillation
threshold. The fibrillation threshold was determined in measurements and is shown in
minimum, maximum and average value. With the simulation results the corresponding voltage
of the fibrillation threshold can be determined.

700
PB-88 [14:30]
Effects of electromagnetic fields from Long Term Evolution on awake electroencephalogram
in healthy humans
Setsu Nakatani-Enomoto1, Madoka Yamazaki2, Hiroyuki Enomoto1 & Yoshikazu Ugawa1
1Department of Neurology, Fukushima, Japan, 960-1295
2Department of Health Science, Saitama, Japan, 355-8501
Keywords: Human, RF/Microwaves, Completed (unpublished)
Presented by: Setsu Nakatani-Enomoto
We analyzed the effects of Long Term Evolution (LTE)-like EMW for 30 min on awake
electroencephalogram (EEG) in healthy adult humans. The maximum local SAR averaged over 10 g
mass was 2.0 W/kg. Thirty-eight subjects aged 20 to 36 years old participated in this study. The
power spectra of θ, α and β frequency ranges at eight cortical sites were analyzed. We showed
enhancement of θ and α waves after both real and sham exposures. The results may be consistent
with the drowsiness of the subjects. However, EEG power spectra were not affected by the
stimulation condition. We conclude that LTE-like exposure for 30 min has no detectable effects on
awake EEGs in healthy humans.
Introduction
Mobile phones are now indispensable for daily life, and, adverse effects of electromagnetic field
(EMF) emitted by the mobile phone have been a great concern. The mobile phones are unlikely to
have harmful thermal effects on the human health [1]: however, several experiments on resting state
electroencephalogram (EEG) in awake human after mobile phone exposure reported inconsistent
results. The EMF exposure from a Global System for Mobile Communications (GSM) system
evoked some changes in EEG coherence. [2] The results of the EEG power spectra were
inconsistent: enhanced [3-5] or reduced [6, 7] α wave power spectra. Some studies showed no
effects. [8, 9] Wideband Code Division Multiple Access (W-CDMA) system showed no effects on
EEG power spectra. [5, 10] Long Term Evolution (LTE) provides high performance technique as the
next-generation communication, which has been widely used in Japan. Previous studies showed
decreased EEG α and β power [11] or modulated synchronization patterns of EEG activation across
the brain [12] after the LTE exposure. The aim of this study was to investigate aftereffects of LTE-
related EMF upon the resting EEG.
Methods
Subjects
Thirty-eight healthy subjects aged 20 to 36 years old were recruited through the public invitation.
Medicines, dietary supplement and healthy food products were prohibited for at least 14 days before
the experiment. Hearing aid users were not included in this study. Their health condition and lifestyle
were checked with a self-completed questionnaire.
Experimental procedures
Experiment began at 1 p.m. or 3 p.m. Alcohol and smoking were forbidden during the whole
experiment. From 4 h prior to the experiment, subjects were permitted to take only water, milk,
100% fruit juice and usual meals. Telephone call longer than 10 min using mobile phone, was
restricted from 0 a.m. to 9 a.m., and phone call was totally prohibited from 9 a.m. to the end of the
experiment. Text messaging with a mobile phone in a short time was only allowed, but, another
performances using a mobile phone were unpermitted throughout the day. Their health condition
and lifestyle were checked again to confirm that they satisfied the entry criteria before each
experiment.
The subjects sat on a comfortable chair during the experiment. [4-12] EEG was recorded for 60 min
in total. EMF was exposed for 30 min [3, 4, 10-12] at the middle of the experiment (Fig. 1B). The

701
subjects were asked to remain awake, and instructed to keep the eyes closed when recording the
EEG. [2, 4, 6, 8, 9, 11] All the subjects jointed two experiments with real and sham stimulation
conditions on two different days.
Electroencephalogram (EEG) recording
EEG data was recorded with 128-channel amplifier system and digitized at 500 Hz sampling rate.
The signals were filtered with a 0.5 to 60 Hz bandpass filter. Figure 1A shows a subject wearing an
EEG sensor net and patch antenna. EEG was recorded with an EEG recording system (Geodesic
EEG Systems 300, Electrical Geodesics, OR, USA). Subjects wore the 128-channel HydroCel
Geodesic Sensor Net (Electrical Geodesics) after soaking the electrodes into the solution of
potassium chloride (Crystalline/USP/FCC, Fisher Scientific International, PA, USA) and Johnson’s
baby shampoo (Johnson & Johnson Consumer Products Company, NJ, USA).
Electromagnetic field (EMF) exposure
The exposure was performed in a shielded room which made 80 dB or more reduction at
frequencies between 5000 kHz and 3 GHz in order to prevent the environmental EMF. The exposure
condition was randomly controlled by the personal computer. This information was decoded after
analyses of the EEG.
The specially designed microstrip patch antenna (Harada Industry, Tokyo, Japan) used in a previous
study [13] was placed in front of the right ear under the EEG sensor net (Fig 1). The EMF was
controlled by a signal generator (N5182A;Agilent Technologies, US-CA, USA) placed outside the
shielded room to transmit continuous wave of LTE-like signal (3GPP Rel.9) at 1950 MHz. The
bandwidth was defined as 20 MHz. We used 64 quadrate amplitude EMF modulation with 100%
resource block allocation.
Prior to the experiment, we made a distribution map of the SAR to confirm that EEG was not
influenced by EEG sensor net. The maximum local SAR averaged over 10 g mass measured at the
surface of the phantom brain under the antenna was 2.00 W/kg with sensor net, and 2.01 W/kg
without the net. No temperature rise was observed at the antenna during the exposure condition.
Data analysis
All the data were analyzed by a specialist who was blind to the experimental conditions. We
eliminated the data with motion or eye blink artifacts by visual inspection using a software of EMSE
Suite (Source Signal Imaging, NC, USA). EEG data were pre-processed to exclude segments with
artifacts. Residual parts with eyes closed condition were divided into 8 second (s) segments with a
Hanning tapering window prior to computing the power spectra. Fifteen segments of the EEG for
total 120 s were used to make average power spectra of one analysis time window for every subject
at each recording sites. The power spectrum was obtained through a power spectrum analysis
program of EMSE. Eight cortical sites (F3, F4, C3, C4, O1, O2, T3 and T4) were the targets of
analyses. The EEG frequency spectrum was subdivided in delta (δ, 0 to 3.8 Hz), theta (θ, 4 to 7.8
Hz), alpha (α, 8 to 12.8 Hz) and beta (β, 13 to 30 Hz) frequency ranges. However, we used only θ, α
and β waves in this study because of large artifacts in the δ range.
Statistical analysis
Statistical analyses were performed using SPSS 17.0 for Windows (SPSS, Chicago, Illinois, USA).
A two-factor repeated-measures analysis of variance (rANOVA) was performed to compare the two
exposure conditions. The analysis included the factor of EXPOSURE (real and sham) and TIME
(pre, during and post sessions). A Greenhouse-Geisser correction for non-sphericity was used in all
analyses. A difference was accepted as significant at a p-value < 0.05 for all analyses.
Results
Three women out of 38 participants stopped experiment because of abnormal EEGs or poor
physical condition. One subject was excluded because of a lot of artifacts. We analyzed, thereafter,
the EEG data of 17 men (22.3 ± 1.3 years old) and 17 women (22.6 ± 3.8 years old). No subjects
were able to differentiate the real exposure from sham exposure.

702
Figure 2 and 3 show power spectra (θ, α, and β wave components) measured at before, during, and
after real or sham exposures(Fig. 2, F3, F4, C3, C4, P3, P4, O1 and O2; Fig. 3, T3 and T4). The
main effect of TIME was significant in θ component recorded at F4, C3, C4, T3 and T4, and α
component at C3, C4, P3, T3 and T4. However, neither the main effect of EXPOSURE nor
interaction between EXPOSURE × TIME was significant (p > 0.05) in every component. Our results
indicate that the EEG power spectra were affected by time during the experiment similarly in the real
and shamexposures, but not affected by the stimulation condition.
Discussion
The present results showed that LTE-like exposure had no effect on awake EEG power spectra in
healthy subjects. Several other studies using GSM or W-CDMA systems reported inconsistent
results to ours. [5, 8-10] The LTE exposure for 30 min using a radio frame paced at right ear
changed α and β bands at front-temporal area. [11] The local SAR averaged over 10 g mass in this
study was 2.0 w/kg, which was stronger than the condition of Yang et al. of 1.34 w/kg. The antenna
was placed on the right side with a distance of 1 cm to the ear in the experiment by Yang, et al.;
whereas in our study, antenna was located in front of the right ear. This difference of exposed area
may explain the inconsistency.
We showed enhancement of θ wave at frontal, central and temporal areas, and α wave at central,
parietal and temporal areas. These effects were observed after both real and sham exposures,
which means EEG was changed with time. Why the EEG was changed during the experiment?
Boredom, fatigue and monotony inducing early drowsiness during long term EEG recordings may
explain these EEG changes. In adult, the onset of drowsiness is characterized by gradual or brisk
alpha dropout. When the drowsiness goes deeper, slow activity with medium voltage increases.
Furthermore, paradoxical α response characterized by higher amplitude than the individual’s regular
α rhythm occurs at this stage. In our study, power spectra showed increased amplitudes of θ (slow
activity) and α waves, which may be consistent with the drowsiness of the subjects.
The present study has several limitations. First, people with electromagnetic hypersensitivity was
not included in this study. Therefore, we cannot exclude the possibility that the EEG may be affected
by LTE exposure in some hypersensitive subjects. Second, our study targeted young adult healthy
peopleonly. One previous study showed age difference of EEG response after exposure of SGM
system. [5] Third, this study used only the continuous-wave signal. However, one previous study
demonstrated changes of power spectra after pulse-modulated EMF, but not after continuous-wave
exposure. [3] These confounding factors should be considered in future studies.
Conclusions
We conclude that LTE-like exposure for 30 min has no detectable effects on awake EEGs in healthy
humans.
Acknowledgement
We thank the Association of Radio Industries and Businesses for their support in technical advice on
the setup of the exposure system and dosimetry assessment.
References
1. Van Leeuwen GM, Lagendijk JJ, Van Leersum BJ, et al. 1999. Calculation of change in brain
temperatures due to exposure to a mobile phone. Phys Med Biol 44:2367-2379.
2. Vecchio F, Babiloni C, Ferreri F, et al. 2007. Mobile phone emission modulates interhemispheric
functional coupling of EEG alpha rhythms. Eur J Neurosci 25:1908-1913.
3. Huber R, Treyer V, Borbély AA , et al. 2002. Electromagnetic fields, such as those from mobile
phones, alter regional cerebral blood flow and sleep and waking EEG. J Sleep Res 11:289-295.
4. Regel SJ, Gottselig JM, Schuderer J, et al. 2007. Pulsed radio frequency radiation affects
cognitive performance and the waking electroencephalogram. Neuroreport 18:803-807.
5. Croft RJ, Leung S, McKenzie RJ, et al. 2010. Effects of 2G and 3G mobile phones on human

703
alpha rhythms: Resting EEG in adolescents, young adults, and the elderly. Bioelectromagnetics
31:434-444.
6. D'Costa H, Trueman G, Tang L, et al. 2003. Human brain wave activity during exposure to
radiofrequency field emissions from mobile phones. Australas Phys Eng Sci Med 26:162-167.
7. Ghosn R, Yahia-Cherif L, Hugueville L, et al. 2015. Radiofrequency signal affects alpha band in
resting electroencephalogram. J Neurophysiol 113:2753-2759.
8. Hietanen M, Kovala T, Hämäläinen AM. 2000. Human brain activity during exposure to
radiofrequency fields emitted by cellular phones. Scand J Work Environ Health 26:87-92.
9. Perentos N, Croft RJ, McKenzie RJ, et al. 2007. Comparison of the effects of continuous and
pulsed mobile phone like RF exposure on the human EEG. Australas Phys Eng Sci Med
30:274-280.
10. Trunk A, Stefanics G, Zentai N, et al. 2013. No effects of a single 3G UMTS mobile phone
exposure on spontaneous EEG activity, ERP correlates, and automatic deviance detection.
Bioelectromagnetics 34:31-42.
11. Yang L, Chen Q, Lv B, et al. 2017. Long-Term Evolution Electromagnetic Fields Exposure
Modulates the Resting State EEG on Alpha and Beta Bands. Clin EEG Neurosci 48:168-175.
12. Lv B, Su C, Yang L, et al. 2014. Whole brain EEG synchronization likelihood modulated by long
term evolution electromagnetic fields exposure. Conf Proc IEEE Eng Med Biol Soc. 2014:986-989.
13. Mizuno Y, Moriguchi Y, Hikage T, et al. 2009. Effects of W-CDMA 1950 MHz EMF emitted by
mobile phones on regional cerebral blood flow in humans. Bioelectromagnetics 30:536-544.
Figures

Figure 1. (A) EEG recording system during the LTE-related EMF exposure. (B) Time course of
the experiment.

704
Figure 2. Power spectra of EEG at F3, F4, C3, C4, P3, P4, O1 and O2 before, during, and
after exposure.

Figure 3. Power spectra of EEG at T3 and T4 before, during, and after exposure.

705
PB-90 [14:30]
STUDENT PAPER
Alpha band analysis in resting EEG after exposure to radiofrequency signal (900 MHz):
EEG/MEG study combined with an anatomic MRI
Jasmina Wallace1, 2, Lydia Yahia-Cherif3, 4, Laurent Hugueville3, 4, Christophe Gitton3, 4 & Brahim
Selmaoui1, 2
1Department of Experimental Toxicology, Institut National de l'Environnement Industriel et des Risques
(INERIS), Verneuil-en-Halatte, France
2Peritox Laboratoire de Périnatalité & Risques Toxiques, Unité mixte INERIS, Université de Picardie Jules
Verne, CURS, Amiens, France
3Centre De NeuroImagerie De Recherche - CENIR, Institut Du Cerveau Et De La Moelle Épinière - ICM,
Paris, France
4Inserm U1127, CNRS UMR 7225, Institut Du Cerveau Et De La Moelle Épinière - ICM, Paris, France
Keywords: Human, RF/Microwaves, Work in Progress
Presented by: Jasmina Wallace
With the extensive use of mobile phones (MP) several studies have been realized to understand the
effects of radiofrequency exposure on brain activity. The results show changes in the alpha band
spectral power. To better understand the cortical structures involved in these changes after MP
exposure (900 MHz), we carried out electroencephalography and magnetoencephalography
recording followed by the anatomical magnetic resonance imaging on healthy volunteers.
Participants were selected according to inclusion criteria. They were healthy males or females, aged
between 18 and 35, non-smokers and with regular sleep habits. Data analyses of 10 subjects are
still under process and ready results will be presented at the conference.
Introduction
The extensive use of mobile phones (MP) and the consecutive increase of exposure to
radiofrequency electromagnetic fields (RF) have raised concerns about the effects of RF on human
health and brain activity. Electroencephalography (EEG) has been largely reported in the literature
and the most consistent effect was seen on the alpha rhythm of the waking EEG. However, changes
observed in the alpha rhythm amplitude were sometimes an increase [1, 2] and sometimes a
decrease [3, 4], while a few studies reported no effects [5]. For this reason, EEG and
magnetoencephalography (MEG) recording, followed by the anatomical magnetic resonance
imaging (MRI), were used to first localize the active brain area with a millimeter spatial and temporal
accuracy, and then to follow the brain connectivity and dynamic during and after a MP exposure, to
better understand the effect of this exposure and the reason why alpha band power reacts
differently.
Aims
The aim of this study is to better understand the alpha band modification and the cortical structures
involved in these changes, due to the RF exposure on healthy volunteers. We investigate the effect
of a GSM (900 MHz) mobile phone exposure on the brain electrical activity during resting
wakefulness. This one is examined through EEG and MEG recordings under controlled
experimental parameters. The combined EEG and MEG recordings with an anatomical MRI allow us
to realise a high resolution spatio-temporal cortical brain neuroimaging and to analyse the whole
cerebral signal dynamic after the RF exposure.
Materials and Methods
So far, the study was conducted on 10 healthy volunteers who completed the whole study but the
analyses are still underway. The objective is to reach at least 26 volunteers at the end of the study.
Participants are selected according to inclusion criteria. They are males or females, aged between

706
18 and 35, they have regular sleep habits, no medication, no chronic diseases or disability, no
recent acute illness and no neurological or psychiatric illness. All participants are right-handed, non-
smoker and they are asked to abstain from consuming alcohol and caffeine 24 hours before each
session. They are asked to not use their personal MP at least 2 hours before the beginning of
experiment and to maintain a regular sleep-wake cycle at least one week before experiment.
Women without oral contraceptive use participate only during the follicular phase of their menstrual
cycle. The participants attend two EEG/MEG recording sessions, separated one week apart,
through a crossover randomized, double-blind, and counterbalanced experimental design. A final
anatomical MRI visit is scheduled at the end of the two sessions. The exposure system is a sham or
a real GSM MP (Nokia 6650, 900 MHz RF, 217 Hz pulse rate, maximum specific absorption rates
were 0, 93 W/kg), that is placed against the left ear. Each recording session is composed of three
phases (Figure 1): the baseline, which is constituted by a 12-min MEG recording and a 12-min EEG
recording; the exposure period with a 26-min EEG recording, and the post-exposure period with a
26-min EEG and MEG combined recording. Each recording phase is characterised by the
alternation of 3-min eyes opened (EO) sequence and 3-min eyes closed (EC) sequence. After each
6-min EO/EC run there is a 30-sec pause. EEG data are recorded using BrainCap (EASYCAP
Products) with 74 electrodes placed according to the international 10/20-system. The cortical
magnetic field are recorded thanks to the Elekta Neuromag Triux MEG system. In addition, the ECG
and the EOG are simultaneously recorded during the whole sessions. Furthermore, at the beginning
of each phase and at the end of the session, a salivary sample is collected with Salivette device
(Sarstedt). From these samples, control immune-enzymatic and proteomic analysis, as cortisol,
alpha-amylase and caffeine, are realised. The participants attend a final anatomic MRI visit
(Siemens, Verio), to provide a brain anatomic scanner, that can significantly improve the workflow
and accuracy of localizing brain activity.
Results
The experimental sessions are ongoing for the remaining subjects. Data analyses of 10 subjects are
still under process and the ready results will be presented at the conference.
Conclusions
This project considers for the first time the recording of EEG combined to the MEG with additional
anatomic MRI, in order to investigate the effect of GSM MP exposure. The MEG is a neuroimaging
technique that allows to measure directly the neural activity with a very high temporal resolution. It
uses an array of sensors positioned over the scalp that are extremely sensitive to minuscule
changes in the magnetic fields produced by small changes in the electrical activity within the brain.
In contrast to the functional MRI, the MEG platform provides real-time information of the neural
involvement, and, unlike the EEG technics, it is less influenced by the changes of electric
conductivity within the brain, skull and scalp. Through EEG and MEG recording and the anatomic
MRI, it is possible to localize the active brain area with a millimeter spatial accuracy and with an
excellent temporal accuracy. So, with this kind of experimental approach we are able to follow the
brain connectivity and dynamic during and after a MP exposure, to better understand the effect of
this exposure and the reason why alpha band power reacts differently.
References
1. Croft RJ, Leung S, McKenzie RJ, Loughran SP, Iskra S, Hamblin DL, et al. Effects of 2G and 3G
mobile phones on human alpha rhythms: Resting EEG in adolescents, young adults, and the elderly.
Bioelectromagnetics. 2010;31(6):434-44.
2. Curcio G, Ferrara M, Moroni F, D'Inzeo G, Bertini M, De Gennaro L. Is the brain influenced by a
phone call? An EEG study of resting wakefulness. Neuroscience research. 2005;53(3):265-70.
3. Ghosn R, Yahia-Cherif L, Hugueville L, Ducorps A, Lemarechal JD, Thuroczy G, et al.
Radiofrequency signal affects alpha band in resting electroencephalogram. Journal of
neurophysiology. 2015;113(7):2753-9.
4. Perentos N, Croft RJ, McKenzie RJ, Cosic I. The alpha band of the resting electroencephalogram

707
under pulsed and continuous radio frequency exposures. IEEE transactions on bio-medical
engineering. 2013;60(6):1702-10.
5. Loughran SP, Benz DC, Schmid MR, Murbach M, Kuster N, Achermann P. No increased
sensitivity in brain activity of adolescents exposed to mobile phone-like emissions. Clinical
neurophysiology. 2013;124(7):1303-8.
Figures

Figure 1. In a double-blind, randomized and counterbalanced crossover design, participants

undergo two different exposure conditions, sham or real mobile phone (MP), at weekly
intervals. The protocol for each recording session is composed of three phases: a baseline,
which is constituted by a 12-min magnetoencephalography (MEG) recording and a 12-min
electroencephalography (EEG) recording; the exposure period with a 26-min EEG recording,
and the post-exposure period with a 26-min EEG and MEG combined recording. The whole
session is characterised by 3-min eyes opened (EO) sequences and 3-min eyes closed (EC)
sequences. After each 6-min run there is a 30-sec pause. The electrocardiogram (ECG) and
the electrooculogram (EOG) are simultaneously recorded. At the beginning of each phase and
at the end of the session, a salivary sample (SAL) is collected.

PB-92 [14:30]
Antenna with directional radiation pattern used in microwave ablation
Pawel Cala1 & Pawel Bienkowski1
1Wroclaw University of Technology, Wroclaw, Poland, 50-370
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Pawel Cala
Microwave (MWA) ablation with a double slot constructed on the basis of a coaxial antenna with a
matching load. Simulations were carried out in the CST environment. It has been shown that it is
708
possible to contruct a microwave ablation antenna with a directional radiation pattern providing a
VSWR <1.5 for the ISM 2.4GHz band. Directionality of the antenna will allow heating only selected
tissues without overheating neighboring ones. Such a solution can definitely improve the treatment
process and not expose the patient to postoperative complications (burns). In further work, the
authors will focus on producing a prototype of one antenna and performing a simulation of an
antenna system consisting of at least three microwave antennas
I. Introduction
Microwave ablation is a relatively new but dynamically developing technique that involves the use of
electromagnetic waves by using one or more antenna within the tissue tumor changes. After
entering the antenna is setting the appropriate time and power of ablation The removal of cancerous
tissue using microwave devices (MWA) is a process of heating the tissue to a level above 55°C,
which causes its permanent damage. Microwave ablation has gained popularity in recent years.
MWA, RF and cryo-stimulation are an alternative treatment option for patients for whom normal
resection of malignant tissue is not possible due to insufficient amount of healthy tissue or tumor
placement, number of tumor spots or other comorbidities. Microwave heating works on the principle
of changes in the position of water molecules in the body. Molecules quickly change orientation
under the influence of EMF to adjust polarization and consequently produce heat through friction.
Microwaves are also used, for example, in a popular household appliance, i.e. Microwave oven,
which is used to heat food meal. The same microwaves are used in antennas for microwave
ablation. The process of heating the object looks equally. The main difference is the place of
heating, in the case of microwave ablation, heating takes place by means of a thin antenna
laparoscopically, percutaneously or carried out by means imaging techniques, such as computed
tomography or ultrasonography. It is worth noting that the effectiveness of the microwave ablation
method strongly depends on the physical properties of the heated tissue, primarily water, which
determines the value of the relative permittivity. Microwave ablation has been recognized as the
most promising technique for the treatment of bone tumors. Thermal ablation antennas work mainly
in the unlicensed ISM 2.4 GHz band. The dielectric properties of tissues are a key factor for the
increase in temperature. Microwave ablation (MWA) can easily penetrate bone tissue , which makes
this technique highly recommended for the treatment of bone tumors. However, its use in the
treatment of bone cancer needs better understanding in long-term effects and benefits. The
disadvantage of antennas for microwave ablation is that they can cause the destruction of healthy
neighboring tissues. The evaporation of water during the procedure changes the dielectric
permittivity of tissue, which changes the impedance matching of the antenna and thus the amount of
power emitted to the object during the treatment. The main benefit of using microwave ablation is:
a) treatment of patients who are not meeting the conditions of the operation,
b) is well tolerated by patients,
c) can be used as a complementary treatment,
d) it is cheaper.
Nevertheless, there are some weaknesses of that method of treatment:
a) it is impossible to treat regions near blood vessels or healthy organs,
b) the generated heat can cause burns and blisters,
c) it is advisable to use a 3D scanner to bring the antenna to the tumor.
II. Modeling and simulations
Antenna simulation was carried out in the CST Microwave 2016 environment. The diameter are
respectively 2.2 mm (Figure 2). Such values have been adopted because they correspond to the
dimensions of the real semi-rigid coax from which the prototype of the antenna will be made. The
antenna is completely covered with a PTFE protective coating with a thickness of 0.5 mm. The
simulation prototype consists of two radiant slots with an angle of 280 °. The antenna constructed in
such a way allows to radiate EM energy in a chosen direction, which allows for partial heating of the

709
tumor tissue without excessive overheating of other healthy human organs. Such a solution opens
the way to building an antenna system consisting of two or more antennas. In order to obtain a
better matching of the antenna to 50Ω, an artificial load was simulated (Figure 1). The simulated
antenna works in the unlicensed ISM 2.4GHz band. In the simulation, an impedance matching was
added to improve the circuit parameters of the antenna. Antenna dimensions are shown in Figure 1.
Authors also working on directional antenna working in ISM 2.4GHz without additional matching
load which in consequence implement additional difficulties during the fabricating prototype.
Fig. 1. Microwave ablation atenna inside the body with tumor
Figure 2 shows an example of the arrangement of three antennas with angular radiation pattern
around the tumor. This solution will eliminate the backward radiation of antennas (A1, A2, A3) and
focus EMF energy on the tumor inside the antenna array. Conducted simulations for the above
parameters. The use of matching (50Ω) at the end of a semi-rigid conductor improves antenna
parameters by 6dB in the ISM 2.4GHz band. The use of impedance matching also causes the
phenomenon of wave guidance, this may lead to excessive heating of the antenna, which will be
tested in further stages of operating. For the proposed model, a targeted radiation pattern in a
selected direction was obtained. From lateral view of the tested antenna model EMF distribution it is
clear that the antenna has properties to concentrate energy in a chosen direction. The S11 factor is
smaller than -14dB, which in practice means that the VSWR antenna in the 2352MHz - 2451MHz
band is smaller than 1.5 (0.2 dB power loss). Figure 3 shows the theoretical radiation pattern inside
the bone. Whole calculations were made with a implemented Pennes equation.
Fig. 2. Example of thee element antenna array
Fig. 3. Cross section of the antenna inside the bone of the dimaeter 30mm
III. Conclusion
Microwave (MWA) ablation with a double slot constructed on the basis of a coaxial antenna with a
matching load. Simulations were carried out in the CST Microwave Studio environment. It has been
shown that it is possible to contruct a microwave ablation antenna with a directional radiation pattern
providing a VSWR <1.5 for the ISM 2.4GHz band. Directionality of the antenna will allow heating
only selected tissues without overheating neighboring ones. Such a solution can definitely improve
the treatment process and not expose the patient to postoperative complications (burns). In further
work, the authors will focus on producing a prototype of one antenna and performing a simulation of
an antenna system consisting of at least three microwave antennas. By means of control of the
signal phase of each antenna, it will be possible to obtain any radiation pattern within the tumor. This
will allow more control of the heating of the selected part of the object during the procedure. The
problem under investigation is extremely complicated and complex, small changes in parameters
during prototype making can have a significant impact on impedance matching or antenna radiation
pattern, therefore in further subjects the influence of as many antenna parameters as possible on
field and peripheral parameters will be determined. The dimensions of the antenna shown provide
the most optimal results while maintaining good reflectance and directionality of the microwave
antenna.
Figures

710
Figure 1. Proposed microwave directional antenna

Figure 2. Example of thee element antenna array

711
 Figure 3. Cross section of the antenna inside the bone of the dimaeter 30mm

PB-94 [14:30]
Visualisation of an nsPEF induced calcium wave using the genetically encoded calcium
indicator GCaMP in U87 human glioblastoma cells
Lynn Carr1, Sylvia M. Bardet1 & Rodney P. O'Connor2
1University of Limoges, Limoges, France, F-87000
2IMT Mines Saint-Etienne , Gardanne, France, F-13541
Keywords: In vitro, Pulsed, Completed (published)
Presented by: Lynn Carr
This study looks at the application of genetically encoded calcium indicators (GECIs), to investigate
the spatiotemporal nature of nsPEF-induced calcium signals, using fluorescent live cell imaging.
GECIs can be targeted to different cellular locations and, by using a plasma membrane expressed
GECI, we demonstrate calcium entry into the cell as a wave, initiated at the cathode side, following
nsPEF treatment. This wave was masked by commercially available chemical, cytosolic, calcium
indicators.
Introduction
Calcium ions (Ca2+) play a key role in a wide range of cellular processes, with elevation of
intracellular Ca2+ concentration ([Ca2+]i) being a significant signalling event. An increase in [Ca2+]i

712
following application of nanosecond pulsed electric fields (nsPEF) has been widely reported[1].
To date all nsPEF induced changes in [Ca2+]i, measured using live cell imaging, have used chemical
calcium indicators, that are generally cytosolic, including FLUO-4[2–4], Fura-2[5–7] and Calcium
Green[8–14]. Genetically encoded calcium indicators (GECIs), such as GCaMP[15], provide an
alternative to chemical indicators and they have yet to be used to investigate the nsPEF calcium
response. They present several advantages over chemical indicators, including their use straight
from the incubator, removing any negative effects of a room temperature indicator loading protocol.
Being easy to target to specific locations within the cell, they could allow a more in depth
investigation of the nsPEF response on particular compartments.
In this study we demonstrate the application of GECIs to investigate the nsPEF-induced calcium
response. We compare the plasma membrane targeted GCaMP5-G and a cytosolically expressed
GCaMP6-S in nsPEF studies with two commonly used cytosolic chemical calcium indicators (Fura
Red and FLUO-4). With GCaMP5-G we show a wave of calcium entry into the cell following nsPEF
treatment, which is not observed with any of the cytosolic indicators, highlighting the potential
interest of being able to target GECIs to specific regions of the cell especially when no chemical
alternative is available.
Methods
Changes in [Ca2+]i were visualised in U87 cells that either expressed a GECI (GCaMP5-G or
GCaMP6-S), or were loaded with a chemical indicator (1.25 µM Fura Red, AM, 0.5 µM FLUO-4, AM
or 1 µM of FFP-18). Cells were imaged, in a standard HBSS solution, by epifluorescence using a
Leica DMI6000 microscope with a 100x objective. Fluorescent excitation was provided by a Spectra
7 light engine (Lumencor). Emitted light was filtered and captured on an electron-multiplying charge-
coupled device camera (EMCCD; Photometrics Evolve 512, Roper) with 512 x 512 pixels.
Using two steel electrodes separated by a gap of 1.2 mm, with 50 Ω impedance in parallel[16] and
an nsPEF generator (FPG 10-1NM-T, FID Technology, Germany) with 50 Ω output impedance, cells
were exposed to 10 ns pulses, with an electric field intensity of 44 kV/cm.
Results
When 100, 10 ns pulses were applied at 10 Hz all three of the cytosolic calcium indicators (FLUO-4,
Fura Red and GCaMP6-S) showed a global, uniform increase in intracellular calcium. The
membrane bound GCaMP5-G however showed a spatially focused increase in fluorescence that
initiated at the cathode before spreading as a wave across the cell (see figure 1). The speed (in
µm/s ± SE) that this wave travelled across the cell was dependent on the frequency of pulse
application: 100P 100 Hz: 7.23 ± 0.89 (n = 4); 100P 10 Hz: 2.65 ± 0.38 (n = 4); 100P 1 Hz: 0.40 ±
0.04 (n = 4). The wave was not propagated by calcium induced calcium release, but could be
recreated in a cell loaded with a membrane bound chemical indicator.
Discussion and conclusion
Whilst synthetic calcium indicators are easy to use, we show in this study an advantage of using
GECIs for studying nsPEF calcium responses. By targeting the GECI to the plasma membrane we
were able to demonstrate the influx of extracellular calcium in a spatiotemporal wave, which was
masked when cytosolic indicators were used. As all calcium indicators are essentially calcium
buffers, they have a perturbing effect on the underlying calcium signal and modelling experiments
have shown that mobile buffers are able to eliminate spatial calcium waves, whereas immobile
buffers have little effect[17]. Targeting GECIs to other locations within the cell, for example
intracellular calcium stores, could be used to further understand the calcium response to nsPEF.
Acknowledgement
This project has received funding fromthe European Union’s Horizon 2020 research and innovation
programme under grant agreement No 737164 SUMCASTEC. Our research was conducted in the
scope of LEA-EBAM, a European Associated Laboratory entitled, “Pulsed Electric Fields

713
Applications in Biology and Medicine”.
References
[1] Batista Napotnik T, Reberšek M, Vernier PT, Mali B, Miklavčič D. Effects of high voltage
nanosecond electric pulses on eucaryotic cells (in vitro): A systematic review. Bioelectrochemistry
2016;110:1–12. doi:10.1016/j.bioelechem.2016.02.011.
[2] Beebe SJ, Chen Y-J, Sain NM, Schoenbach KH, Xiao S. Transient features in nanosecond
pulsed electric fields differentially modulate mitochondria and viability. PLoS One 2012;7:e51349.
doi:10.1371/journal.pone.0051349.
[3] Scarlett SS, White JA, Blackmore PF, Schoenbach KH, Kolb JF. Regulation of intracellular
calcium concentration by nanosecond pulsed electric fields. Biochim Biophys Acta
2009;1788:1168–75. doi:10.1016/j.bbamem.2009.02.006.
[4] Beebe S, Sain N, Ren W. Induction of Cell Death Mechanisms and Apoptosis by Nanosecond
Pulsed Electric Fields (nsPEFs). Cells 2013;2:136–62. doi:10.3390/cells2010136.
[5] Semenov I, Xiao S, Pakhomov AG. Primary pathways of intracellular Ca(2+) mobilization by
nanosecond pulsed electric field. Biochim Biophys Acta 2013;1828:981–9.
doi:10.1016/j.bbamem.2012.11.032.
[6] Semenov I, Xiao S, Pakhomova ON, Pakhomov AG. Recruitment of the intracellular Ca2+ by
ultrashort electric stimuli: the impact of pulse duration. Cell Calcium 2013;54:145–50.
doi:10.1016/j.ceca.2013.05.008.
[7] Ford WE, Ren W, Blackmore PF, Schoenbach KH, Beebe SJ. Nanosecond pulsed electric fields
stimulate apoptosis without release of pro-apoptotic factors from mitochondria in B16f10 melanoma.
Arch Biochem Biophys 2010;497:82–9. doi:10.1016/j.abb.2010.03.008.
[8] Ibey BL, Ullery JC, Pakhomova ON, Roth CC, Semenov I, Beier HT, et al. Bipolar nanosecond
electric pulses are less efficient at electropermeabilization and killing cells than monopolar pulses.
Biochem Biophys Res Commun 2014;443:568–73. doi:10.1016/j.bbrc.2013.12.004.
[9] Tolstykh GP, Beier HT, Roth CC, Thompson GL, Ibey BL. 600Ns Pulse Electric Field-Induced
Phosphatidylinositol4,5-Bisphosphate Depletion. Bioelectrochemistry 2014;100:80–7.
doi:10.1016/j.bioelechem.2014.01.006.
[10] Roth CC, Tolstykh GP, Payne J a, Kuipers M a, Thompson GL, DeSilva MN, et al. Nanosecond
pulsed electric field thresholds for nanopore formation in neural cells. J Biomed Opt
2013;18:035005. doi:10.1117/1.JBO.18.3.035005.
[11] Thompson GL, Roth CC, Dalzell DR, Kulpers M, Ibey BL. Calcium influx affects intracellular
transport and membrane repair following nanosecond pulsed electric field exposure. J Biomed Opt
2014;19:055005.
[12] Vernier PT, Sun Y, Chen M-T, Gundersen MA, Craviso GL. Nanosecond electric pulse-induced
calcium entry into chromaffin cells. Bioelectrochemistry 2008;73:1–4.
doi:10.1016/j.bioelechem.2008.02.003.
[13] Craviso GL, Choe S, Chatterjee I, Vernier PT. Modulation of intracellular Ca2+ levels in
chromaffin cells by nanoelectropulses. Bioelectrochemistry 2012;87:244–52.
doi:10.1016/j.bioelechem.2011.11.016.
[14] Beier HT, Roth CC, Tolstykh GP, Ibey BL. Resolving the spatial kinetics of electric pulse-induced
ion release. Biochem Biophys Res Commun 2012;423:863–6. doi:10.1016/j.bbrc.2012.06.055.
[15] Akerboom J, Chen T-W, Wardill TJ, Tian L, Marvin JS, Mutlu S, et al. Optimization of a GCaMP
calcium indicator for neural activity imaging. J Neurosci 2012;32:13819–40.
doi:10.1523/JNEUROSCI.2601-12.2012.
[16] Wu Y-H, Arnaud-Cormos D, Casciola M, Sanders JM, Leveque P, Vernier PT. Moveable wire
electrode microchamber for nanosecond pulsed electric-field delivery. IEEE Trans Biomed Eng
2013;60:489–96. doi:10.1109/TBME.2012.2228650.

714
[17] Tsai J-C, Sneyd J. Are buffers boring? Uniqueness and asymptotical stability of traveling wave
fronts in the buffered bistable system. J Math Biol 2007;54:513–53.
doi:10.1007/s00285-006-0057-3.
Figures

Figure 1. Spatiotemporal effect on intracellular calcium of 100, 10 ns pulses applied at

10 Hz to U87 cells visualised with different calcium indicators. Representative live cell
images of the response of different calcium indicators to nsPEF treatment as measured in U87
cells. Cells either stably expressed a form of the genetically encoded calcium indicator
GCaMP (membrane bound GCaMP5-G or cytosolic GCaMP6-S) or were loaded with a
chemical indicator (cytosolic indicators FLUO-4 and Fura Red). Fura Red fluorescence
decreases with increasing calcium, the other indicators show an increase in fluorescence.
Images have been pseudo-coloured for clarity. In all experiments the cathode is located in the
top right corner and the anode in the bottom left corner.

PB-96 [14:30]
Local high frequency electromagnetic field, nanoparticles and piezoelectric acoustic
exposition: a potential therapeutic approach in glioblastoma treatment
Flavia Del Vecchio1, Elodie Jobard1, Samir Dekali1, David Crouzier1 & JClaude Debouzy1, 2
1Département EBR/Unité des Risques Technologiques Emergents, Institut de Recherches Biomédicales du
Service de Santé des Armées, Brétigny sur Orge, France, 91223
2CEA-CLINATEC, Grenoble, France, 38000
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Flavia Del Vecchio

715
Our aim was to study the possible therapeutic effects of a co-exposition for 72h to nanoparticles
(SWCNTs), electromagnetic field (EMF, 100kHz, 10V square signal) at non-thermal level and
piezoelectric acoustic emission to have a synergic local toxic effect on an in vitro model of
Glioblastoma. C6 rat glioma cells were randomly assigned to four groups: Sham/Sham, Sham/Nano
exposed Sham/Field exposed, and Nano/Field exposed. Results showed a decrease in the number
of the cells co-exposed compared to the control group. However, the average survival rate of the
cells was still high over the whole sample; and this combination could be interpreted either as all or
nothing local effect, or more diffuse inhibition of the cell proliferation.
Introduction
Glioblastoma (GBM) is the most aggressive malignant brain tumor, with a survival time post-
diagnosis of 15 months, a high recurrent rate and a mortality rate exceeding 95% due to the high
infiltrating properties of the glioma cells.
Common treatments for GBM include surgical resection, radiotherapy, chemotherapy, or
combinations of these three therapeutic approaches (1).
Despite the existence of those different therapeutic treatments, the evolution of the post-diagnosis
mean survival only increased from 13 to 15 months, highlighting the need of more effective therapy.
Several biophysics approaches with a special interest in electromagnetic fields and nanoparticles,
have arisen recently.
For instance the studies performed by Kirson`s group showed a decrease of the malignant cell
proliferation thanks to disintegration of the mitotic spindle using alternative electric fields (also called
TTFields for Tumor Treating Field, 100 kHz for mouse melanoma, 150 kHz for human breast
carcinoma, 200 kHz for rat glioma) with only minor side effects in animal tumor models and human
brain tumors (2,3).
The Magforce AG company, used alternative magnetic field (100 kHz) inducing the vibration of
metallic nanoparticles, resulting in local temperature increase leading to the tumor cells death.
The main drawback for both ED. Kirson and Magforce AG methods was the distance between the
field source (electrodes on the skin or use of external magnetic emitter) and the target and the lack
of specificity.
An interesting approach was used by GK Lewis`s group. (4). Acoustic waves (ultrasounds) were
used to increase the perfusion of locally delivered Evans blue into a brain model (agar gel), an
equine brain and an avian muscle tissue counting for a possible clinic use in order to increase drugs
penetration maintaining a therapeutically useful concentration. More recently, Mano’s group (5)
applied intraparenchymal ultrasound to obtain high drugs concentration directly into a targeted
lesion beyond the blood-brain barrier in rodent and nonhuman primate brain.
On the base of those results, the aim of our study was to evaluate the effects of a co-exposition to
nanoparticles (SWCNTs), electromagnetic field (EMF) at non-thermal level and piezoelectric
acoustic emission to have a synergic local toxic effect on tumor cells.
Methods
C6 glioma cells (from rat`s brain) were co-exposed in a vitro culture for 72h (cell growth and survival
time) using a new patent pending probe.
The probe (fig.1a) was made by three parts: electric, magnetic and piezoelectric. The source of
electromagnetic fields was an Agilent 33120A generator, giving a 100 kHz, 10V square signal,
applied to the probe via coaxial BNC link (ground and line). The EMF was evaluated using a PMOR
03 field meter gave Emax of 15V/m and Hmax of 30A/m in the immediate vicinity of the probe tip.
The temperature control was realized using a FLIR A320 Temperature screen camera. No
temperature increase exceeding 0.2°C was detected, precluding any thermal effect.
The C6 rat glial cells were stored in flasks and randomly assigned to one of the following four
different groups: Sham / Sham, Sham / Nano exposed, Sham / Field exposed, and Nano / Field
exposed. All the cultures were treated in the same conditions (with the probe into the flask) as the
716
double exposed one (except during the sham field exposition, in which the generator was turned
off). The initial C6 cell population was 0.2 * 106 cells per flask. 72h exposure starts immediately after
cells passages. Sham / Sham and Sham / Field groups received 120µl of the stock sham solution
(without SWCNTs), the double exposed and the Sham / Nano groups received 120µl of the Stock
solution containing the SWCNTs. After the exposure, in order to eliminate the SWCNTs in solution,
each flask was rinsed 3 times with Phosphate Buffered Saline. The cells were chemically detached
from the flask using trypsin EDTA. C6 cells were centrifuged 5 min at 300G. The cells pellets were
resuspended in 1 ml cell culture medium. 50 µl of C6 cells solution were added to 50 µl of Trypan
Blue. C6 cells were counted by a TC20TM automatic cell counter.
Results
The cell count results are expressed in percentage calculated on the Sham / Sham group count.
The results showed a decrease in the number of the cells co-exposed compared to the control
group (51% for the Nano / Field group; 36% for the Sham / Nano group; 10% for the Sham / Field
group). The survival rate, evaluated in the whole observation area, was maximal for the Sham /
Sham control group and decreased after the exposition to SWCNTs and EMF (99% survival rate for
the Sham / Field group, 97% survival rate for the Sham / Nano group and 95% survival rate for the
Nano / Field group). It should be noted that, on the observation window, the exposed samples were
rather heterogenous, including normal areas and almost decertified regions.
Discussion
These preliminary results showed a decrease of the C6 population number after the co-exposition,
counting in favor of a synergic effect of the SWCNTs and the EMF.SWCNTs could increase the
toxicity of the EMF by focusing them and inducing an electromagnetic field increase, moreover EMF
could increase the toxicity of the SWCNTs by an augmentation of their dispersion and penetration
into the cells. Both those effects would be magnified by ultrasound exposure. However, the apparent
high cells survival rate was related to the heterogeneity of the samples. Hence, the observed area
included peripherical regions that were close to normal and completely decertified areas, i.e. without
living cell. Moreover, such a coexposition effect would be consistent both with cell destruction or
inhibition of the cell proliferation (2).
Conclusions
The limited number of samples used was not sufficient to draw a conclusion.
Anyway our results showed a promising in vitro response. This study will lead to further in vitro
investigations then to animal experimentations (on rat and then on hogs) with a miniaturized
implantable probe (fig.1b) that could confirm the preliminary results observed in vitro and the safety
of such therapy for a possible human use.
References
1. Stupp R, Hegi ME, Mason WP, van den Bent MJ, Taphoorn MJB, Janzer RC, et al. Effects of
radiotherapy with concomitant and adjuvant temozolomide versus radiotherapy alone on survival in
glioblastoma in a randomised phase III study: 5-year analysis of the EORTC-NCIC trial. Lancet
Oncol. 2009 May;10(5):459–66.
2. Kirson ED, Dbalý V, Tovaryš F, Vymazal J, Soustiel JF, Itzhaki A, et al. Alternating electric fields
arrest cell proliferation in animal tumor models and human brain tumors. Proc Natl Acad Sci. 2007
Jun 12;104(24):10152–7.
3. Kirson ED, Giladi M, Gurvich Z, Itzhaki A, Mordechovich D, Schneiderman RS, et al. Alternating
electric fields (TTFields) inhibit metastatic spread of solid tumors to the lungs. Clin Exp Metastasis.
2009;26(7):633–40.
4. Lewis Jr GK, Schulz ZR, Pannullo SC, Southard TL, Olbricht WL. Ultrasound-assisted
convection-enhanced delivery to the brain in vivo with a novel transducer cannula assembly:
Laboratory investigation. J Neurosurg. 2012;117(6):1128–40.

717
5. Mano Y., Saito R., Haga Y., Matsunaga T., Zhang R., Chonan M., Haryu S., Shoji T., Sato A.,
Sonoda Y., Tsuruoka N., Nishiyachi K., Sumiyoshi A., Nonaka H., Kawashima R., Tominaga T.
Intraparenchymal ultrasound application and improved distribution of infusate with convection-
enhanced delivery in rodent and nonhuman primate brain. J Neurosurg. 2016 May;124(5):1490-500.
doi: 10.3171/2015.3.JNS142152. Epub 2015 Oct 23.
Figures

Figure 1. The figure represents the non implantable (a) and the miniaturized implantable probe
(b) made by three parts: electric, magnetic and piezoelectric.

PB-98 [14:30]
In-vitro study of current induced ventricular fibrillation in rabbits
Michael Koch1, Michael Bartonek1, Jiri Silny3, Bruno Podesser2 & David Santer2
1Innovation,
Codes and Standards, EATON Industries GmbH, Vienna, Austria, 1210
2Department of Biomedical Research, Medical University of Vienna , Vienna, Austria, 1090
3FEMU, RWTH Aachen, Aachen, Germany, 62052
Keywords: In vitro, ELF/LF, Completed (unpublished)
Presented by: Michael Koch
Ventricular fibrillation due to an electric current is one of the most dangerous incident which can
occur in an electrical system. In this paper results on the investigation of current induced ventricular
fibrillation at different frequencies (up to 100kHz) are reported. One of the main targets was to
develop a flexible and adaptable experimental system based on in-vitro experiments of the isolated
rabbit heart within a working heart model approach. In the chosen setup the heart can be electrically
stimulated, driven into a sustained fibrillation state and can be defibrillated afterwards. The research
project is funded by the Austrian Government (FFG, project no. 841285).
Introduction
Ventricular fibrillation caused an electric current is the most dangerous failure case which can
happen in an electrical system. Therefore RCD's/Groundfault-interruptor where developed to
prevent such harmful events. Contemporary knowledge of electrocution is based on historical
animal experiments, experiments on corpses and some self- experiments (e.g. Prof. Dr. G.
Biegelmeier [1]). The problem is that these investigations were not done in a frequency range from

718
DC up to 100kHz. But this is important because of the huge amount of electronics in nowadays life.
Modern electrical consumers lead to residual currents up to 100 kHz. Frequency converters which
are used in washing machines, elevators and also in big industrial applications can be a source of
non 50/60Hz sine waves and also of arbitrary signals or even DC. In this project the current induced
ventricular fibrillation at different frequencies (up to 100kHz) are investigated. One of the big
challenge to explore the threshold values for ventricular fibrillation at different singular frequencies
and also for arbitrary signals. In addition the elecrric field vector needs to be taken into account. To
evaluate the threshold values it is necessary to develop a dedicated procedure which provides a
standardized way for this purpose.. One of the main targets was to develop a very flexible and
adaptable experimental setup. This is the reason why is is also possible to investigate other settings
e.g. the effects of non-lethal weapons, like Tasers. Because of the setting it is also possible to apply
chemical treatments (e.g.drugs) to the heart during or before the experiments) and take other pre-
conditioning of the heart into account (e.g. infacrtion).
Methods
An experimental setup is provided which fulfills all requirements for stimulation, recording of different
heart parameters and test repeatability (Fig. 1).
The experiments are done on rabbit hearts in a working heart model [2], [3]. Due to the high
complexity of the heart itself direct to simulation is not possible and research has to be based on
experiments. In this setup the heart is stimulated, driven into a sustained fibrillation and finally
defibrillated. Stimulation is accomplished by a homogeneous electrical field, similar as in the case of
an electric shock. Stimulation is performed by sinusoidal currents at dedicated frequencies.
withincreasing amplitudfe until ventricular fibrillation starts. Additional experiments enable the
determination of the field density inside of the heart and the dependency of different stimuli (in the
range from DC up to 150kHz and at loads up to 200V and up to 28A).
We are able to:
Perform isolated heart experiments (up to 4h)
Measure the ECG on the heart, the pressure in the heart, the blood flow outside of the heart
Measure the electrical fields inside and outside of the heart
Do chemical analysis for control the parameters of the blood (pH, O2, CO2,…)
Stimulate and take control of the beat of the heart
Take the heart controlled into ventricular fibrillation (at different frequencies) or apply different
stimuli (e.g. Taser signal)
Defibrillate the heart (so we can do several experiments on one heart)
The hearts are placed into a specific frame (which also was developed by EATON Vienna) which
fulfills the necessary boundary conditions for the stimulation tests (Fig. 2).
To determine the ventricular fibrillation threshold a current of a distinct frequency and amplitude was
applied (before the P-wave) for 3 seconds, followed by a pause of 20-90 seconds. If no fibrillation
was detected the amplitude was increased by 3mA in the next stimulation step. If ventricular
fibrillation occurred, two cases have been distinguished. First sustained fibrillation (if it lasts for more
than 5 seconds; this value was derived from the fact that an ICD would fire if the device detects
failing ECG cycles for more than 10 ECG periods) or ventricular fibrillation with spontaneous
remission. If fibrillation was detected the defibrillator was connected to the stimulus electrodes and a
shock-signal was applied. Usually 1-3 shocks were necessary to terminate the fibrillation. After a
period of approx. 5 minutes, if all the parameters were in the expected range, the fibrillation
experiment started again. In a standard experimental plan 5-7 fibrillation experiments were
performed. The specific condition whether the experiments were continued was the shape of the
ECG, the flow, the pressure and also the calculated coronary flow.
One of the big challenges is to find a way to transform values from animal heart experiments to the
human heart. For this we work together with Prof. Silny (an emeritus from the University of Aachen,
Germany). We have developed a method which begins by in-vitro experiments of an isolated heart

719
and leads over a translation mechanism to the human heart to finite-element-simulations of the
whole human body and a mathematical model. So we want to calculate from an critical current
density on the heart (which we measured and transformed to the human situation) backwards to the
situation when a person contacted from e.g. hand-to-hand onto an electric network. The finite-
element simulation are done by Seiberdorf Laboratories and the Matlab simulations are done by
EATON Vienna.
We have also simulated the mechanisms of ventricular fibrillation on heart tissue. Therefore we
implemented the following formulas into Matlab:

So we can simulate the qualitative behavior of the chaotic electrical activity on the heart during
ventricular fibrillation (in a monodomain approach)
Results
One of the studies key results is the experimental setup which is capable to systematically generate
reproducing and reasonable data. A further benefit is the high degree of flexibility due to the concept
of a adjustable frame and a very flexible software model.
The fibrillation thresholds of the isolated rabbit heart were measured up to 10kHz The number of
experiments/laboratory animals was 100 in general. The fibrillation threshold value for 50 Hz was
[mean,STD] = 215,25+/- 53mApp
In (Fig. 3) the upper (yellow) trace shows the output of the electrical field sensor. The second
(magenta) trace shows the stimulus signal. The next trace is the atrial flow signal. Below one can
see the left ventricular pressure signal (red), which is one of the most important signals because it
also delivers information during the current stimulation and is also strongly related to the ECG signal
(below, green line). The blue trace shows the T-wave trigger signal, to check that the stimulation
was started at the intended moment. In the figure we see a complete stimulation experiment. At the
beginning of the timeline, there is a normal heart action followed by a 3 second electric current
stimulation. During this time ventricular fibrillation is initiated. One can see especially the effect on
the pressure signal, which decreases due to the fibrillation. After a certain time the defibrillation
pulse was set which can be seen very well on the E-field sensor signal. The small spike after the
defibrillation pulse originates from the recoupling of the ECG- and stimulation-source as an EMC
interference. Usually it takes a short time (e.g.2s) before the heartbeat starts again and the full heart
action is restored.
Discussion
The presented experimental setup has the advantage of high flexibility, so that it would be also
possible to use this setup for experiments in other fields, e.g. investigation of the effect of conducted
electrical weapons, among others. One of the main indicators for the usefulness of the experimental
system is the comparison of derived results with well-known results that fit to the stimulation mode.
In Geddes [4] a comparable setting for the dog/rabbit is disclosed. The heart vector and the
stimulation time are the same as in the current experimental study and the condition for the heart

720
respectively the surrounding media are also very similar. From this the results reported in this study
fit very well to the results from Geddes [4] and Ferris [5] Furthermore the qualitative behavior of the
measurements converges to the suggestions from the IEC ([6],[7]) which is in a certain sense an
excerpt of the knowledge about the behavior of the heart with respect to ventricular fibrillation from
10Hz up to 10kHz. One can also see that the most sensitive area is around 50Hz, which is a well-
known result. The next steps will be the the translation to the human, compare the result with the
IEC data [7]and the investigation of direct current and mixed signals.
Conclusion
The results of this study clearly demonstrated the suitability of the developed experimental setup for
the investigation of electric current induced ventricular fibrillation in a working heart model
environment. Furthermore, well known result could be reproduced and refined. In this sense the
following measurements (up to 100kHz,DC,…) will bring new and additional necessary data which
should be concerned in the standardization process and also lead to a higher level of safety in the
field of power distribution.
Reference
[1] Biegelmeier, G, Lee, W.R., New considerations of ventricular fibrillation for a.c. shocks at
50-60Hz, IEE Proc. Vol.127 (1980) Nr.2 S103-110
[2] B.Podesser et al., The erythrocyte-perfused ``working heart'' model: Hemodynamic and
metabolic performance in comparison to crystalloid perfused hearts }, J Pharmacol Toxicol 41 (1999)
9-15
[3] B.Podesser et al., New developments in the isolated working heart: a comparison of neonatal,
immature, and adult rabbits after sixty minutes of ischemia in respect to hemodynamic and
biochemical parameters J Pharmacol Toxicol 1993 Dec;30(4):189-96.
[4] Geddes LA, Cabler , Moore , Rosborough , Tacker . Threshold 60-Hz current required for
ventricular fibrillation in subjects of various body weights, IEEE Trans Biomed Eng. 1973
Nov;20(6):465-8.
[5] Ferris, L. P., King, B. G., Spence, P. W. and Williams, H. B., Effect of Electric Shock on the Heart,
Bell System Technical Journal,(1936) 15: 455–468. doi: 10.1002/j.1538-7305.1936.tb03562.x
[6] Cecil S., Schmid G., Überbacher R.,Koch M., Bartonek M., Modelling of ELF-current flow throw
the human body, Bioem 2016
[7] IEC TS60479-2:2005-07, Technical Specification : Effects of current on human beings and
livestock Part2 : Special aspects},

Figures

721
Figure 1. The experimental setup in the General Hospital Vienna (AKH Wien)

Figure 2. The rabbit heart (outside the experimental frame) in the working heart mode.

722
 Figure 3. Diagram for the whole stimulation experiment

PB-100 [14:30]
Protective effects of radiofrequency on β-amyloid-induced toxicity in human neuroblastoma
cells
Anna Sannino1, Stefania Romeo1, Olga Zeni1, Letizia Venturini2, Giovanni Ricevuti2, Salvatore Caorsi3 &
Maria Rosaria Scarfi1
1Institute for Elecromagnetic Sensing of the Environment, CNR, Naples, Italy, 80131
2Dept. of Internal Medicine and Therapeutics, , University of Pavia, Pavia, Italy, 27100
3Dept. of Electrical, Computer and Biomedical Engineering, University of Pavia, Pavia, Italy, 27100
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Stefania Romeo
This study aims to investigate the capability of pre-exposure to 1950 MHz to protect from the
amyloid beta (Aβ)-induced cytotoxicity in human neuroblastoma SH-SY5Y cells. Pre-exposure to
radiofrequency (RF) reduced of about 30% the Aβ-induced DNA damage, while RF exposure alone
did not affect DNA integrity, as assessed by the alkaline comet assay. Moreover, pre-exposure to RF
increased microRNA 107, affecting the enzyme responsible for the Aβ plaques deposition BACE1
and decreased the expression of microRNAs related to long term potentiation and excito-toxicity
(miRNA335 and miRNA26b). On the woole, RF seems to induce a protective effect against the
damage induced by Aβ in human neuroblastoma cells.
Introduction
Alzheimer’s disease (AD) is a known neurological disorder and the major characteristics include
deposition of senile plaques composed of amyloid beta peptide (Aβ) and hyperphosphorylation of
tau protein [1]. Investigations carried out so far have shown a variety of mechanisms involved in Aβ-
induced toxicity, such as disruption of synaptic transmission, variation in the intracellular calcium,
accumulation of neurofibrillary tangles, activation of mitochondrial apoptotic pathway, oxidative DNA

723
damage [2-4].
Despite decades of research, the current pharmacological treatments only partially mask the
symptoms while the disease progresses. Therefore, non-pharmacologic approaches can be a valid
and safe alternative. In this context, the use of radiofrequency (RF) electromagnetic fields has been
recently suggested and the results obtained are very encouraging [5-7].
The aim of this study is to investigate the capability of 1950 MHz RF exposure, UMTS signal, to
reduce the oxidative DNA damage induced by Aβ treatment in human neuroblastoma (SH-SY5Y)
cells and to modulate the expression of selected microRNAs (miRNAs) connected with the principal
mechanisms that characterized AD.
Materials and methods
Exposure system set up and dosimetry
A short-circuited waveguide-based exposure system was employed and described in detail in our
earlier publication [8]. It was designed and tested on the basis of efficiency (up to 80%) and field
uniformity criteria (coefficient variation 30%). An identical not powered waveguide was used to host
sham-exposed samples. Cell cultures were exposed for 20 h at SAR values of 0.3 and 1.25 W/Kg.
Temperature measurements were carried out for 24 h, at regular intervals of every 5 s, with an
accuracy of ± 0.3 ◦C in a separate experiment under the same experimental RF-exposure conditions
at 1.25 W/kg SAR, by using a fiber-optic thermometer (FisoUMI4, FISO Technologies, Quebec,
Canada) with fiber-optic temperature probe (FISO Technologies, FOT-M/2 m) inserted horizontally in
the culture medium. Five independent measurements showed that temperature never exceeded the
accuracy range of the instrument (± 0.3 ◦C).
Cultures and treatments
SH-SY5Y human neuroblastoma cell line was provided by the American Type Culture Collection
(ATCC). Cells were grown in DMEM supplemented with 10% fetal bovine serum, 1% penicillin–
streptomycin, L-glutamine (2 mM), at 37°C, in an atmosphere of 5% CO2 and 95% humidity. Aβ1-42
was dissolved in distilled water following the manufacturer’s instruction at a concentration of 6
mg/ml. The stock solution was diluted in PBS to desired concentrations immediately before use.
Cell cultures were grown for 96 h and during the last 24 h were treated with 30 µg/ml Aβ.
To investigate the role of pre-exposure to RF, four independent experiments have been carried out
and for each experiment eight cultures were set up as follows: a) control, b) Aβ-treated, c) Sham-
exposed, d) Sham + Aβ, e) RF-exposed at 0.3 W/kg, f) RF-exposed at 1.25 W/kg, g) 0.3 W/kg + Aβ,
h) 1.25 W/kg + Aβ. RF was given for 20 h, from 48 to 68 h after cell seeding. A schematic
representation of the treatment schedule is reported in Figure 1.
Evaluation of DNA damage and cell viability
DNA damage was evaluated by applying the alkaline comet assay, following the protocol previously
described [9], by analyzing 500 cells/treatment and was expressed as % of DNA in the tail. Cell
viability was evaluated by applying the Trypan blue exclusion method and calculated as the fraction
of viable cells of the total cells, expressed as a percentage.
Expression of selected microRNA
Total RNA extraction and quantitative real-time reverse transcription PCR (qRTPCR) were
performed as described in details in [xx]. To evaluate the miRNA expression, qRT-PCR was
performed by the Eco Real-Time PCR Instrument (Illumina, San Diego, CA), and the results were
analysed by the comparative ct method (ΔΔct method) using the software package of the Eco Real-
Time PCR System for the calculus of the 2−ΔΔct value [10].
Results
In preliminary experiments, several Aβ concentrations, ranging from 1 to 100 µg/ml, were tested to
select the dose inducing DNA damage, without affecting cell viability. Concentrations of Aβ lower
than 30 µg/ml did not induce DNA damage. When higher concentrations were tested, an increase of

724
about 40 % of DNA in the tail was recorded for all the doses tested (data not shown). Since cell
viability resulted affected at doses of 50 and 100 µg/ml but not at 30 µg/ml Aβ, the latter was used to
perform experiments with RF.
In cell cultures exposed to RF and then treated with 30 µg/ml Aβ, the Aβ-induced DNA damage was
reduced by pre-exposure to RF, at both 0.3 (- 25%) and 1.25 W/Kg SAR (- 38%) and such reduction
resulted statistically significant at the higher SAR tested (ANOVA: P<0,05), as reported in figura 2
where the results of four independent experiments are reported as mean +/- standard error. No
effects on DNA damage were induced by RF exposure alone.
Concerning the the expression of the selected miRNAs, the preliminary results from one of the four
independent experiments carried out seem to indicate that pre-exposure to RF at both SAR values
tested increased miRNA 107, affecting the enzyme responsible for the Aβ plaques deposition
BACE1 and decreased the expression of miRNA335 and miRNA 26b, related to long term
potentiation and excito-toxicity.
Discussion and conclusion
Several results are reported in the literature on the capability of chemical treatments to reduce the
Aβ-induced cytotoxic and/or genotoxicity in SH-SY5Y cells. Musalma and co-workers [11] showed
that treatment with tocotrienol rich fraction, an antioxidant, member of the vitamin E family, resulted
in significantly less DNA damage, higher cell survival and decreased number of apoptotic cells, as
compared to Aβ-treated cells alone.
A protective effect against the cytotoxicity induced by Aβ was also reported by Zhang et al [12]
following treatments of SH-SY5Y cells with pyrroloquinoline quinone (PQQ), an anionic, water
soluble antioxidant. Zhang et al [13] reported a reduction of Aβ-induced oxidative stress, apoptosis
and cell death in cultures treated with salidroside (Sald), an active compound isolated from a
traditional Chinese medicinal plant.
Our preliminary results, are encouraging although the number of observation must be increased to
obtain firm information of the capability of RF exposure in inducing protective effects against the Aβ-
induced cell toxicity.
References
[1] F.M. La Ferla, S. Oddo,Trends Mol. Med., vol. 11, pp.170–176, 2005.
[2] L. Bojarski, J. Herms, J. Kuznicki, Neurochem. Int., vol. 52, pp. 621–633, 2008.
[3] M.R. Hynd, H.L. Scott, P.R. Dodd, Neurochemm. Int., vol. 45, pp. 583–595, 2004.
[4] P. Mao and P.H. Reddy, Biochimica et Biophysica Acta, vol. 1812, pp. 1359–1370, 2011.
[5] G.W. Arendash, T. Mori, M. Dorsey, R. Gonzalez, N. Tajiri, et al., PLoS ONE, vol. 7(4):e35751
doi:10.1371/journal.pone.0035751, 2012.
[6] G.W. Arendash, J. Sanchez-Ramos, T. Mori, M. Mamcarz, X. Lin, et al., J Alzheimers Dis., vol.
19 pp. 191–21, 2010.
[7] N. Dragicevic, P.C. Bradshaw, M. Mamcarz, X. Lin, L. Wang, et al.,Neuroscience, vol. 185, pp.
135–149, 2011.
[8] O. Zeni, A. Sannino, M. Sarti, S. Romeo, R. Massa, Vijayalaxmi and M.R. Scarfì,
Bioelectromagnetics, vol. 33, pp. 497-507, 2012.
[9] Zeni, O. & Scarfi, M. R. Methods Mol Biol 625, 109-119,
doi:10.1007/978-1-60761-579-8_10(2010).
[10] K. J. Livak and T. D. Schmittgen, Med & Health, vol. 4(1), pp. 8-15, 2009.
[11] J. Zhang, R. Zhang, X. Meng, Neuroscience Letters, vol. 464, pp. 165–169, 2009.
[12] L. Zhang, H. Yu, X. Zhao, X. Lin, C. Tan, G. Cao, Z. Wang, Neurochemistry International, vol.
57, pp. 547–555, 2010.
Figures
725
 Figure 1. Treatment schedule for RF exposures and co-exposures

Figure 2. % DNA in the tail in SH-SY5Y cells exposed/sham exposed for 20 h to 1950 MHz
and treated with 30 µg/ml Aβ. Data are means ± SE of 4 experiments. *P<0.05 (ANOVA)

PB-102 [14:30]
STUDENT PAPER
The role of NF-kB and p-SAPK/JNK inflammatory responses of RFR- exposed colon cancer
cells
Fatih Senturk1, Elcin Ozgur1, Gorkem Kismali2, Tevhide Sel2 & Goknur Guler1
1Biophysics Department, Gazi University, Ankara, Turkey, 06500
2Biochemistry Department, Ankara University, Ankara, Turkey, 06110
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Fatih Senturk
The aim of this study is to investigate the effects of 900, 1800, 2100 MHz mobile phone radiation
exposure during 1 hour and 4 hours on NF-kB and p-SAPK/JNK activation levels on HCT116 colon
cancer cells. Results showed that 1h exposure to 900, 1800 MHz caused to statistically significant
increase, whereas 2100 MHz decreased NF-kB levels, but 4 hours of exposure lead to increase the
level of NF-kB for all exposure groups.On the other hand, 1-hour exposure of 2100 MHz leads to
increase in SAPK/JNK levels. However, no difference was found in the other groups with respect to
sham groups.To conclude that 2100 MHz RFR exposure may modulate NF-kB and SAPK/JNK
expression in HCT116 cells.
Introduction
Health effects of Radiofrequency Radiation (RFR) caused by mobile phones and the wireless
726
communication devices led to concern in scientific and social concerns. The molecular mechanisms
underlying RFR induced inflammation mammalian cancer cells are not fully understood. The aim of
this study is to investigate the effects of 900, 1800, 2100 MHz mobile phone modulated RFR
exposure during 1 hour and 4 hours on NF-kB (Nuclear Factor kappa B) and p-SAPK/JNK(Stress-
activated protein kinases/Jun amino-terminal kinases) activation levels of HCT116 colon cancer
cells.
Method
In vitro, HCT116 colon cancer cells were exposed to RFR during 1 hour and 4 hours intermittent
exposure (15 min on/15 min off). RFR exposure system is digitally modulated 3GPP signals at 900,
1800, 2100 MHz were produced by a vector signal generator and a horn antenna in a temperature-
controlled shielded room. NF-kB and p-SAPK/JNK activation levels were measured by ELISA
method. Statistical analysis was performed by using GraphPad Prism 7 software. P<0,05 was
considered statistically significant.
Results
Data showed that 1h exposure to 900, 1800 MHz caused to statistically significant increase,
whereas 2100 MHz decreased NF-kB levels, but 4 hours of exposure lead to increase the level of
NF-kB for all exposure groups. On the other hand, 1-hour exposure of 2100 MHz leads to increase
in SAPK/JNK levels. However, no difference was found in the other groups with respect to sham
exposure.
Conclusion
Activation of NF-kB playing a well-known function in the regulation of immune responses and
inflammation may result from activation of tumor necrosis factor. SAPK/JNK member of the mitogen
activated protein kinase family is activated by a variety of environmental stresses, inflammatory
cytokines and growth factors. To conclude that 2100 MHz RFR exposure may modulate NF-kB and
SAPK/JNK expression in HCT116 cells.
Acknowledgements
The author(s) would like to acknowledge the networking support by COST Action BM1309 (COST
EMF-MED).

PB-104 [14:30]
5 ns electric pulses evoke Ca2+ responses in adrenal chromaffin cells that are longer in
duration than those evoked by a physiological stimulus
Ruby Sukhraj1, Adam Shoaf2, Indira Chatterjee1, Normand Leblanc2 & Gale Craviso2
1Department of Electrical and Biomedical Engineering, University of Nevada, Reno, Reno, Nevada, USA,
89557
2Department of Pharmacology, University of Nevada, Reno, School of Medicine, Reno, Nevada, USA, 89557
Keywords: In vitro, Pulsed, Work in Progress
Presented by: Normand Leblanc

A 5 ns, 5 MV/m electric pulse (NEP) causes a rapid rise in intracellular Ca2+ in bovine adrenal
chromaffin cells that takes much longer to return to pre-stimulus levels relative to Ca2+ responses
evoked by nicotinic receptor stimulation. Although Ca2+ uptake into mitochondria is the main
mechanism by which Ca2+ is rapidly removed from the cytoplasm to achieve baseline levels, we
found that 5 or more NEPs are required to see any evidence of a change in mitochondrial
membrane potential. Understanding the mechanism underlying the slow clearance of Ca2+ from the
cytoplasm following NEP exposure is important since the ability of a chromaffin cell to rapidly clear
the cytoplasm of Ca2+ by a nicotinic receptor agonist is also affected.

727
Key words:
Adrenal chromaffin cells, nanosecond electric pulses (NEPs), intracellular Ca2+, mitochondrial
membrane potential, and nicotinic receptor agonist.
Introduction
We have been exploring the potential for 5 ns duration electric pulses of high intensity (>1 MV/m) to
serve as a novel stimulus to alter neural cell excitability, using isolated bovine adrenal chromaffin
cells as a simple model of neural-type cells. Exposing chromaffin cells to a 5 ns, 5 MV/m pulse
(NEP) causes a rapid rise in intracellular Ca2+[1] that mimics the rapid rise in intracellular
Ca2+evoked by activation of nicotinic cholinergic receptors, the physiological stimulus of these cells
in vivo that evokes Ca2+-dependent exocytosis. However, of significance is that the Ca2+response
evoked by the NEP is much longer in duration than that evoked by nicotinic receptor activation. This
difference is highlighted in the figure shown below for a cell exposed to either a single NEP versus a
single 5 ms application of the nicotinic receptor agonist 1,1-dimethyl-4-phenylpiperazinium (DMPP).
One possibility that could explain the much longer recovery time for intracellular Ca2+to return to
baseline levels in cells exposed to a NEP is that the penetration of the E-field into the cell disrupts
mitochondrial membrane potential [2-3]. This disruption would slow the rate of Ca2+uptake into
mitochondria, which is the main mechanism by which Ca2+is rapidly removed from the cytoplasm to
achieve baseline levels [4-6]. The purpose of this study was to explore whether NEPs cause a
change in mitochondrial membrane.
Methods
Cell culture and fluorescence imaging
Bovine chromaffin cells were isolated from adrenal glands and maintained in suspension. Cell
aggregates were dissociated with the protease, dispase, and attached to glass bottom dishes
coated with fibronectin. Cells were labeled with Calcium Green-1 (Ex: 506 nm, Em: 531 nm) to
monitor intracellular Ca2+ levels. Tetramethylrhodamine ethyl ester, (TMRE), (Ex: 549 nm, Em: 575
nm) was used to label mitochondria and monitor mitochondrial membrane potential. After cell
labeling, cells were maintained in a balanced salt solution (BSS) composed of: 145 mM NaCl, 5 mM
KCl, 1.2 mM Na2 HPO4, 10 mM glucose, and 15 mM Hepes, pH 7.4. The dishes of attached cells
were then placed on a stage on a Nikon TE 2000 inverted epifluorescence microscope. Images
were captured with a iXonEM + DU-897 EMCCD camera and open source Micro-manager software.
5 ns electric pulse delivery
5 ns, 5 MV/m pulses were applied to an attached cell using a pair of gold coated parallel tungsten
electrodes (diameter of 127 μm), spaced 100 μm apart. The tips of the electrode tips were
positioned 40 μm above glass bottom dish, with the exposed cell located in the center between the
electrode tips. Pulses were generated by a custom fabricated pulse generator (Transient Plasma
Systems) and pulse traces were captured with an oscilloscope.
Physiological stimulus application
DMPP was applied to an individual cell for a duration of 5 ms using a pressure ejection system
(Picospritzer). Micropipettes used for agonist delivery were fabricated from borosilicate glass
capillaries having a tip of 1 μm. Pipette tips were placed at a distance of 1.5 cell diameters away
from the target cell.
Results and Discussion
Ca2+ responses evoked by a NEP significantly outlast those evoked by even multiple
applications of DMPP
In Calcium Green-1 loaded cells exposed either to a NEP or one or more applications of DMPP, we
quantified the difference in the duration of Ca2+ responses by determining the half-width of the

728
responses, that is, the time interval for intracellular Ca2+ to decline to 50% of the maximal value. For
a single pulse the half-width ranged from 8 to 38 s, with an average half-width of 25 s (N = 18). This
duration differed significantly from the much shorter half-width determined for a single application of
DMPP that ranged from 4 to 6 s, with an average of 5 s (N = 11). Moreover, the half-widths for Ca2+
responses evoked by 5 and 10 applications of DMPP were still much shorter in duration than that
evoked by a single NEP. Half-widths for 5 applications of DMPP were 10 s and 7s for agonist
delivered at 1 Hz and 10 Hz, respectively. In the case of 10 applications of DMPP, half-widths were
13 s and 11 s for agonist delivered at 1 Hz and 10 Hz, respectively.
Disrupting mitochondrial membrane potential increased the duration of DMPP-mediated Ca2+
responses
As a positive control for establishing that disrupting mitochondrial membrane potential will alter the
duration of Ca2+ responses evoked by a stimulus, we incubated Calcium Green-1 loaded cells with
the protonophore, carbonyl cyanide m-chlorophenyl hydrazone (CCCP; 500 nM for 5 minutes at
37°C) and analyzed the response of the cells to 10 applications of DMPP at 10 Hz. The results
showed that the Ca2+ response was now much longer in duration, exhibiting in some cells the same
time course as the Ca2+ response evoked by a NEP. Clearly, disrupting Ca2+ clearance by
mitochondria results in longer-lived increases in intracellular Ca2+ evoked by a stimulus.
A single NEP had no detectable effect on mitochondrial membrane potential
We next determined whether a single NEP affects mitochondrial membrane potential by labeling
cells with TMRE (20 nM for 20 min at 37°C) and monitoring changes in fluorescence. A decrease in
TMRE fluorescence is indicative of a change in mitochondrial membrane potential. The application
of single 5 ns, 5 MV/m pulse did not result in any change in TMRE fluorescence (N = 0/15). With
multiple pulses, however, decreases in TMRE fluorescence were observed is some cells. For 5 and
10 pulses delivered at 10 Hz, there was a small, gradual decrease in TMRE fluorescence (N = 3/11
cells and N = 3/6 cells for 5 and 10 pulses, respectively). In cells exposed to 20 pulses at 10 Hz,
there was an immediate drop in TMRE fluorescence (N = 2/2). From these results we tentatively
conclude that the slow recovery of intracellular Ca2+ to baseline levels in cells exposed to a single
NEP is not due to an effect on mitochondrial membrane potential. Other strategies are being
explored to examine further this and other effects of a single NEP on mitochondrial function.
A NEP prolonged the time course of Ca2+ responses evoked by DMPP
Because Ca2+responses evoked by a NEP are long-lived, we determined whether the inability of an
exposed cell to rapidly clear Ca2+would result in subsequent Ca2+responses evoked by DMPP to
also be prolonged. For this analysis, a single NEP was applied to a cell, followed 10 s, 20 s or 30 s
later by the application of DMPP. The results showed that even though intracellular Ca2+was still
elevated when DMPP was applied, the agonist was able to evoke a Ca2+response. Thus, NEP-
exposed cells are still able to respond to a physiological stimulus. However, the return of Ca2+to
baseline levels followed the same slow time course as that evoked by the pulse. Thus, the
mechanism that causes the slow decline of intracellular Ca2+to baseline in NEP-exposed cells also
results in a slow decline of intracellular Ca2+to baseline in cells that are later challenged with a
physiological stimulus.
Conclusion
Exocytosis in chromaffin cells, as in nerve terminals, is triggered by Ca2+ influx via voltage-gated
Ca2+ channels. In order for a chromaffin cell to be ready to respond to the next depolarizing
stimulus, which in vivo can arrive at rates of 0.2 – 0.7 Hz at rest, and 10 – 20 Hz when the
sympathetic nervous system is activated, efficient Ca2+-clearance mechanisms are at play. In this
regard, it has been well documented that if the rate of Ca2+ clearance is reduced, leading to longer-

729
lived increases in intracellular Ca2+, catecholamine release is enhanced. This means that NEP
exposure may lead to enhanced exocytosis, which we have found to be the case (unpublished
results). For this reason, it will be crucial for future studies to elucidate the basis for the slow
recovery of Ca2+ to baseline levels.
Acknowledgements
This research was funded by grants AFOSR FA9550-12-1-0018, FA9550-12-1-0023, and MURI
FA9550-15-1-0517.
References
[1] G. L. Craviso, S. Choe, I. Chatterjee and P.T. Vernier. Modulation of intracellular Ca2+in
chromaffin cells by nanoelectropulses. Biolelectrochemistry. 87:244-252, 2011.
[2] P. T. Vernier. Mitochondrial membrane permeabilization with nanosecond electric pulses. Annual
International Conference of the IEEE EMBS. 33:743-745, 2011.
[3] T. B. Napotnik, Y. H. Wu, M. A. Gunderson, D. Miklavcic, and P. Thomas Vernier. Nanosecond
electric pulses cause mitochondrial membrane permeabilization in Jurkat cells. Bioelectromagnetics.
1-8, 2011.
[4] J. Garcia-Sancho and A. Verkhratsky. Cytoplasmic organelles determine complexity and
specificity of calcium signaling in adrenal chromaffin cells. Scandinavian Physiological Society.
1748-1716, 2007.
[5] J. Garcia-Sancho, A. M. G. de Diego, and A. G. Garcia. Mitochondria and chromaffin cell
function. Eur J. Phyiol. 464:33-41, 2012.
[6] M. Montero, M. T. Alonso, A. Albillos, J. Garcia-Sancho, and J. Alvarez. Mitochondrial Ca2+-
induced Ca2+release mediated by the Ca2+uniporter. Molecular biology of the cell. 12:63-71, 2001.
Figures

Figure 1. Ca2+ responses monitored in chromaffin cells labeled with Calcium Green-1. (Left)
Single 5 ms application of 100 μM DMPP delivered via a pressure ejection pipette. (Right)
Single application of a 5 ns pulse. In each figure the arrow indicates the time that the stimulus
was applied to the cells.

730
PB-106 [14:30]
Nanosecond bipolar pulse cancellation below 10 ns — Looking for mechanisms in an
expanding bioelectric puzzle
Esin B Sözer1 & P. Thomas Vernier1
1Frank Reidy Research Center for Bioelectrics, Old Dominion University, Norfolk, VA, USA, 23508
Keywords: In vitro, Pulsed, Work in Progress
Presented by: Thomas Vernier
We present studies of unipolar and bipolar electropermeabilization with 2 ns pulses. Some of our
observations are consistent with previous reports of bipolar cancellation with longer pulses, but we
report also new, unexpected effects of bipolar exposures and discuss mechanistic explanations.
INTRODUCTION
Despite studying interactions of electric fields with cells, and cell membranes over decades, our
understanding of events and responses that are initiated by the electric field exposure is still far from
complete. Nanosecond bipolar pulse cancellation, a recently discovered bioelectric puzzle, is
cancellation of biological and physical effects of a unipolar electric pulse by a second pulse of
opposite polarity [1]. This effect does not apply to longer duration, conventional electroporation
pulses (typically 100 µs) [2], and in some cases cancellation can occur up to 50 µs after the first
unipolar pulse delivery [3].
We report cancellation effect of a bipolar 2 ns electrical pulse, and we discuss the experimental
results in detail, comparing them to results reported for longer pulse exposures, and we reveal new
findings observed with these ultra-short bipolar pulses.
MATERIALS AND METHODS
Cells. U-937 (human histiocytic lymphoma monocyte; ATCC CRL-1593.2) cells [4] were cultured in
RPMI-1640 medium (Corning® glutagro™ 10-104-CV) with 10% fetal bovine serum (Corning,
35-010-CV) and 1% penicillin/streptomycin (10000 U/mL penicillin and 10 mg/mL streptomycin) at
37 °C in a humidified, 5% CO2 atmosphere.
Pulsed Electric Field Exposure. 2 ns, 17 MV/m (HORUS) or 42 MV/m pulses (FID) pulses were
delivered to cells in suspension in cover glass chambers [Nunc™ Lab-Tek™ II] through parallel
tungsten wire electrodes with 50-60 µm interelectrode gap. Bipolar pulses were either symmetric
with equal amplitude of the negative and positive phases or asymmetric with negative phase
amplitude ~30% of the positive phase. Cells were observed at laboratory room temperature on the
stage of a Leica TCS SP8 laser scanning confocal microscope.
Molecular Transport Experiments. Cells were washed and suspended at approximately
5×105cells/mL in fresh medium containing 2 µM YO-PRO-1 for influx experiments, and cells were
loaded in fresh medium containing 0.25 µM calcein-AM for 15 minutes at 37 °C in a humidified, 5%
CO2 atmosphere, then resuspended in fresh RPMI1640 medium at approximately 5×105cells/mL for
efflux experiments.
Imaging. Laser scanning confocal fluorescence microscope images were captured (LeicaTCS SP8)
every 1 or 2 seconds for two to ten minutes from cell suspensions at room temperature in ambient
atmosphere on the microscope stage. For calcein efflux measurements, slice thicknesses as large
as the height of the cells were used, making the imaging roughly equivalent to regular fluorescence
microscopy. This adjustment was made to remove the need to account for cell volume change in
these measurements.
Image Processing. Cells visible in the microscope field between the electrodes were manually
selected for fluorescence photometric image analysis before each pulse exposure. Fluorescence
intensities of each region of interest were extracted using custom MATLAB routines. The following
built-in MATLAB functions were used in custom image processing routines.: ‘imroi’, for manually
choosing regions of interest; ‘regionprops’, for evaluating geometric properties of regions of interest.

731
RESULTS
Bipolar cancellation occurs in YO-PRO-1 uptake with 2 ns pulses, similar to previous reports
with 200 ns pulses.
Figure 1 YO-PRO-1 fluorescence increase after unipolar (magenta), and bipolar (blue), 40, 2 ns, 17
MV/m pulses delivered at 1 kHz repetition rate.
After a 2 ns bipolar pulse exposure, we observe YO-PRO-1 uptake comparable to a sham exposure,
whereas a unipolar pulse exposure of same amplitude results in a 6-7 fold higher transport of the
dye [Figure 1]. This result is consistent with a previous study reporting electropermeabilization by
200 ns pulses [3].
Bipolar cancellation is also observed in calcein efflux.
Unipolar and bipolar pulse exposures result in significantly different permeabilization to calcein.
Calcein-loaded U-937 cells were exposed to unipolar, bipolar-asymmetric, and bipolar-symmetric
pulses of 42 MV/m electric field amplitude at the positive pulse phase. Unipolar pulses cause higher
efflux than bipolar symmetric pulses, which in turn produce higher efflux than bipolar-asymmetric
pulses [Figure 2]. This result shows that nanosecond bipolar pulse cancellation can occur not only
for the influx of cationic molecules as reported here and previously [3,5] but also for the efflux of an
anionic molecule such as calcein.
Figure 2 Calcein efflux before and after unipolar (magenta), bipolar symmetric (green) and bipolar
asymmetric (blue), 40, 2 ns 42 MV/m pulses delivered at 10 Hz repetition rate.
Bipolar pulse exposures cause shrinking of U-937 cells unlike unipolar pulse exposures.
An unexpected observation that occurred during calcein efflux measurements was a slight shrinking
of the cells exposed to bipolar pulses, in contrast to the swelling of cells exposed to unipolar pulses
[6-8]. This new result is under investigation.
Figure 3 Normalized cell area change before and after 40, 2 ns 42 MV/m pulses of unipolar
(magenta), bipolar symmetric (green) and bipolar asymmetric (blue) shape.
ACKNOWLEDGMENT
This work was supported by Air Force Office of Scientific Research grant FA9550-14-1-0123 and
MURI grant FA9550-15-1-0517 on "Nanoelectropulse-Induced Electromechanical Signaling and
Control of Biological Systems", administered through Old Dominion University.
REFERENCES
[1] A. G. Pakhomov et al., “Cancellation of cellular responses to nanoelectroporation by reversing
the stimulus polarity,” Cell. Mol. Life Sci., vol. 71, no. 22, pp. 4431–4441, 2014.
[2] T. Kotnik, L. M. Mir, K. Flisar, M. Puc, and D. Miklavčič, “Cell membrane electropermeabilization
by symmetrical bipolar rectangular pulses: Part I. Increased efficiency of permeabilization,”
Bioelectrochemistry, vol. 54, no. 1, pp. 83–90, 2001.
[3] E. C. Gianulis, J. Lee, C. Jiang, S. Xiao, B. L. Ibey, and A. G. Pakhomov, “Electroporation of
mammalian cells by nanosecond electric field oscillations and its inhibition by the electric field
reversal,” Sci. Rep., vol. 5, p. 13818, Sep. 2015.
[4] C. Sundström and K. Nilsson, “Establishment and characterization of a human histiocytic
lymphoma cell line (U‐937),” Int. J. cancer, vol. 17, no. 5, pp. 565–577, 1976.
[5] B. L. Ibey et al., “Bipolar nanosecond electric pulses are less efficient at electropermeabilization
and killing cells than monopolar pulses,” Biochem. Biophys. Res. Commun., vol. 443, no. 2, pp.
568–573, Jan. 2014.
[6] S. Romeo, Y.-H. Wu, Z. A. Levine, M. A. Gundersen, and P. T. Vernier, “Water influx and cell
swelling after nanosecond electropermeabilization.,” Biochim. Biophys. Acta, vol. 1828, no. 8, pp.
1715–1722, Aug. 2013.
[7] O. M. Nesin, O. N. Pakhomova, S. Xiao, and A. G. Pakhomov, “Manipulation of cell volume and
membrane pore comparison following single cell permeabilization with 60-and 600-ns electric

732
pulses,” Biochim. Biophys. Acta (BBA)-Biomembranes, vol. 1808, no. 3, pp. 792–801, 2011.
[8] E. B. Sözer, Y.-H. Wu, S. Romeo, and P. T. Vernier, “Nanometer-Scale Permeabilization and
Osmotic Swelling Induced by 5-ns Pulsed Electric Fields,” J. Membr. Biol., vol. 250, no. 1, pp.
21–30, 2017.
Figures

Figure 1. YO-PRO-1 fluorescence increase after unipolar (magenta), and bipolar asymmetric
(blue), 40, 2 ns, 17 MV/m pulses delivered at 1 kHz repetition rate

733
Figure 2. Calcein efflux before and after unipolar (magenta), bipolar symmetric (green) and
bipolar asymmetric (blue), 40, 2 ns 42 MV/m pulses delivered at 10 Hz repetition rate.

Figure 3. Normalized cell area change before and after 40, 2 ns 42 MV/m pulses of unipolar
(magenta), bipolar symmetric (green) and bipolar asymmetric (blue) shape, delivered at 10 Hz
repetition rate.

734
PB-108 [14:30]
An ELF magnetic field exposure system for real-time cell monitoring under controlled
temperature conditions
György Thuroczy1, Stefania Romeo2 & Myrtill Simko3
1Department of Non-Ionising Radiation , National Public Health Institute, Budapest, Hungary
2Institute for Electromagnetic Sensing of the Environment (IREA), CNR, Naples, Italy
3SPi-SciProof International AB, Östersund, Sweden
Keywords: In vitro, ELF/LF, Work in Progress
Presented by: György Thuroczy
The aim of present work is to allow real-time microscopic monitoring of cellular and molecular
changes during ELF magnetic field exposure at different temperatures. For this purpose an in vitro
exposure system was designed to expose living cells to ELF magnetic fields under controlled
temperature conditions.
Introduction
The present work is in the framework of a wider project aimed to investigate whether there is a
correlation between reactive oxygen species formation and intracellular localized temperature
increase (“hot spots”) in immune relevant cells under ELF magnetic field exposure. To this aim, an in
vitro exposure system was designed and realized to expose living cells to ELF magnetic fields under
controlled temperature conditions and to allow a real-time microscopic monitoring of cellular and
molecular changes during exposure.
Methods
In order to obtain temperature-controlled exposure conditions during microscopic investigation of
cell cultures in the presence of ELF-MF exposure, an Indium Tin Oxide (ITO)-based glass was
designed as a heating system using DC current. The ITO coated glass is electrically conductive with
a high optical transparency, thus it can be very useful as a temperature regulator during microscopic
investigations. Here, ITO coated glass CEC017S with resistance R=< 17 Ohm/cm2 and 1.1 mm
substrate thickness was used as a conducting glass to heat the cell culture slides or dishes
(Prazisions Glas&Optik GmbH, Germany). Specifically, an ITO-coated glass was connected via
carbon electrodes to the DC current source, which in turn was controlled by a temperature probe
(Luxtron optical probe) placed on the surface of the ITO glass or in the sample fluid (cell culture
medium) to control the temperature in real time. The DC current source was controlled by a
thermometer, equipped with a non-perturbing thermistor probe; the DC current source was limited to
200 mA. According to the temperature measured in the sample, the temperature controller provides
a DC current to the ITO-based system, which heats the sample until it reaches a pre-set
temperature value.
We used a Petri-dish with 2ml liquid (distilled water) that was placed on the conductive side of the
ITO glass heating system. The temperature was measured with two optical temperature probes: the
first probe was placed into the liquid within Petri-dish the second probe was placed at the bottom
(non-conductive) side of the glass plate. The measurements were performed both with covered and
non-covered (open) Petri-dishes. Before the study, we measured the temperature elevation of the
conductive and non-conductive sides of the glass. The temperature was almost similar on both
sides of the glass under heating conditions (by DC current). A double-coil system was designed to
produce the 50 Hz magnetic field. The inner diameter of coils was 12 cm with 85 turns of each coil.
The diameter of copper wire was 0.3 mm. The coils were connected in parallel (as a pair) with the
resultant resistance of 2 ohms. The magnetic field was measured by a Hall-probe connected to
Gauss meter (Lakeshore Model 410).
Results
Figure 1 shows the scheme of the exposure set-up that was designed to enable the exposure to
ELF magnetic fields of live cells during microscopical investigations in real-time and under controlled

735
temperature condition. In Figure 2, the temperature profile of 2 ml of water in the Petri dish, first
heated through the ITO glass plate by 69-180 mA DC current (5-8V voltage) and then left cooling at
room temperature. The system was not perturbed in the presence of 1mT, 50 Hz magnetic field
(Fig.3).
Discussion
The ITO-based heating system here described allows keeping a stable temperature during exposure
to the ELF magnetic field. The system has been mounted under a confocal microscope for real-time
monitoring of cells during exposure, and will be improved according to the experimental
requirements. In particular, the standard Petri dish will be replaced by a “Fluorodish” Petri-dish
which is equipped with cover glass bottom to allow sample inspection trough the microscope. The
temperature profile of the liquid sample was characterized also in the Fluorodish, where we obtained
similar results with respect to the standard Petri dish. Regarding the coils that were employed for the
exposure of the ELF magnetic fields a temperature increase was detected, which depended on the
induced magnetic flux density. The measurements showed that at 1 mT flux density the temperature
increase on the inner wall of the coils’ plastic holder was around 2-3 oC.
Figures

Figure 1. Block diagram of the system

736
Figure 2. Temperature at different DC voltage in the liquid of covered Petri-dish placed on
CEC017S glass plate

Figure 3. Temperature recordings under temperature control by the heated glass with switch
on/off 1 mT 50 Hz magnetic fields. The temperature in the water sample (blue) is fixed at 37.5
oC while the temperature of the glass was changing (red). The temperature of the coil (green)
was also increased during the exposure to magnetic field, but the temperature in the sample
was unchanged.

737
PB-110 [14:30]
STUDENT PAPER
Physical chemical analysis of tomato seeds (Solanum lycopersicum L.) under magnetic
treatment
Javier Torres Osorio2, Laura Maria Echeverry-Cardona1, Elisabeth Restrepo-Parra1, Nelson Ceballos
Aguirre3 & Diana Carolina Galvez-Coy4
1Departamento de Física y Química, Universidad Nacional de Colombia, Manizales, Colombia, 170002
2Departamento de Física, Universidad de Caldas, Manizales, Colombia, 170002
3Department of Agricultural Production, University of Caldas, Manizales, Colombia, 170002
4Tecnoacademia, Servicio Nacional de Aprendizaje SENA, Manizales, Colombia, 170002
Keywords: In vitro, Static, Completed (unpublished)
Presented by: Laura Maria Echeverry-Cardona
Physical chemical analysis of magnetic treatment tomato seeds that exhibited reduction in the mean
germination time (MGT) was carried out for knowing the mechanisms that generate biophysical and
biochemical modifications of the seeds because the magnetic field. The experiment was performed
using a set of dipolar toroidal magnets, 8 doses of stimulation and a control for germination tests.
The seeds with lower MGT were chosen. The chemical and the surface morphology characterization
was performed. Seeds with lower MGT showed an increase of functional groups of lipid molecules,
increase in the free water for reactions and expressions of potassium ions of the seed coat, that can
be related to the increment in the membrane transport.
Introduction
Research on seed magnetic treatment and its physiological response has been documented for
various stages of the plant cycle, from effects on the germination phase to indicators of yield
components [Galland and Pazur, 2005; Maffei, 2014; Pietruszewski and Martínez, 2015], although
most reports focus only on germination. In order to reach a further description and explanation of the
physiological effects of magnetic seed treatment, different hypotheses has been formulated. Some
of these are linked from the modified mechanisms at the biophysical and biochemical level, since
the interpretation of these affectations can contribute to the understanding of the mechanisms that
are unchained in the plant under the effect of the magnetic treatment. The literature reports
influence on enzymatic activation, imbibition [Vashisth & Nagarajan, 2010; Shine et al., 2011],
variations in ionic currents [Socorro & García, 2012], and changes in the cellular membrane
characteristics and RNA quantification [Goodman et al., 1995].
The aim of this work is contributing to the explanation of the biophysical or biochemical action
mechanisms that have been modified by the interaction of seeds with the magnetic field, and thus,
to contribute to the understanding of the mechanisms involved in the seeds magnetic treatment.
Methods
As a magnetic field source, a set of ten toroidal magnetized magnets across its thickness was used,
with nominal magnetic flux density (B) of 100 mT, external radius of 5.00 cm, internal radius of 2.50
cm and 2.00 cm of thickness, positioned with the north polarity on the upper face of the toroid. The
characterization of the magnets was acording with [Torres et al., 2018], allowing to define the spatial
distribution of B, and from this way, to make a relationship between the volume of work used and the
values of B in the place where the seeds are located, determining that the seeds were subjected to
a gradient of magnetic induction up to tens of mT / mm.
An experiment of germination was carried out with magnetically treated seeds, using the
aforementioned nominal values of B, and 8 values of exposition time (texp), from 1 h up to 8 h, with
steps of 1 h, and taking 4 repetitions of 30 seeds per treatment. Of each repetition per treatment, 5
seeds were taken, each 6 h up to 24 h, in order to have 20 seeds per treatment at each time,

738
keeping them at – 15 °C. Of these treatments, two that presented values of MGT with significative
differences respect to the control were chosen for performed the chemical physical analysis, having
the treatments T1 (100 mT – 1 h) and T2 (100 mT – 6 h). MGT of the control was 58.9 h ± 1.3 h,
being MGT for T1 and T2 of 45.4 h ± 2.0 h and 50.3 h ± 1.9 h, respectively.
Wild tomato seeds were used, introduction of IAC 391. After receiving the magnetic treatment, the
seeds were placed in Petri dishes of (100 x 15) mm, using as germination matrix absorbent paper
moistened with distilled water, with four repetitions of 25 seeds per dish. The seeds were kept in
Incucell incubators, of 55 L, without light. The incubation temperature was 30 °C ± 0.14 °C and the
humidity inside was 59.0 ± 3.39%. After 16 h of sowing, the germination was checked every 4 h until
78 h. It is assumed that the tomato seeds are germinated when the radicle reaches a length equal
than 1 mm [Florez et al., 2007].
After being magnetic treatment, 3 samples were taken every 6 h until 18 h. The determination of the
elemental composition of the samples was carried out using the EDS Bruker, Quantax 200, coupled
to the scanning electron microscopy, SEM (TESCAN, model VEGA3). With an increase of 100X in
three different zones of the sample and an acceleration of 20 kV. Functional groups were identified
using FTIR (BRUKER, ATR Alpha Platinum Diamond) with 4 cm-1 resolution. The thermal
degradation was carried out by TGA (TGA Q500 of TA instruments). The heating ramp is 10 °C
min-1, from room temperature to 800 °C, in a nitrogen atmosphere with a flow of 60 mL min-1.
The choice of the statistical test for the analysis of the MGT, was made with the evaluation of two
completely randomized design assumptions: normal distribution of errors (Shapiro-Wilks) and
homoscedasticity (Bartlett). Given the parametric nature of the data, a one-way ANOVA test was
used with the R program. Pairwise comparisons were performed using the Least Significant
Difference test.
Results
FTIR are compared with the indexed values, according to tne literature [Barraza-Garza et al., 2013;
Bonda et al., 2011; Callejas 2011]. These spectra show bands corresponding to O-H, C=O, N-H and
C-N bonds, belonging to aminoacid molecules. This corroborates that the composition of the seeds
is maily a proteic type strcuture (Fig. 1 and Table 1). It is notorious that C-N functional group,
corresponding to a peptidic bond is predominant in these strcutures. This bond comes from the join
of two aminoacids, generating, as a result, a water molecule release. In the EDS spectrum, it was
observed a great quantity of of oxygen and carbon (Fig 3a), agreeing with the results obtained by
means of FTIR, where the bonds identification of each peak is mostly present in these two
compounds.

739
 F 1. FTIR

T 1. V FTIR

F 3. EDS T1 . )
C , )0 , )6
In the SEM images of the seeds, a normal morphology for the episperm was found (Fig.4), similar to
it reported for this kind of samples [Moreno, 2012]. These structures have, as function, the water
absorption and the inner part protection.

740
 F 4. SEM
The behavior of the seeds depending on the increase of the temperature obtained using the thermal
profile by TGA (FIG 5), has allowed to determine the macromolecule structures. In this profile
several concentrations are determined: total proteins 39.15%, lipids 13.96%, inorganic wastes that
do not reach the decomposition, 16.78 % and 3.13 % corresponding to weak joins of water [Cáceres
et al., 2017]

F 5. TGA
On the other hand, regarding to results obtained for magnetically treated seeds, in Fig. 2, FTIR
spectra of these seeds with treatments T1 and T2, and the control are presented. In these spectra, it
was identified, for the magnetically treated samples, the presence of bands corresponding to C-H,
CH2symmetric and C=O stretch bonds, being these, the main absorption bands of the lipids [Bonda
et al., 2011]; moreover, it was observed an increase in the intensity of bands corresponding to C-H
asysymmetric, C-O stretch-flexion and, C-O respectively. These modifications can be attributed to
the magnetic field promotes the generation of changes in the mechanisms related with the
membrane braking off. In fig. 3, the EDS spectra of the samples T1, T2, with 0 and 6 h of imbibition
and control are included. It was observed the presence of potassium in the magnetically treated
samples, T1 and T2, without imbibition. These results agree with those observed in the FTIR
analysis, where it was observed that the samples exhibited changes in their bonds.
Discussion
The data analysis entails two important points: one is evidenced in the non-magnetically treated
seeds. Since there exists high presence of peptidic bonds (3.250 - 3.260) cm-1, it is guaranteed that
seed has availability of molecules of water for reactions at the moment of the dormancy interruption.
These characteristics can contribute to the tomato seeds have a fast germination, since the time of
the imbibition phase was lower that in the case of seeds without magnetic treatment.

741
On the other hand, it is well known that the seed exhibits two type of dormancy, one external and
another internal, and since the magnetically treated seeds evidenced changes in the functional
groups associated to lipids and the presence of potassium ions, these changes are an indicative of
disturbances in the transport systems of the seed inner structures toward the seed coat. These
alterations are related with the increment of the ionic mobility trough the cell membrane, according
to the exposed by Socorro et al., [2012].
Conclusion
The affectations that the magnetic treatment generates on the germinative processes and
specifically, on the average time of germination, are associated to the rupture of the external
dormancy, which is related to the increase of the presence of functional groups associated with
lipids, ionic expression and increase in available water for the occurrence of these reactions.
References
Barraza-Garza, G., de la Rosa, L. A., Martínez-Martínez, A., Castillo-Michel, H., Cotte, M., &
Alvarez-Parrilla, E. (2013). La microespectroscopía de infrarrojo con transformada de Fourier
(FTIRM) en el estudio de sistemas biológicos. Revista latinoamericana de química, 41(3), 125-148.
Baseri, M. K., & Baker, S. (2011). Identification of cellular components of medicinal plants using
FTIR.Rom. J. Biophys, 21, 277-284.
Bonda, M., Perrin, V., Vileno, B., Runne, H., Kretlow, A., Forró, L., ... & Jeney, S. (2011).
Synchrotron infrared microspectroscopy detecting the evolution of Huntington’s disease
neuropathology and suggesting unique correlates of dysfunction in white versus gray brain matter.
Analytical chemistry, 83(20), 7712-7720.
Cáceres, V. I., Sánchez-Molina, J., & Chaparro-García, A. L. (2017). Evaluación de arcillas
caoliniticas-illiticas provenientes de la formación guayabo del Área Metropolitana de Cúcuta, Norte
de Santander, Colombia. Revista Ion, 30(1), 117-127.
Callejas FR. Tablas de Espectroscopía Infrarroja. online
http://depa.fquim.unam.mx/amyd/archivero/TablasIR_24985.pdf. 07/03/2018
Florez, M., Carbonell, M., & Martínez, E. (2007). Exposure of maize seeds to stationary magnetic
fields: Effects on germination and early growth. Environmental and Experimental Botany, 59(1),
68–75.
Galland P, Pazur A. 2005. Magnetoreception in plants. Journal of Plant Research 118: 371-389.
Goodman, E., Greenebaum , B., & Marron, M. (1995). Effects of electromagnetic fields on
molecules and cells. Int Rev Cytol, 158, 279–338.
Maffei M. 2014. Magnetic field effects on plant growth, development and evolution. Frontiers in plant
Science 5, article 445.
Moreno Guarín, D. C. (2012). Evaluación y estandarización de las condiciones del proceso de
deshidratación de tomate (Doctoral dissertation, Universidad Nacional de Colombia).
Pietruszewski S, Martínez E. 2015. Magnetic field as a method of improving the quality of sowing
material: a review. International Agrophysics 29: 377-389
Shine, M., Guruprasad, K., & Anand, A. (2011). Enhancement of germination, growth and
photosynthesis in soybean by pre-treatment of seeds with magnetic field. Bioelectromagnetics,
32(6), 474–84.
Socorro, A., & García, F. (2012). Simulation of magnetic field effect on a seed embryo cell.
International Agrophysics, 26, 167-173.
Torres, J., Hincapie, E., & Gilart, F. (2018). Characterization of magnetic flux density in passive
sources used in magnetic stimulation. Journal of Magnetism and Magnetic Materials, 449, 366–371.
https://doi.org/10.1016/j.jmmm.2017.10.037
Vashisth, A., & Nagarajan, S. (2010). Characterization of water distribution and activities of enzymes

742
during germination in magnetically-exposed maize (Zea maysL) seeds. Indian journal of
biochemistry & biophysics, 47, 311–318.

PB-112 [14:30]
Effects of microwave radiation on distribution of dendritic spines and SNK-SPAR pathway in
primary hippocampal neurons
Lifeng Wang1, Weijia Zhi1, Haijuan Li1 & Xiangjun Hu1
1Beijing Institute of Radiation Medicine, Beijing, China, 100850
Keywords: In vitro, RF/Microwaves, Completed (unpublished)
Presented by: Lifeng Wang
In the study, we found that 30mW/cm2 microwave radiation could result in the density of dendritic
spines and percentage of mature spines decreased in cultured primary hippocampal neurons; the
increased SNK which led to the decrease of SPAR and PSD-95; low expression of Calcineurin
which induced the decrease of CREB, the increase and nuclear translocation of p-CREB, then lifted
the transcription level of SNK; the phosphorylated modification site in GKBD, which is the specific
binding structure domain with PSD-95 of SPAR, transformed from ser-1567 and ser-1472 to
ser-1603, and the interaction of SPAR and MAP2. It might lead to obstacle in spine maturation and
debilitate synaptic connections.
Epidemiological and experimental studies showed that microwave radiation could result in sleep
disorder, abnormal EEG and deterioration of the memory. Synaptic plasticity is the neural basis of
learning and memory, and the vast majority of excitatory synapses are composed of dendritic spines
as their postsynaptic structure. SNK-SPAR pathway plays a key role in synaptic reconstruction
depending on activation of neurons. It is reported that the microwave radiation could induce
abnormality in dendritic spines plasticity, however, the mechanism are rarely reported. In the study,
we found that 30mW/cm2 microwave radiation could result in the density of dendritic spines and
percentage of mature spines decreased in cultured primary hippocampal neurons; the increased
SNK which led to the decrease of SPAR and PSD-95; low expression of Calcineurin which induced
the decrease of CREB, the increase and nuclear translocation of p-CREB, then lifted the
transcription level of SNK; the phosphorylated modification site in GKBD, which is the specific
binding structure domain with PSD-95 of SPAR, transformed from ser-1567 and ser-1472 to
ser-1603, and the interaction of SPAR and MAP2. It might lead to obstacle in spine maturation and
debilitate synaptic connections.
Key words: microwave; dendritic spines; SNK; SPAR
(This work supported by National Natural Science Foundation of China (61501492))

PB-114 [14:30]
STUDENT PAPER
Effects of RF-EMF on APP-processing and cell death in mouse hippocampus cell line
Kyeonghee Yoon1, Sojeong Choi1, Nam Kim2, Hyung Do Choi3, Hae June Lee4 & Yun-Sil Lee1
1College of Pharmacy & Graduate School of Pharmaceutical Sciences , Ewha Womans University , Seoul,
Korea
2School of Electrical and Computer Engineering, Chungbuk National University, Chungju, Korea
3Departmet of EMF Research Team, Radio and Broadcasting Technology Laboratory, ETRI, Seoul, Korea,
139-706
4Division of Basic Radiation Bioscience, Korea Institute of Radiological and Medical Sciences, Seoul, Korea,
01693
Keywords: In vitro, RF/Microwaves, Work in Progress
Presented by: Kyeonghee Yoon

743
In vitro experiments were performed after exposure of LTE RF-EMF (1745 GHz) because currently,
the spread of LTE mobile communication network has increased. The LTE type RF-EMF was
exposed to the cells for 24 hours at an intensity of 8 W/kg, which is the maximum intensity of our
exposure system and Amyloid precursor protein (APP) processing and cell death effects were
examined. RF-EMF exposure showed no significant changes of ADAM10 and BACE1 as well as
expression of CTFβ and CTFα. RF-EMF, no significant cell death was induced. Even though more
detailed experiments is needed, unlike to IR, exposure of 8 W/kg LTE type RF-EMF for 24 hours did
not affect APP processing and cell death in HT22 hippocampus cells.
Introduction
Alzheimer disease (AD) is known to be one of the most common neurodegenerative diseases, and
its prevalence is strongly correlated with aging. AD is characterized by a gradual decline in memory
and cognitive ability. Alzheimer 's has a steady increase in the incidence of elderly people aged 65
years and older. Amyloid precursor protein (APP) is cleaved to amyloid beta by β-secretase.
Alzheimer's has a pathological feature that an abnormal accumulation of amyloid beta occurs in the
brain. Recent studies have shown that long term RF-EMF (1.95 GHz, SAR 5W/kg, 2 hr/day for 8
months) exposure to Alzheimer's animal model 5XFAD mice alleviates symptoms of Alzheimer's
including passive avoidance performance, alteration behavior (spatial working memory) and
amyloid-beta (Ab) deposits in histopathological analysis. However, RF-EMF exposure for 3 months
did not show any significant effects. Therefore, additional experiments in vitro were required to
elucidate these different effects. In this study, further in vitro experiments were performed after
exposure of LTE (Long term evolution) RF-EMF (1745 GHz) because currently, the spread of LTE
mobile communication network has increased. We investigated APP processing and cell death by
RF-EMF in HT22 hippocampus cells. We also investigated the role of heat shock protein 27
(HSP27) on the effects of RF-EMF.
Results
The LTE type RF-EMF was exposed to the cells for 24 hours at an intensity of 8 W/kg, which is the
maximum intensity of our exposure system and APP processing and cell death effects were
examined. Exposure of ionizing radiation (IR) was used as a positive control and cells were
irradiated at 5 or 10 Gy. The expression of BACE1 (β-secretase), but not ADAM10 (a-secretase)
was increased after IR in both control and APP overexpressing cells. However, amyloidogenic APP
processing product, CTFβ (C-terminal fragment β) and nonamyloidogenic APP processing product,
CTFα were not changed by IR. Degradation of APP by IR was obviously observed accompanied
with decreased expression of CTFβ and CTFα. In the case of RF-EMF exposure, no significant
changes of ADAM10 and BACE1 as well as expression of CTFβ and CTFα were observed.
Overexpression of HSP25 increased the protein stability of APP by binding directly to APP, which
affects increased half life of APP. IR-induced cell death was protected by HSP25 overexpression,
However, APP overexpression increased IR-induced cell death. In the case of RF-EMF, no
significant cell death was induced. Even though more detailed experiments is needed, unlike to IR,
exposure of 8 W/kg LTE type RF-EMF for 24 hrs did not affect APP processing and cell death in
HT22 hippocampus cells.
References
1) Evans DA, Funkenstein HH, Albert MS, Scherr PA, Cook NR, Chown MJ, Hebert LE, Hennekens
CH, Taylor JO (1989) Prevalence of Alzheimer's disease in a community population of older
persons. Higher than previously reported. JAMA 262:2551–2556
2) McKhann G, Drachman D, Folstein M, Katzman R, Price D, Stadlan EM (1984) Clinical diagnosis
of Alzheimer's disease: report of the NINCDS-ADRDAWork Group under the auspices of
Department of Health and Human Services Task Force on Alzheimer's Disease. Neurology
34:939–944
3) Shoji M, Golde TE, Ghiso J, Cheung TT, Estus S, Shaffer LM, Cai XD, McKay DM, Tintner R,

744
Frangione B, Younkin SG. 1992. Production of the Alzheimer amyloid beta protein by normal
proteolytic processing. Science 258: 126–129.
4) Jeong YJ, Kang GY, Kwon JH, Choi HD, Pack JK, Kim N, Lee YS, Lee HJ. 2015. 1950MHz
electromagnetic fields ameliorate Ab pathology in Alzheimer’s disease mice. Curr Alzheimer Res
12:481–492.
5) Son Y, Jeong YJ, Kwon JH, Choi HD, Pack JK, Kim N, Lee YS, and Lee HJ. 2016.
1950MHzRadiofrequencyElectromagnetic Fields Do Not Aggravate Memory Deficits in 5xFADMice.
Bioelectromagnetics 37:391-399
Figures

Figure 1. A) External appearance of the whole RF-EMF cell exposer and an incubator
containing RF-EMR exposer. Amplifier to emerge RF-EMF and computer are interconnected.
A CO2 tank that supplies CO2 to the incubator is located next to the incubator. The water
bath, which is used to adjust the media temperature to 37 degrees, is located next to the
incubator. B) An internal view of the RM-EMF exposer. C) The above shows that the exposed
W is kept constant. The below shows it is constant that the percent of CO2 and temperature
inside incubator and the temperature inside the exposer. D) It was confirmed that the
temperature of medium did not rise above 37 degrees during exposing RF-EMF.The RF-EMF
was exposed to the cells for 24 hours at an intensity of 8W/kg.

PB-116 [14:30]
Effect of chronic RF-EMF exposure on aging-induced oxidative stress and

745
neuroinflammation in C57BL/6 mice
Ye Ji Jeong1, Hyung-Do Choi2, Jeong-Ki Pack3, Nam Kim4, Yun-Sil Lee5 & Hae-June Lee1
1Division of Basic Radiation Bioscience, Korea Institute of Radiological and Medical Sciences, Seoul, Korea
2Department of EMF Research Team, ETRI, Daejeon, Korea
3Department of Radio Sciences and Engineering, Chungnam National University, Daejeon, Korea
4School of Electrical and Computer Engineering, Chungbuk National University, Cheongju, Korea
5Graduate School of Pharmaceutical Sciences, Ewha Womans University, Seoul, Korea
Keywords: In vivo, RF/Microwaves, Completed (unpublished)
Presented by: Hae-June Lee
In this study, we determine whether chronic RF-EMF exposure affect oxidative stress and
inflammatory response of the brain in aging mice. The middle aged C57BL/6 mice (14 months old)
were exposed to 1,950 MHz electromagnetic fields for 8 months (SAR 5W/kg, 2 h/day, 5
days/week). No changes were observed in the protein levels of DNA damage, apoptosis, GFAP,
Iba-1 and heat shock proteins by long-term RF-EMF exposure.
Introduction
As people are continuously exposed to electromagnetic fields from mobile phone in everyday life
and due to its the proximity of mobile phones to the head, public concern whether they can be
harmful to the brain has raised [1]. The use of mobile phone has gradually increased over the last
10 years, its cumulative exposing periods of users is also increasing. Radiofrequency
electromagnetic field (RF-EMF) is known to interact with biological systems and may cause
oxidative stress under certain circumstances. Many neurological diseases, including Alzheimer’s
disease (AD) and amyotrophic lateral sclerosis (ALS), are now recognized to share atypical
inflammatory reactions as a major pathological feature [2]. Therefore, it is necessary to investigate
whether long-term exposure to RF-EMF affects oxidative stress and neuroinflammation on the aging
brain.
Methods
C57BL/6 mice at the age of 14 months were assigned to 2 groups (the sham- and RF-EMF-exposed
group, 12 mice per group). Mice in the RF-EMF group were placed in a reverberation chamber and
exposed to 1,950 MHz electromagnetic fields for 8 months (specific absorption rate (SAR) 5 W/kg, 2
h/day, 5 days/week). After RF-EMF exposure for 8 months, alterations in the levels of glial fibrillary
acidic protein (GFAP), ionized calcium binding adaptor molecule 1 (Iba-1), heat shock proteins, DNA
damage and apoptosis were analyzed in brain tissue after by Western blotting,
immunohistochemistry, or qPCR.

Results and conclusion

Compared to young brain, aged brain had increased inflammatory markers (GFAP and Iba-1), lipid
damage marker 4HNE and DNA damage makers (8-OxodG). However, no changes in the protein
levels of DNA damage, apoptosis, GFAP, Iba-1 and heat shock proteins were detected in the aged
brain by chronic RF-EMF exposure. Moreover, no alteration was found in the expression levels of
inflammatory marker genes in aging brain by RF. Therefore, these findings indicate that chronic RF-
EMF exposure did not influence on aging-induced oxidative stress and neuroinflammation in
C57BL/6 mice.

References
[1] Morgan LL, Miller AB, Sasco A, Davis DL. 2015. Mobile phone radiation causes brain tumors and
should be classified as a probable human carcinogen (2A) (review). Int J Oncol. 46(5):1865–1871.
[2] Amor S, Peferoen LA, Vogel DY, Breur M, van der Valk P, Baker D, van Noort JM. 2014.
Inflammation in neurodegenerative diseases--an update. Immunology. 142(2):151–166.

746
exposure.
Acknowledgement
This work was supported by Institute for Information & communications Technology Promotion (IITP)
grant funded by the Korea government (MSIT) (2017-0-00961, Study on the EMF Exposure Control
in Smart Society)

PB-118 [14:30]
STUDENT PAPER
Magnetic field effects on H2O2 and possible biological implications including cancer
Sahithi Kandala1, Rodolfo Bruzon1, Hakki Gurhan1 & Frank Barnes1
1Electrical, Computer and Energy Engineering, University of Colorado, Boulder, Colorado, USA, 80309
Keywords: In vivo, Static, Work in Progress
Presented by: Sahithi Kandala
The links between Reactive oxygen species, Electromagnetic Fields and Biological Implications are
explored. Experimental data is used as background to show the changes that are observed due to
static magnetic fields on fibrosarcoma cells.
This paper talks about the possible pathways that link the change in Hydrogen Peroxide observed
due to Magnetic Fields and the changes in the Biological system due to the variation of Hydrogen
Peroxide. It particularly concentrates on the pathways that could lead to Cancer, Alzheimer's and
Aging. Data from previous studies from the literature and experimental data on fibrosarcoma cells
we have observed serves as the background. The effects due to varying concentrations of ROS
(Reactive Oxygen Species), antioxidants such as SOD (Superoxide Dismutase) and NADH
(Nicotinamide Adenine Dinucleotide – Hydrogen) are discussed.
The changes observed in the concentration levels of Hydrogen Peroxide and others (such as Ca2+)
may increase the rate of mutation as well as induce mutation. Mutation happens as a result of
changes in the activation/deactivation of protein/enzymes. These changes in activation/deactivation
may be due to changes in the transcription factor, which can be modified by changes in radical
concentrations. Carcinogens can change the rate of production of radicals and the possible
carcinogens considered here are Static Magnetic Field and Time-Varying Electromagnetic Field.
Hydrogen Peroxide is a ROS present in biological systems. It is generated by the dismutation of
Superoxide (O2-). Changes in Oxygen radicals and other ROS, in this instance Hydrogen Peroxide,
may cause modifications of proteins. There are two types of changes that occur in a biological
system due to environmental conditions (here External Magnetic Fields): Adaptive Response and
Pathological Changes. Here, we concentrate on the pathological changes (activation/deactivation of
Transcription Factors).

PB-120 [14:30]
Generation of a syngeneic orthotopic graft murine model of peritoneal carcinomatosis for
studying in vivo nsPEF anticancerous effects
Abdelkader Taibi1, Marie-Laure Perrin2, Lynn Carr2, Catherine Yardin1, Sylvaine Durand-Fontanier1 & Sylvia
M. Bardet2
1CNRS, CHU Limoges, XLIM, UMR 7252, University of limoges, Limoges, France, F-87000
2CNRS, XLIM, UMR 7252, Limoges, France, F-87000
Keywords: In vivo, Pulsed, Work in Progress
Presented by: Sylvia M. Bardet
Colorectal cancer (CRC), a common cancer with poor prognosis, affects the peritoneum by
metastasis, leading to an incurable peritoneal carcinomatosis (CPc). High-intensity nanosecond

747
pulsed electric fields (nsPEF) have emerged as a promising tool for tumor ablation. We have
developed 4 models of CPc suitable for in vivo nsPEF studies, by syngeneicintravenous,
subcutaneous, intraperitoneal or laparotomygraft of CT-26 cells in immunocompetent mice, and
compared growth by bioluminescence. Tumoral tissue was observed by multiphoton microscopy. In
vitro CT-26 cells were analyzed for their sensitivity to nsPEF (survival, mitochondrial potential,
permeabilization). Our CPc models will enable robust experiments for future in vivo nsPEF analysis.
Introduction
Carcinoma of the colon and rectum, known as colorectal cancer (CRC), is a common cancer that
causes around 700000 deaths annually, with more than 3.5 million patients living with it [1]. In
addition to the liver, which is the most frequently affected organ, the peritoneum, the membrane that
covers the entire abdominal cavity and viscera, is another metastatic site (found in 10-25% of cases)
[2,3]. Until recently, oncologists and digestive surgeons have considered colorectal peritoneal
carcinomatosis (CPc, the implantation of tumor cells throughout the peritoneal cavity) as an
incurable disease [4]. Currently, the only treatment is based on a complete cytoreduction associated
with hyperthermic intraperitoneal chemotherapy, supervised by systemic chemotherapy [5]. This
treatment has improved the long-term survival of these highly selected patients with a median
survival of between 22 and 60 months [6,7].
However, this curative surgical treatment is only available to patients with resectable CPc. The
majority of patients are considered unresectable at the time of diagnosis [8] and are not able to
access this treatment, despite systemic neoadjuvant chemotherapy. In addition, this surgical
treatment is associated with a morbidity of 30% in these already fragile patients [9].
It is for these reasons that it is imperative to work on therapeutic alternatives to treat the many
patients with unresectable, peritoneal carcinomatosis.
High-intensity pulsed electric fields with nanosecond durations (3-300ns; nsPEFs) have emerged as
a promising tool for tumor ablation [10-12]. The physical mechanisms and specific type of cell death
that occurs following the application of nsPEFs is under investigation by several groups and initial
studies have shown effects on cell signaling through calcium and cell death pathways [13-14],
mitochondria [15] and cytoskeleton [17-19].
In order to develop a new model of peritoneal colorectal cancer in mouse suitable for studies on
nsPEF treatment, we grafted murine colon carcinoma CT-26 cells expressing luciferase in
immunocompetent BALB-c mice by intravenous, subcutaneous, intraperitoneal injection. In a fourth
model, the injection of CT-26 cells was performe after scraping the entire peritoneum (laparotomy).
Tumoral growth was then followed by bioluminescence, over 15 days post-grafting. In vitro CT-26
cells were analyzed by flow cytometry for their sensitivity to nsPEFs (survival, mitochondrial
potential and permeabilization). In parallel, peritoneal tumoral tissue were observed by multiphoton
microscopy in order to evaluate the tumor microenvironment.
Methods
1. Cell culture and flow cytometry
CT-26 cells were grown at 37°C with 5% CO2 in supplemented Modified Essential Medium (10%
fetal bovine serum, 0.2% Glucose, 2 mM L-glutamine, 100 U/ml penicillin 100 μg/ml streptomycin).
The cells were suspended at 1.2 million/ml in complete medium in a volume of 55 µl in a wideband
electroporation cuvette containing complete medium (conductivity of 1.5 S.m-1) or with 10 µM
CCCP treatment for mitochondrial depolarization controls. We exposed volumes of cells for a
duration that limited the sedimentation (less than a minute). The cells were homogeneously
distributed within the whole volume with respect to the electric field. Due to the ultra-short duration
of the pulses, it can be assumed that the electric field does not induce significant displacement of
the cells. The temperature variation inside the exposed sample was measured using a fluoro-optic
probe (Luxtron). Immediately after exposure the cells were transferred to tubes and co-labeled with
3 nM of TMRM and 3 μM of 7-AAD. Determination of the mitochondrial membrane potential (ΔΨm,
by TMRM) and cell viability and permeabilization (by 7-AAD) was simultaneously measured using a

748
flow cytometer (Muse® Cell Analyzer, Millipore).
2. In vivo model
5-weeks old Balb-c mice (Charles Rivers, France)were grafted with firefly luciferase-expressing
murine CT-26 colon carcinoma cells by intravenous, subcutaneous, intraperitoneal or laparotomy-
combined injection and followed the tumoral growth by bioluminescence over 15 days post-grafting.
The injection volume was 40 µL with 105 cells and mice maintained under isoflurane anesthesia. 12
mice from each group were studied daily between day 3 and 15 postgrafting. First, mice were
injected i.p. with 25 mg/kg of XenoLight D-Luciferin under isoflurane anesthesia and imaged with a
CCD camera system (IVIS® Lumina, P. Elmer).A circular region of interest (ROI) was placed over
the tumor, and the total signal (photons/s) in this ROI was determined using Image Analyst MKII.
3. Exposure system
We developed in previous studies in vitro and in vivo nsPEF exposure setup [20,21] composed of an
nsPEF generator, a digital phosphor oscilloscope (DPO), a high-voltage measurement device (tap-
off), an electrode-based delivery system and a microscope stage. The nsPEF generator (FPG
10-1NM-T, FID Technology, Germany) delivered 10ns pulses with amplitudes between 4.5kV and
10kV and with rise-times around 5ns. A 1GHz oscilloscope (DPO 4104, Tektronix, USA) using 30dB
attenuator connected to the 40dB attenuation tap-off was used to display the time-domain
measurements of each delivered electrical pulse. The numerical dosimetry of the nsPEF exposure
setup was performed using numerical modeling and full-wave 3D simulations based on a
discretization of Maxwell's equations in integral form [22,23].
4. Multiphoton fluorescence
Peritoneal biopsy or histologic sections were positioned on a stage of a customized Olympus
multiphoton microscope BX61WI/FV1200MPE with a 25X immersion objective (1.05NA) coupled
with a tunable femtosecond Ti:Sapphire pulsed laser (Chameleon Ultra II, Coherent) for the
excitation. Image stacks were taken at 2μm steps, acquired under 810nm excitation for second
harmonic generation (SHG, collagen) and auto-fluorescence (elastin) wavelength with FluoView
FV1200 software (Olympus). The different components of the emitted light from the sample were
separated using a dichroic mirror (570nm) and detected by a pair of photomultiplier tubes.The
images obtained were analyzed with Imaris software (Bitplane AG) or Fiji/ImageJ (NIH).
Results
The dosimetry of the in vivo and in vitro delivery systems showed that the electric field in the central
region between the electrodes had a good homogeneity, as based on 3D numerical simulations. The
intensity of the electric field than can be delivered in vitro in the electroporation cuvette was around
60 kV/cm and in vivo to the tumour was between 30-50 kV/cm (Fig.1c).
Using conventional in vivo bioluminescence imaging of wild-type green-emitting luciferase, we
characterized and compared the growth rates and dissemination of luc CT-26 tumors between the
four different grafting methods (Fig.1b). The aim was to obtain a limited peritoneal carcinomatosis
using peritoneal carcinomatosis index in mice (IPC <10). Upon implantation, intravenously injected
cells did not demonstrate a significant growth rate and subcutaneous injection resulted in a limited
carcinosis in the peritoneal cavity. Both intraperitoneal or laparotomy injection showed elevated
growth rates at 15 days post-surgery. The IPCmedianfor subcutaneous, intraperitoneal or
laparotomy groups was 5 (0-13), 9 (1-12) and 24 (3-29) respectively.Anatomic and histologic
analysis corroborated the existence of various tumoral nodules. Multiphoton microscopy showed
tumoral fibrosis (SHG) and infiltrated tumoral cells, in comparison with non-pathologic peritoneum
(Fig.1a).
In vitro data are also consistent with our previously published results on 5 tumoral cell lines [21], we
observed a strong dose-dependent effect of nsPEF on cell viability.
Conclusion
We have compared and developed a model of peritoneal colorectal cancer in mouse suitable for

749
nsPEF studies based on grafting murine colon carcinoma CT-26 cells expressing luciferase in
immunocompetent BALB-c mice by intravenous, subcutaneous, intraperitoneal or laparotomy
injection. We have demonstrated that these CT-26 cells are sensitive to 10 ns nsPEF. For futurein
vivo nsPEFstudies, we have chosen the i.p. injection mode as it provided a fast, but limited, tumoral
growth. We can then combine fluorescent and bioluminescent reporters to measure tumour
regression and visualize the signaling pathways activated in vivo in the hours and days following
nsPEF application.
Acknowledgments
This study has received funding from the European Union’s Horizon 2020 Framework Programme
FET OPEN Research and innovation under grant agreement N° 737164 (SUMCASTEC).This
research was conducted in the scope of LEA-EBAM, a European Associated Laboratory titled
“Pulsed Electric Fields Applications in Biology and Medicine”.
References
1. Ferlay et al.(2013) "Cancer incidence and mortality ...". Eur J of Cancer.
2. de Cuba et al. (2013) "Cytoreductive surgery and HIPEC for peritoneal ...". Cancer Treat Rev.
3. Koppe et al. (2006) "Peritoneal carcinomatosis of colorectal origin". Ann Surg.
4. Ceelen and Bracke (2009) "Peritoneal minimal residual disease in colorectal cancer". Lancet
Oncol.
5. http://www.snfge.org/content/4-cancer-colorectal-metastatique.
6. Goéré et al. (2013) "Is There a Possibility of a Cure in Patients With Colorectal Peritoneal
Carcinomatosis ..." Ann Surg
7. Verwaal et al. (2005) "Long-term survival of peritoneal carcinomatosis of colorectal origin". Ann
Surg Oncol.
8. Pelz et al. (2010) "Evaluation of best supportive care and systemic chemotherapy...". BMC
Cancer.
9. Malfroy et al. (2016) "Complications after cytoreductive surgery with hyperthermic intraperitoneal
chemotherapy for treatment of peritoneal carcinomatosis". Surg Oncol.
10. Chen et al. (2012) "Long term survival of mice with hepatocellular carcinoma after pulse power
ablation with nanosecond pulsed electric fields." Technol Cancer Res Treat.
11. Nuccitelli et al. (2006) "Nanosecond pulsed electric fields cause melanomas to self-destruct."
Biochem Biophys Res Commun.
12. Chen et al. (2012) "Tumor ablation with nanosecond pulsed electric fields." Hepatobiliary
Pancreat Dis Int.
13. Zhang et al. (2008) "Nanosecond pulse electric field (nanopulse): a novel non-ligand agonist for
platelet activation." Arch Biochem Biophys.
14. Morotomi-Yano et al. (2014. "Different involvement of extracellular calcium in two modes of cell
death induced by nanosecond pulsed electric fields." Arch Biochem Biophys.
15. Pakhomova et al.(2014). "Calcium-mediated pore expansion and cell death following
nanoelectroporation." Biochim Biophys Acta.
16. Beebe et al. (2013). "Induction of Cell Death Mechanisms and Apoptosis by Nanosecond Pulsed
Electric Fields (nsPEFs)" Cells.
17. Thompson et al. (2014). "Calcium influx affects intracellular transport and membrane repair
following nanosecond pulsed electric field exposure." J Biomed Opt.
18. Rassokhin et al. (2014). "Cellular regulation of extension and retraction of pseudopod-like blebs
produced by nanosecond pulsed electric field (nsPEF)" Cell Biochem Biophys.
19. Pakhomov et al. (2014) "Disassembly of actin structures by nanosecond pulsed electric field is a

750
downstream effect of cell swelling." Bioelectroch.
20. Bardetet al. (2016) "Multiphoton imaging reveals that nanosecond pulsed electric fields collapse
tumor and normal vascular perfusion in human glioblastoma xenografts". Sci Rep.
21. Soueid et al.(2018) "Delivery devices for exposure of biological cells to nanosecond pulsed
electric fields". Med Biol Eng Comput.
22. Kenaan et al. (2011) "Characterization of a 50-Ω Exposure Setup for High-Voltage Nanosecond
Pulsed Electric Field Bioexperiments." Biomedical Engineering, IEEE Trans.
23. Wu et al. (2013) "Moveable Wire Electrode Microchamber for Nanosecond Pulsed Electric-Field
Delivery." Biomedical Engineering, IEEE Trans.
Figures

Figure 1. a) Two-photon imaging of the peritoneum (SHG signal in green, elastin auto-
fluorescence red), b) Assessment of in vivo tumor growth : bioluminescence imaging of CT-26
tumors 15 days after laparotomy in mice, c) nsPEF exposure setup : spatial distribution of the
electric field magnitude of the delivery system around the tumor for a 6.1-kV delivered voltage
in a 2D vertical plane perpendicular to the electrode tip.

751
PB-122 [14:30]
STUDENT PAPER
Cellphone radiofrequency signal induced inflammatory response oxidative stress and effect
on memory in rat brain
Kumari Vandana1, 2, Jayprakash Nirala1, 2, Ranjan Nanda2 & Paulraj Rajamani1, 2
1School of Environmental Sciences, Jawaharlal Nehru University, New Delhi, India, 110067
2Translational health group, International Center for Genetic Engineering and Biotechnology, New Delhi,
India, 110067
Keywords: In vivo, RF/Microwaves, Work in Progress
Presented by: Kumari Vandana
The exponential worldwide growth in the use of mobile phones has raised concerns regarding
potential harmful effect of exposure to radiofrequency modulated electromagnetic field radiations. Of
particular concern has been the effect on brain and hippocampus in specific. Present study was
undertaken to determine if long term exposure to RF cause inflammatory cytokines mediated
oxidative stress in brain and if there is any functional consequence of it in terms of memory
impairment.
Introduction
The exponential worldwide growth in the use of mobile phones has raised concerns regarding
potential harmful effect of exposure to radiofrequency modulated electromagnetic field radiations. Of
particular concern has been the effect on brain and hippocampus in specific. Oxidative stress has
been proposed as the possible underlying mechanism responsible for radio frequency (RF) effects
on nervous system [1]. It has also been known that Stressful events affect the immune system by
reducing the cellular response to mitogen stimulation, decreasing production of natural killer cell
activity and altering levels of cytokines. An important and detrimental consequence of increased
cytokine production is the induction of nitric oxide (NO) and reactive oxygen species [2]. However
reports on direct correlation between exposure to RF radiation and immune response has been
inconsistent. Thus the present study was undertaken to determine if long term exposure to RF
cause inflammatory cytokines mediated oxidative stress in brain and if there is any functional
consequence of it in terms of memory impairment.
Method
Study was carried out on twelve adult male Wister rats randomly divided in to two groups, sham
exposed (n = 6) and exposed (n =6 ). Animals of exposure group were subjected to cellphone
radiation at frequency 1966.1MHz with SAR value of 0.1839W/Kg in a specially designed anechoic
chamber for 90 days (2hr/day, 5 days a week). At the end of exposure period animals were
subjected to behavioral paradigm (contextual fear conditioning) and at the end of it animals were
sacrificed to collect blood sample and isolate hippocampus. Level of inflammatory cytokines (IL1,
IL6, TNFα), stress hormones (Corticosterone & ACTH) and oxidative stress markers were estimated
in the serum and hippocampus respectively.
Result & Discussion
Long term exposure to cell phone radiofrequency signal resulted in significant increase in level of
inflammatory cytokines (IL1, IL6, TNFα) and stress hormone (CORT, ACTH) in exposed group as
compared to sham exposed. A significant increase in level of oxidative stress markers viz. reactive
oxygen species (ROS), malonaldehyde (MDA) and protein carbonyl (PCO) and decrease in total
antioxidant capacity (TAC) was observed in exposed group. However decrease in freezing memory
was pronounced but not significant in exposure group. Present study indicate that RF exposure may
have acted as a chronic stressor and activated immune response. These blood born cytokines may
have crossed the blood brain barrier and induced oxidative stress in brain and activated
hypothalamic-pituitary-adrenal (HPA) axis resulting in increased level of corticosterone.
Corticosterone has strong influence on hippocampal dependent memory due to abundant presence

752
of receptors [3, 4]. Corticosterone have both memory impairment as well as enhancement effect
depending on interaction between level of corticosterone and state of mind (sleep or awake) and
mechanism is different in both cases [5]. In the present study there is no significant impairment in
memory. It might be due to adaptive effect of high level of corticosterone.
Conclusion
Present study shows that long term RF exposure induces immune response as well as a state of
oxidative stress in brain which may affect normal functioning if exposed for longer period.
References
[1] Megha, K., Deshmukh, P. S., Banerjee, B. D., Tripathi, A. K., Ahmed, R., & Abegaonkar, M. P.
(2015). Low intensity microwave radiation induced oxidative stress, inflammatory response and DNA
damage in rat brain. Neurotoxicology, 51, 158-165.
[2] Hu S, Peterson PK, Chao CC (1998) Kappa-opioid modulation of human microglial cell
superoxide anion generation. Biochem Pharmacol 56: 285–288.
[3] Lupien SJ, Lepage M. Stress, memory, and the hippocampus: can't live with it, can't live without
it. Behav Brain Res. 2001;127:137–158.
[4] Joëls M. Functional actions of corticosteroids in the hippocampus. Eur J Pharmacol.
2008;583:312–321.
[5] Kelemen, E., Bahrendt, M., Born, J., & Inostroza, M. (2014). Hippocampal corticosterone impairs
memory consolidation during sleep but improves consolidation in the wake state. Hippocampus,
24(5), 510-515.

PB-124 [14:30]
Effect of radiofrequency electromagnetic fields on in vivo C6 brain tumors in Wistar rats
Anne-Sophie Villegier1 & Nihal Ouadah1
1Experimental toxicology, INERIS, Verneuil-en-Halatte, France, 60550
Keywords: In vivo, RF/Microwaves, Completed (unpublished)
Presented by: Anne-Sophie Villegier
Previous literature suggested that brain cancer patients with heavy cell phone use showed reduced
survival. Here we aimed to test survival and tumor aggressiveness in the C6 cells glioblastoma
model in Wistar rats subjected to controlled brain averaged specific absorption rate (BASAR)
induced by radiofrequency electromagnetic field (RF-EMF) exposures. For the first time, BASAR-
dependent effects were reported in vivo by the reduction of apoptosis and by the reduction of
inflammation in the tumor micro-environment. In accordance with the literature, our data did not
suggest any impact on survival or tumor volume. Further replication studies are needed to confirm
these observations.
Purpose
Glioblastoma is a malignant brain tumor which has one of the poorest prognosis. It is not clear if
toxic environmental factors can influence its aggressiveness. Recently, it was suggested that brain
cancer patients with heavy cell phone use showed reduced survival.
Here we aimed to assess the effect of controlled brain averaged specific absorption rate (BASAR)
from cell phone radiofrequency electromagnetic fields (RF-EMF) on in vivo C6 brain tumors in
Wistar rats.
Methods
C6 cells grafted male rats were exposed to GSM 900 MHz signal at environmental BASAR, 0
(sham), 0.25 or 0.5 W/kg (5 days a week, 45 min a day in restraint), or were cage controls (no
restraint). At death, tumor volume and immunohistochemistry for CD31, cleaved caspase (CC) 3

753
and Ki67 were assessed to examine vascularization, apoptosis and cellular divisions, respectively.
Index of inflammation, necrosis, infiltration and mitosis were determined.
Results
Results showed no BASAR effect on survival (31 days post-graft median), tumor volume, mitotic
index, vascularization, infiltration, necrosis or cell division. However, results suggested a BASAR-
dependent reduction of inflammation and apoptosis.
Conclusions
Our data suggested an action of RF-EMF by reducing inflammation and glioblastoma cell apoptosis,
at probably too low amplitude to impact survival. Further replication studies are needed to confirm
these observations.
Figures

Figure 1. Absence of BASAR-related effect on survival in the C6 GBM rat model. Seven days
after their graft, rats were exposed to RF-EMF at 0 (n = 31), 0.25 (n = 18) or 0.5 W/kg BASAR
(n = 39) or were cage controls (CC, n = 15). There was no effect of RF-EMF exposure or
restraint on survival.

754
Figure 2. Grades for (a) inflammation (b) necrosis (c) infiltration of immune cells, and (d) CD31
immunohistochemistry in the C6 GBM rat model subjected to RF-EMF exposures. At death,
brain were collected and histochemical observations were performed on HE colored brain
sections. Rats were cage controls (CC, n = 15), or RF-EMF exposed at 0 W/kg (n = 31), 0.25
W/kg (n = 18) and 0.5 W/kg (n = 39). Inflammation was reduced by the RF-EMF exposure with
a BASAR-dependent effect.

755
Figure 3. (a) CC3, (b) Ki67 immunohistochemistry and (c) mitotic index in the C6 GBM rat
model subjected to RF-EMF exposures. At death, brains were collected and histochemical
observations were performed on HE colored brain sections. Rats were cage controls (CC, n =
15), or RF-EMF exposed at 0 W/kg (n = 31), 0.25 W/kg (n = 18) and 0.5 W/kg (n = 39). CC3
immunohistochemistry was reduced by the RF-EMF exposure with a BASAR-dependent effect
after adjustment on survival (a: graphic and picture).

756
PB-126 [14:30]
An investigation into a proteomics analysis of breast cancer cells during ELF-MF induced
apoptosis
Reyhane Ghadirian1 & Alireza Madjid Ansari1
1Integrative Oncology Department, Motamed Cancer Institute, ACECR, Tehran, Iran, 1517964311
Keywords: Mechanistic/Theoretical, ELF/LF, Work in Progress
Presented by: Alireza Madjid Ansari
Protein expression analysis has become a promising tool in exploring apoptotic events in breast
cancer cells. In search of a more effective treatment, extremely low frequency magnetic fields (ELF-
MF) have been suggested as a possible adjuvant therapy.
Introduction
Breast cancer is one of the most common cancers among women worldwide. Finding new specific
and more effective solution, extremely low frequency magnetic field (ELF-MF) has been suggested
as a possible adjuvant therapy for cancer treatment. It has been shown that ELF-MF can induce
apoptosis pathway in several cancer cells, including breast and prostate cancers, but its mechanism
of action is still ambiguous regarding to the effects on growth, development and more importantly
the potential positive outcomes of this field in apoptotic response. In this study, protein expression of
normal and cancerous breast cells were exposed to ELF, and analyzed using proteomics approach
in order to elucidate the possible mechanism of action leading to cell death.
Methods
Breast cancer cell line BT-474 and non-tumorigenic epithelial cell line MCF10-A were cultured in
DMEM with 10% FBS. They were then exposed to 1 mT and 100 Hz (ELF-EMF) 2 h/day for 2, 3, 4
and 5 days. Flowcytometry was conducted in order to determine and analyze apoptotic events. 2D
gel electrophoresis, image analysis and mass spectrometry were performed to find out subtle
changes in protein expression after and during exposure.
Results
Apoptosis was examined in cells exposed to magnetic fields and increase of apoptosis was
observed compared to sham (control). In addition, proteomics studies and Bioinformatics analysis
on the obtained 2D electrophoresis and mass spectrometry showed that exposure to 1 Hz, 100 mT
ELF-MF led to alteration of the protein expression pattern of cells in exposed groups in spite of the
sham exposure.
Discussion
The proteomic analysis strategy of matched normal and malignant breast cancer cells from invasive
breast carcinoma was performed. This strategy resulted in a better understanding of proteomics
approaches in evaluation of extremely low frequency electromagnetic field effects on breast cancer
cells during apoptosis. In addition, cells exposed to ELF were characterized with inhibition of cell
division and induction of pro-apoptotic protein expression, as shown by flowcytometry.
Conclusions
Indeed, it has been revealed that exposure to ELF-MF have major influences on the increase of
apoptosis and changing protein expression. Altogether, this study clarifies that further proteomic
analyses may shed light upon recent hypotheses (i.e. Radical pair mechanism) pertaining to how
ELF exerts its effects in cells.

PB-128 [14:30]
Effects of radiofrequency fields on RAS and ERK kinases activities in live cells using the

757
real-time Bioluminescence Resonance Energy Transfer technique (BRET)
Emmanuelle Poque-Haro1, Delia Arnaud-Cormos2, Hermanus Ruigrok1, Florence Poulletier De Gannes1,
Annabelle Hurtier1, Rémy Renom1, Isabelle Lagroye1, Bernard Veyret1, Philippe Leveque2 & Yann
Percherancier1
1Bioelectronic group , Bordeaux University, Talence, France, 33400
2Bioelectromagnetics team, Limoges University, Limoges, France, 87060
Keywords: Mechanistic/Theoretical, Static, Completed (unpublished)
Presented by: Isabelle Lagroye
Using ERK and RAS molecular probes based on the Bioluminescence Resonance Energy Transfert
technique, we assessed the effect of 1800 MHz exposure (CW and GSM signals) at two different
levels: 6W/kg and 1.5 W/Kg, on ERK and RAS activities in HuH7 cells.
Introduction
The mitogen-activated protein kinase (MAPK) pathway plays a role in the regulation of gene
expression, cellular growth, and survival. In this signaling pathway, Ras activates three mitogen-
activated protein kinases (MAPKs) including ERK, JNK, and p38. Abnormal MAPK signaling may
lead to increased or uncontrolled cell proliferation and resistance to apoptosis. Research into the
MAPK pathway has shown it to be important in some cancers and some bioelectromagnetics
studies suggest that both RF and ELF may affect it through the modulation of ERK activity. Using
ERK and RAS molecular probes based on the Bioluminescence Resonance Energy Transfert
technique, we assessed the effect of 1800 MHz exposure (CW and GSM signals) at two different
levels: 6W/kg and 1.5 W/Kg, on ERK and RAS activities in HuH7 cells.
Method
BRET, like FRET, is a biophysical technique used to study interactions and conformational changes
in proteins by measuring the non-radiative energy transfer between a bioluminescent energy donor
(luciferase or Luc) and a fluorescent acceptor, e.g., "Yellow fluorescent Protein" (YFP), fused to
proteins of interest. In the present study, HuH7 cells transiently expressing RAS or ERK BRET
probes were sham-exposed or exposed to RF for 24h before being activated with increasing doses
of PMA for 15 minutes. The BRET signal was recorded immediatly after PMA activation.
Results
BRET basal signal is not modified and neither efficacy nor potency of PMA to activate ERK or RAS
activities are modified following RF exposure in comparison to the sham condition.
Conclusion
We did not find any evidence for an effect of RF exposure (CW and GSM signals) on RAS and ERK
proteins.

PB-130 [14:30]
Alterations in photosynthetic system key factors after exposure of Microcystis aeruginosa to
1.8GHz electromagnetic field
Chao Tang1, Ziyan Zhang1, Hui Yu1 & Xiaomei Huang1
1Key Laboratory of Urban Environment and Health, The Institute of Urban Environment (IUE), Chinese
Academy of Sciences (CAS), Xiamen, China, 361021
Keywords: Mechanistic/Theoretical, RF/Microwaves, Work in Progress
Presented by: Chao Tang
Electromagnetic radiation played a crucial role in the phycoerythrin content regulation. M.
aeruginosa may alleviate the electromagnetic radiation damage of photosynthetic system by
adjusting the content of phycobiliprotein.

758
Photosynthesis is the most basic material and energy metabolism in the biosphere. Light is an
electromagnetic wave. Quantity (duration and intensity) and quality (color or wavelength) of
electromagnetic waves are important factors that are involved in regulating growth and physiological
processes in photosynthetic organisms. Electromagnetic radiation is an important environmental
factor. How electromagnetic radiation affects photosynthesis and whether electromagnetic radiation
and light have a homologous receptor or receptors should be investigated.
Microcystis aeruginosa belongs to the order cyanophyta, which is the dominant population in many
bloom-forming lakes, which occupy important niches in the environment and is a common habitat of
algae with high environmental sensitivity. The most common and widely used mobile communication
frequency in mainland China is 1.8 GHz. The effects of 1.8 GHz and 40 V/m electromagnetic
radiation on chlorophyll fluorescence, the activities of photosynthesis catalytic enzymes (PGK,
GAPDH, TPI, FBPase, SBP) and the content of phycobiliprotein of Microcystis aeruginosa were
examined to explore the mechanism of electromagnetic radiation on photosynthesis.
The results showed that the value of Fv/Fm, Yield and the activities of photosynthesis catalytic
enzymes (PGK, GAPDH, TPI, FBPase, SBP) in M. aeruginosa cells decreased under the
electromagnetic radiation exposure, but no significant effect was observed. The content of
phycobiliprotein (PC, AP, PE) was significantly decreased after 0.5h, 4h, 12h and 48h exposure to
electromagnetic radiation (P <0.05). After 1h, 16h electromagnetic radiation exposure, the content of
phycobiliprotein (PC, AP, PE) increased significantly (P <0.05).
In conclusion, electromagnetic radiation played a crucial role in the phycoerythrin content regulation.
M. aeruginosa may alleviate the electromagnetic radiation damage of photosynthetic system by
adjusting the content of phycobiliprotein.
Figures

Figure 1. Chlorophyll fluorescence of Microcystis aeruginosa (The control mark was C and the
processing marker was E). (To exclude the impact of light on the experiment, a dark condition
was chosen for the experiment. The exposure group was treated with 1.8GHz and 40V/m
electromagnetic radiation in the dark for 24 hours, whereas the control group was not exposed
to electromagnetic radiation and other conditions remained constant.)

759
Figure 2. The activities of photosynthesis catalytic enzymes (PGK, GAPDH, TPI, FBPase,
SBP) of Microcystis aeruginosa (The control mark was C and the processing marker was E).
(To exclude the impact of light on the experiment, a dark condition was chosen for the
experiment. The exposure group was treated with 1.8GHz and 40V/m electromagnetic
radiation in the dark for 24 hours, whereas the control group was not exposed to
electromagnetic radiation and other conditions remained constant.)

760
 Figure 3. The content of phycobiliprotein of Microcystis aeruginosa (The control mark was C
and the processing marker was E). (To exclude the impact of light on the experiment, a dark
condition was chosen for the experiment. The exposure group was treated with 1.8GHz and
40V/m electromagnetic radiation in the dark for 0.5, 1, 2, 4, 8, 12, 16, 20, 24, 36, 48 hours,
whereas the control group was not exposed to electromagnetic radiation and other conditions
remained constant.)

PB-132 [14:30]
Study on the effects of oxidative stress of Microcystis aeruginosa under the exposure of
electromagnetic radiation
Hui Yu1, Chao Tang1, Xiaomei Huang1 & Chuanjun Yang1
1Key Laboratory of Urban Environment and Health, The Institute of Urban Environment (IUE), Chinese
Academy of Sciences (CAS), Xiamen, China, 361021
Keywords: Mechanistic/Theoretical, RF/Microwaves, Completed (published)
Presented by: Hui Yu
The effect of the activities of Antioxidant enzyme systems (SOD, MDA) of Microcystis aeruginosa
were studied under the exposure of three kinds of typical electromagnetic radiations. The results
showed that oxidative stress of M.aeruginosa could be induced under the exposure of
electromagnetic radiation.
The number of artificial sources of electromagnetic radiation in modern society has increased
tremendously because of the current demands for electricity, telecommunication services, and
electronic devices. The electromagnetic environment in cities is becoming increasingly deterioration.
Electromagnetic radiation not only has a potential long-term effect and threat to public health but
may also impact the ecological environment. Due to the water eutrophication, algae bloom occurred

761
frequently. Large-scale blooms and red tides pose a serious threat to ecosystems. Whether
electromagnetic environment affects the growth of cyanobacteria, and how electromagnetic
radiation works on cyanobacteria should be investigated.
In this paper, the effect of the activities of Antioxidant enzyme systems (SOD, MDA) of Microcystis
aeruginosa were studied under the exposure of three kinds of typical electromagnetic radiations. To
exclude the impact of light on the experiment, a dark condition was chosen for the experiment. The
exposure group was treated with 1.8GHz and 40V/m radiofrequency electromagnetic fields (RF-
EMF), 50Hz and 3mT extremely low frequency electromagnetic fields (ELF-EMF), and 5000V
alternating electric field (AEF) in the dark for 0.5, 1, 2, 12, 24, 48 hours, whereas the control group
was not exposed to electromagnetic radiation and other conditions remained constant.
The results showed that the activities of SOD in M. aeruginosa cells increased significantly under
the RF-EMF, ELF-EMF and AEF exposure of both short (0.5h) and long term (48h) (p＜0.01). The
contents of MDA in M. aeruginosa cells increased significantly under the short term (2h) exposure of
RF-EMF, ELF-EMF, and AEF (p＜0.01). The contents of MDA in M.aeruginosa cells decreased
significantly under the long term (48h) exposure of RF-EMF and AEF (p＜0.01). The results
mentioned above indicated that oxidative stress of M.aeruginosa could be induced under the
exposure of RF-EMF, ELF-EMF and AEF.
Figures

Figure 1. SOD activity of Microcystis aeruginosa

762
 Figure 2. MDA content of Microcystis aeruginosa

PB-134 [14:30]
Wide-band evaluation of EMF emitted by portable computer devices
Krzysztof Gryz1, Jolanta Karpowicz1 & Patryk Zradziński1
1Laboratory of Electromagnetic Hazards, Central Institute for Labour Protection - National Research Institute
(CIOP, Warszawa, Poland, 00-701
Keywords: Occupational, All Frequencies, Work in Progress
Presented by: Krzysztof Gryz
The aim is identifying and evaluating the electromagnetic field (EMF) emitted by portable computer
devices (e.g. laptops, tablets). The electronic components of portable devices (external AC/DC
power supply, internal graphic card and processor interfaces, wireless communication modules etc.)
are wide-band sources of EMF, currently from static field up to microwave of approximately 6 GHz
frequency. The broadband and selective measurement of EMF near laptops housing were
performed. Obtained results shown that dominant components of EMF exposures near portable
computers are: ELF magnetic field emitted by AC/DC adapters and RF electric field emitted by WiFi
interfaces (2G or 5G).
Introduction
Computer devices, of common use in the work and living environment, emit electromagnetic fields
(EMF) of frequency from static till several GHz. Traditional desktop computers and screens are
replacing by small, light, portable devices (laptops, tablets), functioning in various configurations
(e.g. with or without active function of wireless communication and internet access). EMF exposure
and hazards near desktop computers (used at least in the distance of 50 cm from operator) are

763
considered as negligible, and many times it was confirmed by measurement campaigns that
operators exposure is compliant with the exposure limits provided with respect to protection of
workers, general public (e.g. ICNIRP) and also technical environmental requirements concerning
low emission of EMF given by certified organization (e.g. TCO Development). [1-7].
However, special attention should be draw on portable computer devices because of possibility of
their use near to the operator’s body (even in touch), and because of such exposure scenario –
possibility of much stronger EMF influence than that during the use of desktop computers.
The aim of the presented investigations is identifying and evaluating of the wide-band EMF near the
portable computer devices.
Methods
The investigations covered the spot measurements of root-mean-square (RMS) and exposure
evaluation for in compliance with weighted RMS and weighted poeak values of magnetic flux
density, B (for extremely low frequency, ELF) and electric field strength RMS value, E (for radio
frequency, RF). The investigations were worked out near various laptops (minimum distance of
geometrical centre of used isotropic EMF measurement probes set to be 10 cm away from device
housing) using:
− for B-field spatial distribution: broadband field meter ELT-400 (Narda Safety Test Solutions,
Pfullingen, Germany) – measurement range of magnetic flux density (0.01- 20000) µT in the
frequency band 1 Hz – 400 kHz, with the modes of shaped sensitivity according to
Recommendation 1999/519/EC or Directive 2013/35/EC requirements
− for E-field spatial distribution: broadband field meter NBM-550 (Narda Safety Test Solutions,
Pfullingen, Germany) with probe EF0691 – measurement range of electric field strength (0.4 - 650)
V/m in the frequency band 100 kHz – 6 GHz
− for E-field spatial distribution: selective field meter SRM-3006 (Narda Safety Test Solutions,
Pfullingen, Germany) with probes: 3501/03 and 3502/01– measurement range of electric field
strength (0.01 - 160) V/m in the frequency band 27 MHz – 6 GHz
− for waveform and spectrum identification: scopemeter Fluke 190-104 (Fluke, USA), with a Fast
Fourier Transform (FFT) analyse mode or by selective radiation meter SRM-3006 (Narda Safety
Test Solutions, Pfullingen, Germany).
Measurements were performed for 10 laptops in shielding room (shielding effectiveness exceeding
100 dB in frequency range 1 MHz – 1 GHz) with walls covered by absorber of EMF to eliminate
impact on measurement results from EMF emitted by external wireless communication system
sources and reflected EMF emitted by device under test. During the measurements laptops are
operated in a normal conditions with the use of data transfer via internal WiFi interface to router
located inside shielding room (at distance 5 m).
Results and Discussion
Data collecting is ongoing. Identification and measurements shown that in wide spectrum of EMF
emitted by different electronic components of laptops both components, ELF and RF, are present.
The most significant sources of exposure are: external AC/DC adapters and internal wireless
interfaces. Attention is needed for pulsed AC/DC adapters which emit EMF of fundamental
frequency 50 Hz with significant contents of higher harmonics (Figure 1 present an example of
parameters of waveform of B-field emitted by pulsed AC/CD adapter – up to 13th harmonic at level
exceeding 20% of fundamental). The maximum values of the measured B-field not exceed 10 μT
(near laptop’s AC/DC adapter with output voltage 19 V and power up to 130 W). Taking into account
non sinusoidal characteristic of emitted EMF also weighted B-field was analysed. Normalized results
of B-field according to exposure limits from Recommendation 1999/519/EC were below 10% and
below 3% when normalised according to Low Action Levels from Directive 2013/35/EC. ELF B-field
around laptops is lower.
Results of selective measurements of E-field from WiFi interfaces were below 7 V/m (at frequencies

764
used by laptops: system 2G in the range (2400 – 2484) MHz and system 5G in the range (5150 –
5725) MHz) (fig. 2). Such level of exposure is relatively low in comparison with international
guidelines on the exposure limits (general public limit 61 V/m provided by Recommendation
1999/519/EC, occupational exposure limit 140 V/m provided by Directive 2013/35/EC). In the case
of laptops realise IEEE 802.11h protocol with TPC (transmit power control) option, the level of
emitted E-field may be higher than our results of our measurements when the distance to router is
longer than 5 m used in our measurements protocol.. E-field measured in other frequencies ranges
(e.g. from LCD displays, hard disc and graphic card interfaces) not exceed 1-2 V/m.
Conclusions
Results of the study shown that wide-band EMF emitted by laptops is below international exposure
limits (for general public and workers), however exceed average level of the population exposure
(not exceeding 1 μT @ power frequency (ELF component) and 1 V/m @ RF component [8]). RF
EMF near to laptops may be also higher than 3 V/m and 10 V/m limits used in the immunity tests for
the medical electronic devices, including medical implants in frequency range 80 MHz – 2.7 GHz [9].
Discussed EMF exposures need also attention in the worst case exposure scenario – when portable
computer is used in touch of user’s body, but such exposures need evaluation in the context of
thermal effects of electromagnetic exposure evaluated by specific energy absorption rate (SAR)
values.
Acknowledgments
Research supported in Poland within the National Programme “Improvement of safety and working
conditions” (2017-2019) - within the scope of state services - by the Ministry of Science and Higher
Education, Poland (II.N.18). The CIOP-PIB is the Programme’s main co-ordinator.
References
[1]. European Recommendation, 1999/519/EC; [2]. Directive 2013/35/EC; [3]. ICNIRP 2010. Health
Phys. 2010;99(6):818-836; [4]. ICNIRP 1998. Health Phys. 1998;74(4):494-522; [5]. TCO
Development, TCO Certified Displays 6.0, 5.03.2012; [6] TCO Development, TCO Certified
Desktops 5.0, 11.11.2015; [7]. TCO Development, TCO Certified All-in-one PCs 3.0, 11.11.2015; [8]
SCENIHR, 2015. [9]. IEC 60601-1-2;
Figures

Figure 1. Waveform and spectrum of magnetic field emitted by laptop AC/DC adapter (time
base – 10 ms/div)

765
 Figure 2. Spectrum of electric field emitted by laptop’s WiFi 2G interface

PB-136 [14:30]
STUDENT PAPER
Design of a triple band monopole antenna using parasitic elements with SAR reduction for
WLAN and WiMAX Applications
Hanul Bong1, Niamat Hussain1, Min-Joo Jeong1, Ji Woong Park1, Seungwoo Lee2 & Nam Kim1
1Chungbuk National University, Cheongju, Korea, 28644
2Korea Electric Power Corporation, Naju, Korea, 58217
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Hanul Bong
A triple band antenna consists of a dipole, a monopole and parasitic elements is presented. The
aimed frequency band is WLAN (2.4/5.8 GHz) and WiMAX (3.5 GHz) applications. The parasitic
elements around feed line is introduced for improving the performance of the antenna. The antenna
shows acceptable radiation pattern along with - 10 dB return loss at all three bands. Moreover, the
proposed antenna with reflector got reduction of 85.12 %, 50.06 % and 36.87 % in 1g SAR value at
2.4 GHz, 3.5 Ghz and 5.8GHz, respectively when compared with international standard value.
Introduction
In recent years, with a rapid pace of radio frequency technologies development and various wireless
communications. The design of an antennas which satisfying such a multifunction radio
communication is required to have wide operating band, stable unidirectional radiation pattern and
gain. In addition, Specific Absorption Rate (SAR) is also an important factor in design of wireless
communication antennas. In this paper, we have proposed a triple band antenna which is capable of
multiple functions for nowadays wireless communication systems.
Method
The proposed antenna consists of a dipole, a monopole and parasitic elements. The antenna which
printed on a 1.6 mm thick FR4-PCB (relative dielectric constant = 4.6) has total size of 100 x 75 x
1.6mm3. The antenna was optimized for better performance and low SAR values. The reflector was
used to increase antenna gain and reduce SAR value. The distance between the reflector and the
antenna was around of λ/4, which was optimized for gain and impedance matching of antenna,

766
which was found 25 mm. To calculate SAR value, the antenna was placed at 0 mm from the human
head. The input power of the antenna was 1 W. The SAR analysis was performed using Sim4Life
(SEMCAD-X), simulation software.
Results
Proposed antenna using parasitic elements improve performance such as impedance and gain. The
antenna shows acceptable radiation pattern along with - 10 dB return loss at all three bands. Fig 1.
shows return loss of the antenna with and without reflector, while the radiation pattern has been
shown in Fig. 2. We compared the SAR values of the proposed antennas with and without reflector.
Human phantom models and averaged SAR distributions are displayed in Fig.3. The reflector was
introduced in antenna design to observe effect on SAR value. The average SAR1g value without
reflector was 1.98 W/kg at 2.4 GHz, 2.29 W/kg at 3.5 GHz and 5.42 W/kg at 5.8 GHz. On the other
hand, these values with reflector were decreased to 0.238 W/kg at 2.4 GHz, 0.799 W/kg at 3.5 GHz
and 1.01 W/kg at 5.8 GHz at 2.4 GHz, 3.5 GHz and 5.8 GHz, respectively.
Conclusion
We proposed a triple band antenna using parasitic elements and investigated its radiation and SAR
characteristics. The antenna provided acceptable gain, radiation pattern and - 10 dB return loss
bandwidth at all considered bands. The antenna with reflector has SAR1g values of 0.238 W/kg,
0.799 W/kg and 1.01 W/kg at 2.4 GHz, 3.5 GHz and 5.8 GHz, respectively. The proposed antenna
achieved 85.12 % reduction in SAR1g, at 2.4 GHz, 50.06 % at 3.5 GHz and 36.87 % at 5.8 GHz
compared to the international standard value.
Figures

Figure 1. Reflection loss of antenna with and without reflector

767
Figure 2. Radiation pattern of the antenna with and without reflector.

Figure 3. The 1g averaged SAR distribution when the antenna was placed at 0 mm distance
from human head.

768
PB-138 [14:30]
Trends in scientific research on mobile communications: Temporal and geospatial analysis
based on data of the EMF-Portal
Dagmar Dechent1, Anne-Kathrin Petri1, Lambert Bodewein1, Tanja Emonds1, David Graefrath1, Thomas
Kraus1 & Sarah Driessen1
1Research Center for Bioelectromagnetic Interaction (femu), University Hospital RWTH Aachen, Aachen,
Germany, 52074
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Dagmar Dechent
The EMF-Portal (www.emf-portal.org) summarizes systematically scientific research data on the
effects of electromagnetic fields (EMF) and provides all information free of charge in English,
German and Japanese. The core of the EMF-Portal is an extensive literature database with an
inventory of about 26,000 publications and 6,100 summaries of individual scientific studies. Using
the database of the EMF-Portal we analyzed the global scientific efforts in publications on
medical/biological effects of exposure to mobile phone related radiofrequency EMF. We present
temporal trends in number of publications and analyzed the mostly investigated research topics,
authors with most publications, funding and origin of the publications.
Introduction
Since the introduction in the early-to-mid 1980s, the spread of mobile communication devices has
increased exponentially leading to an estimated 7,740 million number of subscriptions in 2017 with
103.5 subscriptions per 100 inhabitants worldwide [1]. At the same time, concerns about health
effects caused by exposure to radiofrequency electromagnetic fields (RF-EMF) emitted by the
mobile communication devices have been raised and several national (e.g., German Mobile
Telecommunication Research Programme) and international research programs (e.g., Interphone
Study, REFLEX project) were initiated. Based on the scientific knowledge the International Agency
for Research on Cancer (IARC) of the World Health Organization (WHO) classified radiofrequency
electromagnetic fields into group 2B "possibly carcinogenic" to humans [2]. WHO is currently
conducting a formal risk assessment of all studied health outcomes from RF-EMF exposure and the
evaluation should be released in 2018.
In 2005, the internet information platform EMF-Portal (www.emf-portal.org, Fig. 1) of the RWTH
Aachen University, Germany, was launched with the aim to inform the public about potential effects
of exposure to electric, magnetic or electromagnetic fields (EMF). The EMF-Portal summarizes
systematically scientific research data and provides all information free of charge in English,
German and Japanese. The core of the EMF-Portal is an extensive literature database with an
inventory of about 26,000 publications and 6,100 summaries of individual scientific studies on the
effects of EMF. In 2017, an evaluation of the EMF-Portal stated a completeness of the relevant
scientific literature of more than 97% [3].
The aim of the presentation is to analyze trends in scientific literature on mobile communication
devices and medical/biological effects in experimental und epidemiological studies based on the
data in the EMF-Portal.
Methods
We conducted our search in the EMF-Portal and included peer-reviewed experimental and
epidemiological studies on mobile communications devices (i.e., analog and digital mobile phones,
base stations, GSM, UMTS and LTE) and medical/biological effects. All identified publications were
analyzed by year of publication, research topics, authors and funding. Journals were ranked by the
number of articles. Origin of publications (i.e., country of affiliation address of the first author) was
visualized by density equalizing mapping.

769
Results and Discussion
We identified 1,415 scientific peer-reviewed publications on exposure to mobile communication
devices and medical/biological effects published from 1992 to 2017. After a steadily slow increase
up to 2006, the annual number of publications was about 92 publications per year (range: 81–102)
in the past decade (Fig. 2). We divided the publications into two nearly equally sized groups up to
the year 2009 (n=667) and since 2010 (n= 748) for further analysis of temporal trends.
A total of 458 journals were identified in which the articles were published. The journal
‘Bioelectromagnetics’ led the field (178 publications, 12.6%), followed by ‘Radiation Research’,
‘International Journal of Radiation Biology’ and ‘Electromagnetic Biology and Medicine’ (85, 75 and
69 publications, respectively; Fig. 3). About one third of the articles were concentrated in only five
journals whereas 292 (63.6%) of the journals published only a single article once.
Further results will be presented and discussed at the meeting.
References
[1] International Telecommunication Union (2018). ICT Statistics – Global and regional data.
https://www.itu.int/en/ITU-D/Statistics/Pages/stat/default.aspx (Assessed 20.02.2018)
[2] International Agency for Research on Cancer. Non-Ionizing Radiation, Part 2: Radiofrequency
Electromagnetic Fields. IARC Monographs on the Evaluation of Carcinogenic Risks to Humans, Vol.
102. Lyon, France: WHO, 2013. http://monographs.iarc.fr/ENG/Monographs/vol102/mono102.pdf
[3] Driessen S, Dechent D, Gollnick F, Gräfrath D, Schmid G, Cecil S, Hirtl R, Schneeweiß P,
Klemcke S, Janke G, Trübswetter A (2017). Evaluierung des EMF-Portals und Ableitung von
Erkenntnissen und Empfehlungen für dessen weitere Gestaltung. Abschlussbericht zum Vorhaben
FM 8857. Bundesamt für Strahlenschutz (Hrsg.). BfS-RESFOR-124/17,
URN:nbn:de:0221-2017071314292

Figures

770
Figure 1. Homepage of the EMF-Portal

Figure 2. Number of publications and journals per year between 1992 and 2017. (Green –
number of publications, blue – number of journals).

771
 Figure 3. Top ten journals in relation to the number of publications by time periods (green
1992–2009, blue 2010–2017).

PB-140 [14:30]
Scientific knowledge changes people's RF-EMF risk perception
Alberto Nájera1 & Judith Maciá1
1Medical Sciences Department, University of Castilla-La Mancha, Albacete, Spain
Keywords: Public Health Policy, Static, Completed (published)
Presented by: Alberto Nájera
We analyzed if providing scientific information about personal exposure to RF-EMF and possible
health effects prior to the placement of a new mobile base station, could minimize citizen
mobilizations.
With the development of the information society and the new communication technologies, human
exposure to RF-EMF has increased. Despite the available scientific evidence, the concern about its
possible effect on human health has also been increasing.
Faced with this situation of concern and uncertainty about the potential effects on health, the World
Health Organization (WHO) in 2005 showed that it is not only the lack of knowledge about the
consequences on health that generates this uncertainty but, the differences in personal risk
perception also influence this positioning.
The main objective of the study was to evaluate, on the one hand, the impact of the source from

772
which the information comes; and on the other hand, how the lack of communication of scientific
evidence, leads to generate a situation of unrest and unfounded fears.
For this evaluation, two surveys were carried out, before and after, to assistants (Ne = 16), people
living in a neighbour in which there was a movement against the installation of a mobile phone base
station, to a conference in which the results of personal exposure measures in 10 homes before and
after the start-up of the base station were communicated. The content of the talk included:
information about RF-EMF, electromagnetic spectrum, intensity, propagation and interaction with
living beings, typical levels of RF-EMF exposure, emission of radiation by mobile phone antennas,
distance effect, legislation, etc. A control group was taken into account, a random sample (Nc = 53)
to which the same survey was carried out as to the study participants but who did not attend it. In
the survey, age, training level, subjective level of exposure to RF-EMF (little, very, very much) were
collected, if the exposure could be associated with symptoms and diseases, a possible effect of
distance to the antenna, sources of information, the reliability of the information obtained according
to the source.
The answers were analysed through SPSS (version 15), through chi-2 tests and McNemar tests
before and after the talk. The small size of the sample prevents analysis and we did not find
statistically significant differences between groups. However, the subjective perception of personal
exposure offered, on a scale of 1 to 3 (Figure 1), with 3 being the maximum value, similar results
before and control, but with a very different pattern with the experimental group that reduced that
perception very significantly. On the perception of risk related to symptoms and exposure, the
attendees changed their thinking before and after the conference (Figure 2). The reliability of the
scientific information through the Internet was also evaluated, observing a change in the pattern
between before and after (Figure 3).
In conclusion, providing scientific information could minimize citizen mobilizations against installation
of mobile phone base stations.
Figures

Figure 1. Subjective perception of personal exposure level

773
Figure 2. ¿Exposure and symptoms? After,

Figure 3. Is the Internet a reliable source of scientific information?

774
PB-142 [14:30]
New technologies of the smart cities and public exposure of extremely low frequency
electromagnetic fields
Rauno Pääkkönen1 & Leena Korpinen2
1TMI Rauno Pääkkönen , Tampere, Finland
2Clinical Physiology and Neurophysiology Unit , The North Karelia Central Hospital , Joensuu, Finland
Keywords: Public Health Policy, ELF/LF, Completed (unpublished)
Presented by: Leena Korpinen
New technology brings more and more extremely low frequency electromagnetic fields around us.
These fields can be more complex than before; therefore, exposure to these fields must be
evaluated. We have performed this work for example with smart electric meters, electric cars, or
work related electromagnetic exposures. The fields have generally been less than 10% of the
recommendation values.
I. Introduction
A smart city concept is an urban development vision to integrate multiple information and
communication technology (ICT) solutions in a secure fashion like local departments information
systems, schools, libraries, transportation systems, hospitals, power plants, water supply networks,
waste management, law enforcement, and other community services. The wireless technology,
Internet including equipment, and new technologies of smart cities are new technical trends in
Europe. In addition, the use of electric cars has increased, and it is possible include intelligence, for
example, in electric storages and as part of smart homes and smart buildings. At the same time,
part of the European population is concerned that the exposure to electromagnetic fields will
increase and affect human beings. Some persons have described symptoms that they associate
with use of electrical equipment; for example, cell phones or laptops. In the media, there are
sometimes discussions about sensitivity to electromagnetic fields, for example, near cell phone base
stations [1, 2].
The aim of this paper is to collect the EMF exposure data from our earlier measurements and
analyze which EMF sources are important for smart cities. Our focus is on the extremely low
frequency-magnetic fields (ELF-EMF).
II. The smart grid and smart cities
A smart city includes typically, e.g., smart buildings (heating, mobility, and energy management),
transport (traffic), and renewables. The idea is the improvement of cities' overall energy efficiency.
Figure 1 shows the components of the smart grid. [3] The electric cars are also part of the smart
cities and smart electric grid.
III. EMF sources at smart cities
The smart cities include different electric and magnetic field sources of city systems like electric and
intelligent grids. For example, in an ELF area, there are smart meters, electric cars, and the
charging technology of EVs (electric vehicles). In addition, all cities also include electrical cables,
equipment, and illumination systems. The smart homes include typically wireless communication
technologies, but those sources are not in the ELF area. The following topics are examples of the
different sources that have been measured in Finland.
A. Smart meters
Finnish utilities use smart meters, which can send information or communicate using either a
50–100 kHz electromagnetic field signal through cables or a radio frequency (800–3000 MHz) aerial
signal to masts. The duration of communication of smart meters is typically from a few seconds to
tens of seconds each time. In our earlier study [4], we employed 46 smart meters that used power
line communication (PLC). The highest magnetic field was 0.48 μT.

775
B. Occupational fields
Examples of magnetic flux densities found at Finnish work places. Instantaneous values can be
high, but when average exposures are considered, then the values are typically low, for example
less than 5 μT.
IV. Discussion
Generally, the measured field values are less than recommendation values (<10%). Although known
fields are considered to be low, there are lots of concerns about possible health risks; hence, these
evaluations must be continued because while EMFs are becoming more complex, there is a
possibility that some kind of new adverse effects may appear directly or indirectly.
V. Conclusion
Thus far, we have not noticed any special problems in electromagnetic environment of smart cities,
but we have to gather data in order to respond to health concerns. Moreover, with data, it is possible
to shield sensitive equipment and reduce possible health risks.
VI. References
[1] L. Korpinen, R. Pääkkönen, “Self-report of physical symptoms associated with using mobile
phones and other electrical devices,” Bioelectromagnetics, vol. 30, pp. 431-437, September 2009.
[2] World Health Organization, 2005. Electromagnetic fields and public health: electromagnetic
hypersensitivity. Read 18.4.2016 http://www.who.int/peh-emf/publications/facts/fs296/en/.
[3] H. Rahimi-Eichi, U. Ojha, F. Baronti, and M.-Y. Chow M-Y, “Battery management system, an
overview of its application in the smart grid and electric vehicles,” IEEE industrial electronics
magazine, June 2013, 14-16.
[4] R. Pääkkönen, L. Korpinen L, Emission of smart meter electromagnetic field (50-100 kHz) in
Finland. CIRED 2015. 23rd International Conference on Electricity Distribution. Lyon, France 15-18
June 2015, paper 1484, 5 p.
Figures

Figure 1. Components of the smart grid. [3]

776
PB-144 [14:30]
STUDENT PAPER
Design of a triple band folded dipole antenna with low Specific Absorption Rate (SAR) for
LTE/WLAN applications
Ji Woong Park1, Min-Joo Jeong1, Niamat Hussain1, Hanul Bong1, Seungwoo Lee2 & Nam Kim1
1Chungbuk National University, Cheongju-si, Korea, 28644
2Korea Electric Power Research Institute, Cheongju-si, Korea, 58217
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Ji Woong Park
This paper presents a triple band folded dipole antenna with a reflector for 1.8 GHz LTE and 2.4/5.8
GHz WLAN applications. The antenna is composed of a monopole, via, two folded dipoles, and a slit
etched on the folded dipole. To obtain the optimized parameters, we used the simulator, Sim4Life
and the antenna shows an acceptable radiation pattern along with -10 dB return loss at all three
bands. The maximum value of the SAR1g (1 g-averaged SAR) for 1.8 GHz, 2.4GHz, and 5.8 GHz
band was 0.076 W/kg, 0.071 W/kg, and 0.0381 W/kg, respectively. These values are much lower
than the international standard values of SAR1g (1.6 W/kg), even though the distance between
antenna and head is zero.
Introduction
The development of wireless communication technology has been broadened and multi-banded with
the frequency band of operation frequency, which is multi and broadband communication capable of
voice and multimedia transmission and reception. Recently, researches on antenna design and
manufacture for wireless communication have been actively conducted. Therefore, the
measurement of electromagnetic waves can adversely affect the human body and the SAR of
human head is restricted from exceeding the standard in each country. In this paper, we propose a
folded dipole antenna with triple band characteristics that can be used in 1.8 GHz of LTE and 2.4/5.8
GHz band of WLAN, and analyzed the SAR. In addition, SAR was reduced by using reflector on the
antenna.
Method
A design of an antenna radiation characteristics and SAR were analyzed by Sim4Life. The antenna
has modified planar monopole antenna connected to the feed line, patch with two planar folded
dipole antennas on the back, and via that connects the monopole and the folded dipole. Also, the
proposed antenna in order that reduced the SAR value has a 65mm(≈ λ / 2) distance between the
antenna and the reflector. Since the international standard has been measured at less than 10 mm,
we simulated it with the most stringent criteria by attaching an antenna to the head at input power of
1 W. In the simulation method, SAR1g was analyzed for 1.8 GHz, 2.4 GHz and 5.8 GHz by bringing
the head and the antenna into contact with each other, and SAR was compared with the presence
or absence of the reflector in Figure. 1.
Results
The antenna was designed on a 1.6 mm thick FR4-PCB (relative dielectric constant = 4.3, loss
tangent = 0.025) has total size of . The optimal distance between the antenna and the reflector is 65
mm (≈ λ / 2) to reduce the SAR value. The total size of reflector was . The antenna shows an
acceptable radiation pattern along with -10 dB return loss at all three bands. The return loss with
and without reflector has been showed in Figure 1, while the far field radiations are shown in Figure
2. The calculation result shows that SAR1g value with reflector and SAR1g value without reflector in
Figure 3.
Conclusion
We propose a triple band folded dipole antenna for LTE/WLAN and compared the SAR1g values

777
according to international standards with and without reflector at all three bands. An antenna with a
reflector has a lower SAR1g value than an antenna without a reflector. Due to the shielding by
reflector, the reduction in SAR is achieved about 99 % at all three bands, compared to the
international standard value of SAR1g (1.6 W/kg).
Figures

Figure 1. Simulated results of return loss

Figure 2. Simulated results of radiation pattern

778
 Figure 3. SAR1g distribution at each band in human head model

PB-146 [14:30]
A preliminary presentation of a national EMF exposure survey program concerning sensitive
land use
Charilaos Tyrakis1, G.A. Gourzoulidis1, 2, A. Alexias1, G. Manios1, 3, V. Tsitsia1, Y. Kiouvrekis4, 6 & C
Kappas5
1Medical Physics Lab, University of Thessaly, Larissa, Greece, 41110
2Research & Measurements Center of OHS Hazardous Agents, Ministry of Labor, Athens, Greece, 10110
3Nuclear Medicine Department, Evangelismos General Hospital, Athens, Greece, 10676
4Mathematics Department, National Technical University of Athens, Athens, Greece, 15780
5Medical Physics , University of Thessaly, Larissa, Greece, 41110
6Institute for The Future , University of Nicosia , Nicosia, Cyprus, 24005 CY-1700

779
Keywords: Public Health Policy, RF/Microwaves, Work in Progress
Presented by: Charilaos Tyrakis
The degree of social concern regarding EMF exposure is considerable in Greece, while the entire
protection framework, both for the general public and for the workers is set on a really sound basis,
internationally speaking; moreover, the balance of the scientific evidence further supports it. What
seems to be missing is the communication of the appropriate information as well as the clarification
of the misinformation, especially as far as the EMF exposure conditions of kids and/or people with
sensitivities (pregnant women, people with chronic diseases, elderly) are concerned. Consequently,
a national EMF exposure survey has been proposed and approved to highlight the field; the basic
parameters of this program are provided.
INTRODUCTION
EMF exposure assessment and adverse health effects remains as a controversy issue for many
decades. The rationale behind the International Commission’s on Non-Ionizing Radiation Protection
(ICNIRP) approach to this issue, which is the scientific basis for the legislative protection framework,
is to apply higher safety factors for the general public (g.p.) than for workers, in order to compensate
for: i) severe environmental conditions (i.e. high temperature) and/or high activity levels; ii) a
potentially higher thermal sensitivity in certain population groups, such as the frail and/or the elderly,
infants and young children, and people with diseases or taking medications that compromise
thermal tolerance [1].Nevertheless, the social ‘anxiety’ about EMF exposure increases when it
comes to children and patients. Moreover, the Hellenic legislation has ‘arbitrarily’ incorporated this
‘anxiety’ through reduced RF limits (60 and 70% of the EU) [2] while it has not incorporated the 50
Hz limit increase (from 100 μΤ to 200 μΤ) implied by ICNIRP 2010 [3]. In other words, the social
concerns are somehow strengthened by the local policies and need to promote the scientific
evidence is urgent.
Fear-mongering
Many attempts have been recorded to spread fear when speaking about EMFs. Severe bullying and
violence incidents have been recorded against landowners (where base stations are installed),
especially in areas close to kindergartens and schools. Additionally, websites host warnings that
EMFs could be responsible for: i) cancer [4]; ii) leukemia and AIDS [5]; iii) suicide prompting; iv)
coagulation of red blood cells; v) melting of the mercury of dental fillings; vi) serious sleep
disturbances; vii) reduction of the bees’ populations [6]; viii) lighting attraction [7]; and even ix)
popcorn production and egg baking when radiated by mobile phones.In many cases, the above-
mentioned attempts are combined with the promotion (online shopping) of ‘EMF-protection
equipment’.
METHODS
The National EMF Survey Program EDBM34
Τhe Νational Program led by the University of Thessaly evaluates the EMF environment especially
the ELF & RF exposure in areas of sensitive land use (schools, hospitals, Kindergardens, elderly
houses). The program’s objective is to clarify this field; coded as EDBM34 under the
OPERATIONAL PROGRAM ‘DEVELOPMENT OF HUMAN RESOURCES, EDUCATION AND
LIFELONG LEARNING’ of the Ministry of Economics and Development and the European Social
Fund.
The parameters of the program
EMF measurements have been designed according to table 1. The main idea, behind highlighting
the EMF exposure, is to focus on the main sources of concern, such as mobile base stations and
high voltage pylons; Sensitive land use is identified in the vicinity of such EMF sources. According to
the Greek legislation, there is a reduction of 40% in the General Public EMF limits at a distance
range of 300m around Mobile Base Stations. In this sense, high frequency measurements have
been carried out in a range of 300m radius from mobile base station, especially in areas of sensitive
land use, and in a range of 200m radius from high-voltage pylons for low frequency measurements.
780
The main parts/innovations of the survey include:
Concise recording, for the first time in Greece, of the radiation exposure of sensitive parts of the
population, both from high and low EMF frequencies.
Standardization of the measurements, following the proper approach in order to ensure the
repeatability and the accuracy of the results (use of dedicated and calibrated equipment, concise
measurements according to standards, appropriate data and statistical processing).
Epidemiological data processing, correlating the radiation exposure to the social stratification of
the population.
Guidelines provisioning, concerning sensitive population protection, based directly on dedicated
measurements.
In addition, the main contribution is:
The social and epidemiological concern, since the knowledge gap, the misinformation and the
psychological burden of the general public, rather than the EMF exposure itself, are considered as
the core issues. In this sense, the real radiation exposure of kindergardens and schools operationg
close to mobile base stations installations and high voltage pylons, will be clarified, resulting in the:
a)abolition of the EMF exaggerations and the related fear-mongering and b)enhancement of the
general public’s sense of security, enabling them to trust scientists that are entitled to monitor EMFs.
Table1. Measurements sites of sensitive land use, for high and low frequencies; distribution
throughout the country

Peloponnese
Macedonia- Thessaly-
Sites Attica Thessaloniki Central Crete Total
Thrace Epirus
Greece

kindergartens
70 40 20 30 20 20 200
*

Schools * 70 40 20 30 20 20 200

High
voltage 50 30 20 30 20 20 170
pylons**

Mobile
base 50 30 20 30 20 20 170
stations

* Close to mobile base stations (300m range) ** Located closer than 200m
RESULTS
The first RF results indicate (Figures 1-2):
Very low field levels, most of them well below 1 V/m (broadband value).
Considerable EMF contribution from Radio (FM)/Television sources.
Many ‘hidden & unknown’ to the population EMF sources like meteorological and defense
satellite signals.
The distribution of a sample result concerning a hospital case exhibits a total RF exposure ratio of
0.00018 (total broadband field around 0.9 V/m) (Figure 1). The type B expanded (k=2) uncertainty
was calculated around 21% and the measurements were conducted at three heights (1.10, 1.50 and
781
1.70m from the ground).As a result, the spectrum analysis shows that the highest RF exposure level
measured in the hospital area is 5,555 times lower than the Greek public exposure limits that have
adopted a safety factor of 60% compared to the EU/ICNIRP Guidelines.The distribution of a sample
result from a school case in the area of Larissa (Central Greece) exhibits a total RF exposure ratio
of 0.00082 (total broadband field around 0.92 V/m) (Figure 2).As a result, the spectrum analysis
shows that the highest RF exposure level measured in the area of the school is 1,220 times lower
than the Greek public exposure limits.
DISCUSSION
The use of the words ‘radiation’ for the whole electromagnetic spectrum seems to contribute to the
‘EMF high concern’ as this is being confused with the ionizing regions and effects (radioactivity). In
this sense the communication of the exposure results is the most challenging part of the project, a
fact highlighted also from the relevant occupational survey [8]. Moreover the use of different (local)
limits, which are many orders of magnitude below the official limits [9], proposed by certain ‘EMF
safety companies and/or organizations’, not only gives rise to public concern strengthening the
general anxiety but also degrades the trust to science and scientists. This mix up of ‘scientific’ and
‘non-scientific’ approaches results in many people being afraid of EMFs and having a very limited
view of the EMF sources. The mobile phone base stations are on the top of the list of ‘scary’
sources and many studies reflect the fact that the exposure perception is the key indicator of the risk
perception, as most of the individuals questioned seem to believe that a safety distance from a base
station must be thousands of meters away from them [10]!
Nevertheless, as far as ‘EMF-proofs ‘are concerned, it should be stated that:
In principal, science cannot prove the non-existence of an effect. Actually, every epidemiological
survey concerning EMF effects has to be first checked against the null hypothesis [11] i.e. the
hypothesis that an EMF exposure doesn’t cause adverse health effects; this part is almost
always missing. This leads to a critical issue in risk communication: The chances for giving a
sound and undisputable message in risk communications are asymmetrically distributed
between the risk and the no-risk-messages; in principle, warning people can be better
substantiated than reassuring people. Unfortunately, this corresponds to a preference of human
nature for negative information, as negative information weights stronger in human information
processing than positive information [12].
There are many additive factors on health status; In any case, this is the main criticism
concerning epidemiological survey as far as adverse EMF effects are concerned.
Last but not least, there is a great concern about different EMF limits adopted in different countries.
Local safety factors or complete differentiation from WHO & ICNIRP Guidelines may result in an
increase of the public concern and a severe degradation of the worldwide scientific work. In fact
these different policies lead to an untrustworthy environment which prevents technology
improvement and society wellbeing. Need for a worldwide limits harmonisation is essential.
CONCLUSION
The first results prove that the general public exposure is far below the WHO and ICNIRP's limits
(1998, 2010 guidelines) and also below the Greek limits concerning areas/land of sensitive use
(2006, 2014). These reliable quantitative results, even if limits might be reviewed in the future, set a
sound basis for the real exposure conditions, highlighting a huge safety margin.The communication
of the non-ionizing radiation issues is the most challenging part of the survey and a lot of effort
should be directed towards this.
REFERENCES
[1] ICNIRP Guidelines (up to 300 GHz), Health Phys 74(4):494-522;1998.
[2] Hellenic Presidential Decree (FEK) 346/B/3.3. 2008.
[3] ICNIRP Guidelines for limiting exposure to time-varying electric and magnetic (1 Hz - 100 kHz).

782
Health Physics 2010;9(6):19.
[4] Khurana G. ‘Mobile Phones and Brain Tumours – A Public Health Concern’, www.brain-
surgery.us, 2008.
[5] http://nhsblogdoc.blogspot.com/2008/06/mobile-phones-cause-hiv-aids.html
[6] Kirschvink JL et al., Measurement of the threshold sensitivity of honeybees to weak, extremely
low-frequency magnetic fields, J of Experim Biology, V.200(9), pp.1363–1368, 1997.
[7] www.mmfai.org/public/docs/eng/MMF_Viewpoint_Lightning.pdf
[8] Gourzoulidis GA et al. Occupational exposure to electromagnetic fields, the situation in Greece.
Phys Medica 2016;32(276).
[9] Directive 2013/35/EU OJ L 179/1.
[10] Freudenstein F et al. Exposure Perception as a Key Indicator of Risk Perception and
Acceptance of Sources of Radio Frequency Electromagnetic Fields. J of Environmental and Public
Health, http://dx.doi.org/10.1155/2015/198272.
[11] SCENIHR opinion, http://ec.europa.eu/health/scientific_committees/emerging
/docs/scenihr_o_041.pdf
[12] Ito TA et al. Negative information weights more heavily on the brain: the negativity bias in
evaluative categorizations. J Pers Soc Psychol. 1998;75:887–900.
Figures

Figure 1. Sample results from a hospital (MT stands for mobile telephone).

783
 Figure 2. Sample results from a school.

PB-148 [14:30]
North Atlantic Treaty Organization (NATO) Science and Technology Organization (STO)
Research Task Groups (RTG) 189 and 298. Creating Basis for Standards Paradigm Shift New
Radio Frequency Exposure Standards
B. Jon Klauenberg1, 2, Jeffrey Whitmore1, 2, Roel Escobar1, 2, Jason Payne1 & Bennett Ibey1
1Directed Energy Bioeffects Division, United States Air Force Research Laboratory, San Antonio, Texas,
USA, 78234
2Collaboration Support Office, North Atlantic Treaty Organization STO Research Task Group, Paris, France,
75008
Keywords: Standards, RF/Microwaves, Completed (published)
Presented by: B. Jon Klauenberg
Military electromagnetic (EM)-based systems must be continually evaluated for establishing safe
operational exposure limits. Standardization of safety guidance for exposure of personnel is required
for interoperability during multi-nation actions including training, demonstrations, exercises or
combat engagements. NATO STO RTG-189 reached consensus to eliminate high peak power ultra-
short pulsed EM field limits. RTG 298 initiated literature review to determine the conditions under
which tissue damage thresholds can be incorporated into safety standards. We will review the 2018
RTG-189 STO Technical report “Electromagnetic Fields Exposure Limits” and describe the RTG-298
planned processes for establishing the tissue damage thresholds.
Introduction
Commonality of standards is required for NATO interoperability. NATO Nations are developing new
capabilities of extremely high-peak-power ultra-short electromagnetic pulses. Coordination of
international laboratory expertise ensures NATO Nations are leveraging research efforts and will

784
expedite establishing appropriate safety guidance. As Custodian of STANAG 2345 [1], USA has led
three revisions of that personnel safety standard over a 15-year period and recognizes the
importance of the STANAG to continued interoperability.
Five NATO nations (France, Hungary, Netherlands, Sweden, Turkey, United Kingdom, United
States) participated in a three year Research Technology Organization, Human Factors and
Medicine (HFM) Research Task Group (RTG) review of the emerging military technologies that
employ High-Peak-Power ultra-short Pulsed Electromagnetic Fields (HPPP-EMFs) ranging from
milliseconds (e.g. radars) to micro- and nano-seconds (e.g. high-power microwave and directed-
energy devices) [2].
The EMP generated by a high-altitude nuclear detonation that can cover a continent is termed High-
altitude EMP (HEMP) [3]. The immense amount of energy liberated by a nuclear explosion changes
the surrounding environment through conversion of the weapon gamma-ray energy in the earth’s
ionosphere to RF electromagnetic energy which propagates toward the earth’s surface [4]. These
EMPs can extend well beyond the thermal or ionizing radiation components, especially for HEMPs
and represent a threat to electronics and communications. HEMP has been of scientific interest
since 1989 [5]. Electromagnetic emitters capable of delivering 100 Gigawatt (GW) pulses to a
transmitting antenna producing hundreds of kV/m peak electric (E) fields had been developed by the
1990’s [6]. As early as the 1960s EMP Bioeffects had the attention of government and industry. The
United States Occupational Safety and Health Act [7], employer concern for worker safety, and the
rapid development of EMP simulators in the 1960’s and 1970’s led to efforts to establish worker
health and safety EMP exposure standards. Experiments were being conducted by the United
States military on the effects of EMP that would occur with nuclear events.The effects they were
looking at were those on electrical and electronic materiel. The ultra-short HPPP-EMF limits that
were inserted into standards over 40 years ago were set because of concerns that were not
scientific in nature, but rather overly cautious concerns of program managers, policy makers and
some scientists who believed any limit was better than none.
The original goal for HPPP-EMF emission standardization was to establish safe exposure levels for
personnel working with EMP simulators. The hypothetical concern was that the extremely high peak
power may interact with personnel in a manner differing from the established thermal-biological
basis. Studies carried out on both animals and humans for over 40 years failed to provide support
for this hypothesis to limit HPPP-EMPs, yet the limits remained.
The first guidance for limiting exposure to EMP “Provisional Safety Criteria for Use in
Electromagnetic Pulse (EMP)” was published by the Radiobiology Division United States Air Force
(USAF) School of Aerospace Medicine (USAFSAM) in 1971 [8]. As a highly conservative measure,
a “provisional” arbitrary limit of 100 kV/m was established. The “provisional” interim standard stated:
“No hazardous effects of fields now in use in EMP simulators have been documented or calculated.
Pulsed microwave exposures four orders of magnitude greater than those being used in current
EMP simulators have given no demonstrable ill effects to several species of animals including man.
The recommended standards should be conservative at least one order of magnitude, but are
conservative by three orders of magnitude based on experimental data.”
This “provisional” EMP safe-tolerance limit was based on acute thermal burden and related to the 10
mW/cm2 (100 joules meter -2 seconds -1) that would later become the 1974 ANSI standard [9]. The
limit was reissued in 1975 [9] and in 1987 [10]. The limit of 100 kV/m fit well with the measured fields
occurring due to nuclear blast which rarely reach 100 kV/m [11].
Analysis/Results
In summary, no noticeable adverse health effects following exposure to HPPP-EMPs were reported
for the reviewed USAF (Fig. 1), Boeing and Bell Laboratories studies which included human studies
with years of medical follow-up. HFM RTG-189 found no published and replicated adverse health
effects or biological mechanisms, beyond thermal interaction, for pulses shorter than 100 ms which
suggested that neither the peak E-field limit in the IEEE C95.1TM-2005 safety standard nor the
proposed limit in the Directive 2004/40/EC [12] (subsequently promulgated as 2013/35/EU [13])

785
have scientific basis. The group noted that physical laws governing the propagation of E-fields in air
already limit the maximum allowable peak E-field at ~3 MV/m (air breakdown). Current research
efforts by members to expose biological organism(s), tissues, and cells to environmental fields up to
this magnitude have been unable to elicit an acute biological response.
Conclusions
The 100 kV/m limit was not based on science, but was basically a policy decision. Today we know
that setting exposure limits without sufficient data, a precaution against imaginable effects, can be
problematic later as the data-base is expanded demonstrating no need for such caution.
Unfortunately, there is resistance by some to remove any limits once established even in the face of
50 years of negative studies. “The peak field standard should be removed as it has no historical
precedence, no published support, and also little to no impact on safety as the thermal standard is
already in place and well supported” [14]. A consensus statement was drafted which recommended
that the exposure limitation based on HPPP E-field be eliminated. This consensus recommendation
was developed by the NATO Research Task Group HFM-189 consisting of 17 subject-matter
experts from seven countries. The recommendation of the HFM-189 was adopted by the IEEE
International Committee on Electromagnetic Safety (ICES) TC-95 in IEEE Standard C95.1-2345TM,
“Military Workplaces-Force Health Protection Regarding Personnel Exposure to Electric, Magnetic,
and Electromagnetic Fields, 0 Hz to 300 GHz”. The standard has been adopted by NATO under
STANAG 2345 Edition 4. [1]
Injury Thresholds of High Power Radiofrequency Emissions (NATO STO HFM RTG-298)
Introduction
Establishing validated exposure limits for emerging EM-based systems will enable operational
implementation of systems emitting directed energy (DE) with novel power, pulse characteristics,
waveforms and other physical properties that are not considered in civilian safety standards.
Specific characteristics of the EM spectrum have been exploited in development of directed energy
weapons (DEW). The RTG-298 will focus on garnering sufficient scientific bioeffects data to support
development of safe exposure limits for personnel in high energy EM environments and target,
operator and collateral DEW. The RTG will strive to elucidate the roles of exposure parameters such
as power, pulse characteristics, frequency, modulation, polarization, duration, and intermittence of
exposure to ensure safe operations and access to RF environments that may exceed current
exposure limits. Importantly, the threshold at which damage occurs within the human body is not
well documented. Data on actual damage thresholds will enhance accuracy of SOH standards and
allow commanders to balance mission needs against the probabilistic risk of injury. Commanders
need to know the consequences of exposure above the present behaviorally-based conservative
level. The potential for shock and burns has been the number one concern of the IEEE International
Committee on Electromagnetic Safety (ICES) and must be further elucidated.
SCIENTIFIC OBJECTIVE(S) AND EXPECTED ACHIEVEMENTS
Coordinate and leverage multinational trans-Atlantic scientific knowledge for the identification of
the potential for injury to military personnel exposed to novel emerging military technologies
employing RF.
Identify gaps through a review of published data and computer modeling that will provide
direction for future research.
Develop guidance for preventing injury from emerging technologies.
SCIENTIFIC TOPICS TO BE COVERED
Topic 1. Injury thresholds for exposure to high power RF emissions
Topic 2. Computer-based modeling of RF-induced injury
S&T ADVANCES EXPECTED
Establish military relevant thresholds and response curves for injury that will enable operational
decision making about potential exposure. This activity supports multinational interoperability by
maximizing operational capability while minimizing injury, thus optimizing human protection,

786
sustainability and survivability in RF environments.
REFERENCES
1. NATO STANAG 2345 MEDSTD (Edition 4) – Military Workplaces – Force Health Protection
Regarding Personnel Exposure to Electric, Magnetic, and Electromagnetic Fields, 0 Hz to 300 GHz,
2015.
2. Electromagnetic Fields Exposure Limits. NATO STO Technical Report TR-HFM-189,
AC/323(HFM-189)TP/716, ISBN 978-92-837-2041-6, 2017.
3. Baum, C., “From the Electromagnetic Pulse to High-Power Electromagnetics”, Proceedings of the
IEEE, 80, pp. 789-817, June 1992.
4. Patrick, E.L. and Vault, W.L., “Bioelectromagnetic Effects of the Electromagnetic Pulse (EMP)”,
AD-A221 552, HDL-TL-90-2, March 1990.
5. Vittitoe, C., “Did High-Altitude EMP Cause Hawaiian Streetlight Incident”, System Design and
Assessment Note 31 June 1989 reported in Baum, C.E. From the Electromagnetic Pulse to High-
Power Electromagnetics, Proceedings of the IEEE, Vol. 80, pp. 789-817, June 1992.
6. Agee, F.J., et al. “Ultra-Wideband Transmitter Research”, IEEE Transactions on Plasma Science
26:860-873, 1999.
7. OSHA, John Stender, Assistant Secretary of Labor [Docket No. S-101] Exposure to
Electromagnetic Pulses. Determination Not To Propose Standard. FR Doc 75-14995 filed 6-6-1975.
8. USAFSAM, Radiobiology Division, “Provisional Safety Criteria for Use in Electromagnetic Pulse
(EMP)”, 15 December 1971.
9. ANSI, “Safety Level of Electromagnetic Radiation with Respect to Personnel”, ANSI C95.1-1974.
10. USAF, AFOSH REGULATION 161-42, “Radiofrequency Radiation Health Hazards Control”,
Washington, DC, USA, 7 November 1975. Department of the Air Force, AFOSH Standard 161-9,
“Exposure to Radiofrequency Radiation”, Washington, DC, USA, February 1987.
11. Foster, Jr., et al., “Report of the Commission to Assess the Threat to the United States from
Electromagnetic Pulse (EMP) Attack”, National Research Council, Washington, DC, USA,
Committee on Electromagnetic Pulse Environment, 2004.
12. Directive 2004/40/EC of the European Parliament and of the Council of 29 April 2004 on the
minimum health and safety requirements regarding the exposure of workers to the risks arising from
physical agents (electromagnetic fields) (18th individual Directive within the meaning of Article 16(1)
of Directive 89/391/EEC).
13. Directive 2013/35/EU of the European Parliament and of the Council of 26 June 2013 on the
minimum health and safety requirements regarding the exposure of workers to the risks arising from
physical agents (electromagnetic fields)
14. Ibey, B.L., E-mail to RTG-189, 6 December 2012.
15. Huber, P., “Galileo’s Revenge: Junk Science in the Courtroom”, Basic Books, Harper Collins,
New York, NY, USA, Basic Books, 1991.
Figures

787
Figure 1. B-52 Stratofortress on Atlas-1 “Trestle” Kirtland AFB World’s Largest EMP Simulator

Figure 2. Determine Spread between Radiofrequency (RF) Sensation and Tissue Damage

788
Figure 3. Modeled Injury Thresholds

789
 Gold Sponsors

The ELES Company ensures the safe, reliable and uninterrupted transmission of electricity. ELES is the
guardian of Slovenia’s electric power transmission system, which is closely connected to the
transmission networks of neighbouring countries and integrated into the European energy system. Our
responsibility is that electricity is always at hand, ensuring the consumers can meet their needs and
wishes at any
me. ELES is the operator of the electric power transmission network of the Republic of
Slovenia. With a professional approach, know-how and advanced technology, ELES has been providing
safe, reliable and uninterrupted electric power transmission throughout Slovenia and across the borders
for 90 years. Thus, the company connects people and ensures quality of life. ELES endeavours to
strategically, responsibly and sustainably plan, construct and maintain Slovenia’s high-voltage
transmission network in three voltage levels: 400 kV, 220 kV and a part of 110 kV.

The Electric Power Research Ins
tute (EPRI) conducts research, development, and demonstra
on
projects for the benet of the public in the United States and interna
onally. As an independent,
nonprot organiza
on for public interest energy and environmental research, we focus on electricity
genera
on, delivery, and use in collabora
on with the electricity sector, its stakeholders and others to
enhance the quality of life by making electric power safe, reliable, aordable, and environmentally
responsible.

790
 Gold Sponsors

The MWF is an interna
onal associa
on of companies with an interest in mobile and wireless
communica
ons including the evolu
on to 5G and the Internet of Things.

We are a development-stage medical device company using a novel and proprietary plaorm technology
called Nano-Pulse S
mula
on or NPS. NPS is a local and drug-free technology that u
lizes ultra-short,
nanosecond pulsed electric elds to induce cell signaling and the ac
va
on of cellular pathways in

ssue. One of the more promising applica
ons of NPS is the treatment of solid tumors, where pre-
clinical data developed by Pulse Biosciences and others have shown that NPS provides eec
ve local
tumor control and ini
ates an adap
ve immune response with a vaccine-like eect by inducing
immunogenic apoptosis of the treated cells. We believe we will establish NPS as a new treatment
modality across a variety of applica
ons, including both veterinary and human oncology, dermatology,
aesthe
cs and other minimally invasive applica
ons where current abla
on modali
es do not provide
the benets of NPS. We are developing a proprietary system for the delivery of NPS treatments,
currently iden
ed as the “PulseTx.” The PulseTx system delivers NPS pulses through its proprietary
tunable pulse generator and its suite of electrodes. The PulseTx system pulses are applied directly to

ssue through electrodes, crea
ng transient nanometer pores in cell and organelle membranes. We
believe we are the only medical device company with the intellectual property, technology, and know-
how to be able to produce this natural cell death using NPS to ini
ate cell signaling that induces the
targeted adap
ve immune response.

791
 Silver Sponsors

Ortho
x Internaonal is a diversi
ed, global medical device company focused on improving paents'
lives by providing superior reconstrucve and regenerave orthopedic and spine soluons to physicians
worldwide. Headquartered in Lewisville, TX, the company has four strategic business units that include
BioSm, Biologics, Extremity Fixaon and Spine Fixaon. Our products are distributed in more than 50
countries around the world via Ortho
x sales representaves, stocking distributors and subsidiaries.
Founded in Verona, Italy in 1980, Ortho
x has approximately 900 employees around the world who are
dedicated to the development, manufacturing and distribuon of orthopedic and spine products and
regenerave ssue forms. Many of our innovave product oerings were culvated through aliaons
with surgeons and premiere health care organizaons. Ortho
x is proud to collaborate on research and
development acvies with leading clinical? organizaons such as the Musculoskeletal Transplant
Foundaon (MTF), the Orthopedic Research and Educaon Foundaon and the Texas Scosh Rite
Hospital for Children.

792
 Bronze Sponsors

A1 Slovenija is the leading private provider of integrated communica
on services in Slovenia. More than
500 employees are commied to crea
ng solu
ons that enrich life in the digital age to more than
700.000 users. A1 Slovenia develops meaningful communica
on solu
ons that full current and
emerging needs of users so they will experience work play, crea
ng, and sharing in a more connected
way. A1 Slovenija is in a 100-percent ownership of Telekom Austria Group (ATX:TKA) and is an important
part of the leading digital solu
ons and communica
ons provider in Central and Eastern Europe with
more than 24 million users in 8 countries, more than 4 billion EUR income and more than 17.500
employees. The Group is the European unit of América Móvil, the third largest provider of wireless
services in the world. A1 Slovenija began oering its services in March 1999 as the rst privately owned
mobile operator in Slovenia. It has been retaining a posi
on of a pioneer in the mobile communica
on
eld and constantly strengthening its posi
on in the integrated communica
ons market ever since. A1
Slovenija is using advanced technology to enable the users to shape the experience they want to live.

The GSMA represents the interests of mobile operators worldwide, uni
ng nearly 800 operators with
almost 300 companies in the broader mobile ecosystem, including handset and device makers, soware
companies, equipment providers and internet companies, as well as organisa
ons in adjacent industry
sectors. The GSMA also produces industry-leading events such as Mobile World Congress, Mobile World
Congress Shanghai, Mobile World Congress Americas and the Mobile 360 Series of conferences. For
more informa
on on GSMA ac
vi
es related to electromagne
c elds please visit
hp://www.gsma.com/publicpolicy/consumer-aairs/emf-and-health.

793
 Bronze Sponsors

Telstra is Australia’s leading telecommunica
ons and technology company, oering a full range of
communica
ons services and compe
ng in all telecommunica
ons markets. In Australia Telstra provide
17.6 million retail mobile services, 5.1 million retail xed voice services and 3.5 million retail xed
broadband services.

794
 Exhibitors

Narda is a leading provider of measuring equipment for the EMF Safety, RF Test & Measurement and
EMC sectors. With its standards-compliant EMF measurement solu
ons (for electrical and magne
c
elds) from 0 Hz to 90 GHz, it covers almost all mobile radio services - including 5G. The EMF Safety
product spectrum includes wideband and frequency selec
ve measuring instruments, full coverage wide
area monitors, and personal safety monitors that are worn on the person. The RF Test & Measurement
range includes analyzers and devices for measuring and iden
fying RF sources. The PMM brand of EMC
instruments consists of devices for measuring the electromagne
c compa
bility of equipment. Servicing
and calibra
on, together with training programs complete the range of services provided. The company
management system is ISO 9001:2015 cer
ed, and it operates a DIN EN ISO/IEC 17025:2005 accredited
calibra
on laboratory. Narda has development and produc
on facili
es at three loca
ons: Hauppauge,
Long Island/USA, Pfullingen/Germany and Cisano/Italy. It also has its own representa
ve in Beijing/
China. A global network of Sales Partners ensures close customer contact. Narda is part of L3
Technologies, New York.

Regional authorised distibutor

WiLLTRON Technologies GmbH

Am Kanal 8-10
2352 Gumpoldskirchen
j.saustingl@willtron.at
www.willtron.at

795
 Conference Supporters

Since 1980, the year the company was founded, IGEA has been working to provide innova
ve and
eec
ve therapies to pa
ents. IGEA's constant commitment and knowledge, built on solid scien
c
founda
ons, led to the birth of Clinical Biophysics, a branch of medicine based on the use of non-ionising
physical energy. Over the years, scien
c research and intellectual curiosity have always dis
nguished
the work of IGEA enabling the company to develop new and ecacious orthopaedic medical devices to
promote repara
ve osteogenesis and joint protec
on. In 2000, with the aim of oering the pa
ent an
eec
ve and safe therapy that is easy to use, with minimal side eects, IGEA decided to take on a new
challenge: the development of oncological biophysics for the treatment of solid tumours. The signicant
progress made in science, technology and con
nous iden
ca
on of innova
ve and ecacious
treatment solu
ons for the improvement of pa
ents’ quality of life, have made IGEA the leading
company in the eld of Clinical Biophysics.

It is ZonMw’s goal to ensure that healthy people stay that way for as long as possible, that ill people
recover as quickly and completely as possible and that people who require care and nursing receive the
highest standard of services. To achieve this, we need to focus on preven
on: on stopping people from
becoming ill. And we need good health care for people who nevertheless fall ill. The ques
on ZonMw
faces is how to improve disease preven
on and health care. One thing is clear: you need a lot of
knowledge, and therefore a lot of research. And it is important that people actually use that knowledge.
With this in mind, ZonMw funds and promotes research, development and implementa
on.

796
Abramson, Michael: S14-3, S05-3, PA-71 Beauvois, Veronique: PA-83, S10-3
Abundis-Gutierrez, Emanuel: S13-3 Bechet, Paul: PA-49
Achermann, Peter: S16-3 Benassi, Barbara: S03-3, S03-2
Addison, Darren: S11-3 Benke, Geza: PA-71, S14-3, S05-3
Aerts, Sam: S14-4, S14-6 Benke, Kurt: S05-3
Agneessens, Sam: PB-44 Benquet, Pascal: PA-75, FA-8
Ahlbom, Anders: S05-5 Bereta, Martin: PA-125
Ahn, Hyungsang: PA-47 Besnier, Philippe: S14-2
Ahn, Young Hwan: PA-111 Bhatnagar, Megha: S05-5
Aimoto, Atsuko: FB-3, PB-48 Bhatt, Chhavi: PA-71, S14-3
Akahane, Manabu: S04-2 Bienkowski, Pawel: PB-92, PB-34, PA-1, PA-53
Akutagawa, Masatake: FA-10, PA-99 Birks, Laura: S11-1
Al Hajj, Maarouf: S11-4 Bishop, Jonh: PA-61
Alekseev, Stanislav: PA-91, FA-9 Blystone, Chad: T3-1
Alexias, A.: PB-146 Bodemann, Ralf : S16-4
Alon, Leeor: PW1-7 Bodewein, Lambert: PA-131, PB-138
Aloui, Saifeddine: S14-5 Bogdanov, Todor: PA-117
Alpuente-Hermosilla, Jesus: PA-45 Boillée, Severine: S08-1
Alteköster, Carsten: S04-3 Bollati, Valentina: S03-2
Aminzadeh, Reza: PB-44, S14-2 Bonato, Marta: PB-82
Amoon, Aryana: S05-5, S05-6 Bonelli-Salvadori, Aurélie: PA-113
An, Ho Sung: PB-52 Bong, Hanul: PB-144, FB-14, FB-13, PA-133, PB-136,
An, Junoh: PA-47 S13-1
Anand, Snjezana: S12-1 Bordarie, Jimmy: S10-3
Andre, Franck: FA-7, PA-73 Bories, Serge: S14-5
Apollonio, Francesca: S15-5, W3-3, PB-56, W3-7, PA-63, Bornkessel, Christian: S11-2, S14-1
PA-119, W3-5, PB-72 Botha, Shirmone: PA-73, FA-7
Arah, Onyebuchi: S05-6 Bouisset, Nicolas: PB-84, FB-6, S10-2
Ardelean, Ioan: PA-49 Bouland, Catherine: PA-83
Ardelean, Ioan I.: PA-49 Bouldi, Mélina: S05-4
Arnaud-Cormos, Delia: S08-1, PB-128 Boutillier, Laurette: PA-113
Atanasov, Nikolai: PA-117 Bray, Issy: S05-5
Atanasova, Gabriela: PA-117 Brecelj, Erik: PB-74
Attal-Trétout, Brigitte: W3-7 Breda, Cestnik: PB-12
Augustsson, Torsten: S05-2 Breton, Marie: W3-6
Azan, Antoine: W3-8, W3-6 Brulais, Sébastien: S14-5
Azimov, Uktam: S13-1 Bruzon, Rodolfo: FB-11, S06-1, PB-118
Bachmann, Philipp: PA-55 Brzozek, Christopher: S14-3, S05-3, PA-71
Bae, Sanghyuk: PB-78 Bunch, Katherine: S05-5
Bailey, William: S16-4 Burdio, José-Miguel: FA-7, PA-73
Bajla, Ivan: PA-125 Burm, Eunae: PB-78
Baker, Cadence: S01-4 Bushberg, Jerrold: S16-4
Bakos, Jozsef: PA-89 Butera, Alessio: S03-3
Balzano, Quirino: PB-56, S13-2, PB-36 Bömmels, Ingo: S16-5
Banin, Alessandro: PA-63 Cadossi, Ruggero: PA-119, S15-5
Barbosa, Carlos: PA-35 Cala, Pawel: PB-92, PA-1
Barbu-Tudoran, Lucian: PA-49 Calderon, Carolina: S11-3
Bardet, Sylvia M.: PB-94, PB-120 Camera, Francesca: S15-5
Barnes, Frank: S06-1, FB-11, PB-118 Campi, Tommaso: PA-33
Barnes, Ronald: PB-32 Canova, Aldo: S15-2
Bartonek, Michael: PB-86, PB-98 Caorsi, Salvatore: PB-100
Batista Napotnik, Tina: S07-6, FA-6, PA-65 Capone, Fioravante: S15-5

797
Caprio, Emilio: S11-6 Cvetkovic, Mario: PB-30, FB-2
Capstick, Myles: PA-113 Dalecki, Anna: S01-2, S10-1
Cardis, Elisabeth: S11-1, PB-44 Daniels, Dianne: S07-2
Carr, Lynn: PB-94, PB-120 Danker-Hopfe, Heidi: S01-3
Carrasco, Eduardo: S13-2 Dassonville, David: S14-5
Carvallo, Andres: FA-8, PA-75 Davis, Ilan W.: PA-61
Casciati, Arianna: PA-61 De Geeter, Nele: S14-2
Casciola, Maura: PA-63, S12-2, W3-3 De Santis, Valerio: PA-33
Cassara, Antonino Mario: S09-4, S15-1, S09-3 De Seze, Rene: PB-44
Castellani, Federica: S12-4 Debaillie, Björn: S11-5
Ceballos Aguirre, Nelson: PB-110, FB-9 Debouzy, JClaude: PB-96
Cecil, Stefan: PB-86, PA-25 Dechent, Dagmar: PA-131, PB-138
Chafai, Djamel E.: W3-2 Degirmenci, Elif: PB-26
Chang, Irene: S05-1 Dekali, Samir: PB-96
Chatterjee, Indira: S12-6, PB-104 Del Signore, Federico: PA-119
Chavannes, Nicolas: S09-4 Del Vecchio, Flavia: PB-96
Chen, Guangdi: S06-3, PA-97 della Valle, Elena: PA-119, W3-3
Chen, Hong: PA-69 Demeester, Piet: FA-3, S11-5, PA-23
Chiaramello, Emma: PB-82 Deng, Guanlei: PA-67
Chobineh, Amirreza: S02-6 Denzi, Agnese: W3-5, PA-63
Choi, BeomJin: PA-19 Dermol-Cerne, Janja: PB-66
Choi, Dong Geun: PA-31 Deschamps, François: FB-6, S01-4, PB-84, S10-2
Choi, Heungjae: S06-4 Di Lazzaro, Vincenzo: S15-5
Choi, Hyung Do: PB-114, FB-10, PA-15, PA-111 Dieudonné, Maël: S10-3
Choi, Hyung-Do: PB-116, PB-18, PA-109, PA-85, FA-11, Dijk, Gerwin: FB-4, PB-68
PB-78 Djamel, Allal: PB-38
Choi, Jong Hyuk: PB-78 Dodig, Hrvoje: PB-30, FB-2
Choi, Sojeong: FB-10, PB-114 Doi, Tsukasa: S04-2
Chopra, Nishtha: S11-3 Domingues, Luis: PA-35
Choromańska, Anna: PB-76 Dongus, Stefan: W1-1, S16-3, PB-44
Chou, C. K.: S16-4, PW1-4 Dorn, Hans: S01-3
Christ, Andreas: S13-2, PB-36, PW1-8 Driessen, Sarah: PA-131, PB-138
Chung, Beom Sun: S15-1 Dumontheil, Iroise: S05-1
Chung, Min Suk: S15-1 Dupré, Luc: S14-2
Cifra, Michal: PA-125, W3-1, W3-2 Durand-Fontanier, Sylvaine: PB-120
Cirimele, Vincenzo: S15-2 Durrenberger, Gregor: W1-5
Clavel, Jacqueline: S05-5 Echeverry-Cardona, Laura Maria: FB-9, PB-110
Cleveland, Robert: S16-4 Edeline, Jean-Marc: S08-1
Cocco, Davide: PA-119 Edhemovic, Ibrahim: PB-74
Cockburn, Myles: S05-6 Eeftens, Marloes: S11-1, PB-44, S05-1
Colella, Micol: S15-5 Egblomassé-Roidl, Cornelia: S13-4
Collard, Jean-Francois: PA-83 Eggert, Torsten: S01-3
Colombi, Davide: PW1-9, PB-26, PB-36 El Hajj, Wassim: PW1-11
Conil, Emmanuelle: S02-6 El Khoueiry, Corinne: S03-5, PA-81
Consales, Claudia: S03-3, S03-2 Elias-Viñas, David: PB-2
Corbacio, Michael: S01-4 Elliott, Paul: S05-1
Cosquer, Brigitte: PA-113 Emonds, Tanja: PB-138
Craviso, Gale: S12-6, PB-104 Emoto, Takahiro: PA-99, FA-10
Crespi, Catherine M.: S05-6, S05-5 Enomoto, Hiroyuki: PB-88
Crespo-Valero, Pedro: S04-5, S09-4 Esaki, Kaoru: FB-3, PB-48
Croft, Rodney: S01-5, S16-1, S01-1, S01-2, S10-1, P2-1, Escobar, Roel: PB-148
PA-71 Estenberg, Jimmy: S05-2
Crouzier, David: PB-96 Evans, Ian: S01-5
Cruciani, Silvano: PA-33 Fall, Abdou Khadir: S14-2

798
Faraone, Antonio: S16-4 Gryz, Krzysztof: PA-127, PB-134, PB-28
Farcito, Silvia: S15-1 Guez, David: S07-2
Farrugia, Lourdes: PA-49 Guidon, Manuel: S09-4
Felix-Gonzalez, Nazario: PB-8, PA-51 Guler, Goknur: PA-93, PB-102, FB-8
Feliziani, Mauro: PA-33 Gurhan, Hakki: FB-11, PB-118, S06-1
Feo, Francesco: PB-56 Guxens, Mònica: S11-1
Ferber, Jeannette: PA-69 Ha, Eun-Hee: PB-78
Fernandez, Andrea: PB-8 Ha, Mina: PB-78
Feychting, Maria: S05-5 Haba, Yvonne: PA-125
Findlay, Sheridan: S01-1 Habauzit, Denis: S08-4
Fiocchi, Serena: PB-82 Haberl, Saša: S07-6
Fite, Johanna: W1-3 Hadjiolova, Radka: PA-117
Fleming, Charlotte: S05-1 Hagiwara, Masaki: FB-3, PB-48
Foster, Kenneth: S16-4, PW1-6, PB-36 Hancock, Christofer P.: PA-61
Fourquaux, Isabelle: S03-6 Hansson Mild, Kjell: FA-4, PA-37
Frankel, Jennifer: FA-4, PA-37 Hao, Yanhui: PA-123
Freschi, Fabio: S15-2 Harakawa, Shinji: PA-107
Freudenstein, Frederik : S16-1, S10-1, PA-71 Hartman, Holly: S12-1
Fröhlich, Jürg: S16-3, PA-55, PA-57, W1-1 Hassan, Mahmoud: S01-6
Fujita, Hideki: S04-2 Hather, Catherine: S06-4
Fukunari, Masafumi: S15-6 Havelka, Daniel: W3-2
Funahashi, Daisuke: S02-2 Hayashi, Kiyotaka: PA-139
Gajšek, Peter: PB-62, PA-39, S11-1, S02-4 Hayashi, Shin’ichiro: S06-5
Gallot, Guilhem: W3-8 He, Chengqi: S08-3
Galvez-Coy, Diana Carolina: FB-9, PB-110 He, Hongchen: S08-3
Gao, Peng: PA-59 Heberling, Dirk: S11-2, S14-1
García-Sánchez, Tomás: FA-7, FB-5, W3-8, PB-70, PA-73 Hein, Matthias: S14-1, S11-2
Garibaldi-Covarrubias, Roberto: S13-3 Heinrich, Hannah: S16-5
Gasc, Geraldine: S03-6 Hemon, Denis: S05-5
George, Christopher: S06-4 Herrala, Mikko: S06-2
Geschwentner, Dirk: S13-4 Hikage, Takashi: PB-58, PB-14
Geuzaine, Christophe: S10-3, PA-83, PA-11 HIkichi, Takeo: S04-2
Ghadirian, Reyhane: FB-13, PB-126 Hinsenkamp, Maurice: PA-83
Ghasemi, Negareh: S07-5 Hirao, Kouji: FA-10, PA-99
Ghorbel, Amina: W3-7, PB-70, FB-5 Hirata, Akimasa: S16-4, S02-2, PW1-3, S10-4, S02-1,
Giaccone, Luca: S15-2 S15-4
Gibot, Laure: S03-6 Hirtl, Rene: PB-60
Giremus, Audrey: PA-81 Hjouj, Muhammad: S07-2
Girkontaitė, Irutė: S12-5 Hoffmann, Mathias: S16-5
Gitton, Christophe: PB-90, FB-7 Hofner, Nora: S09-3
Gobinet, Cyril: W3-6 Hong, Seung-Cheol: PA-135
Goethals, Michel: PA-11 Hong, Yun-Chul: PB-78
Golzio, Muriel: S03-6 Hori, Takuya: PA-107
Gomez-Tames, Jose: S02-1, S15-4 Hu, Xiangjun: PB-112, PA-129
Gonzalez-Rubio, Jesus: PA-45 Huang, Xiaomei: PB-132, PA-3, PB-130, FA-1
Goswami, Ishan: PA-121, FA-12 Huang, Yuanyuan: S11-4, S02-5, S02-6
Goulet, Daniel: S01-4, FB-6, S10-2, PB-84 Hugueville, Laurent: FB-7, PB-90
Gourzoulidis, G.A.: PB-146 Hurtier, Annabelle: PB-128, S08-4
Grabner, Karol: PB-12 Huss, Anke: S11-1, PB-44, S04-1
Graefrath, David: PA-131, PB-138 Hussain, Niamat: FB-14, PB-136, PB-144, FB-13, S13-1,
Graf, Kevin: S16-4 PA-133
Gresits, Ivan: PB-42 Hwang, Taewook: PB-46
Griffo, Raffaele: S11-6 Höytö, Anne: S06-2
Grošelj, Aleš: S07-1, FA-6 Ibey, Bennett: PB-148, PW2-10, PW2-3, PB-32

799
Iftode, Cora: PB-16 Kinouchi, Yohsuke: PA-99, FA-10
Ikehara, Toshitaka: FA-10, PA-99 Kiouvrekis, Y.: PB-146
Ikehata, Masateru: S03-1 Kismali, Gorkem: FB-8, PB-102, PA-93
Ikuyo, Miwa: PB-48, FB-3 Klauenberg, B. Jon: PB-148, S16-4
Imai, Shinya: S04-2 Klein, Marie-Muguet: PA-113
Ishiwatari, Hiromi: S15-3 Koch, Michael: PB-86, PB-98
Ito, Takahiro: S02-2, S10-4 Kodera, Sachiko: S02-1
Iyama, Takahiro: PB-14, S02-2, PB-6 Koerber, Rainer: S09-3
Izawa, Shuhei: S04-2 Kojima, Masami: PW1-5, S15-6
Jagielski, Kai: FA-5, FA-2, PA-21, PA-41 Kolosnjaj-Tabi, Jelena: S03-6, PA-95
Jang, Judong: PA-31 Kondo, Tsukasa: S15-3
Jeon, Sangbong: PA-15, PB-18, PA-111, PA-47 Kopacz, Thomas: S11-2, S14-1
Jeong, Min-Joo: PB-144, PB-136, FB-13, S13-1, PA-133, Korpinen, Leena: PB-142
FB-14 Kos, Bor: S02-4, PB-74
Jeong, Ye Ji: PB-116, PA-109, FA-11 Kotnik, Tadej: PB-74
Jeschke, Peter: S16-5 Kouzmanova, Margarita: PA-117
Jin, Yumin: S06-3, PA-97 Koyama, Shin : PA-87, S06-6
Jobard, Elodie: PB-96 Kralj, Slavko: S03-6
Johansen, Christoffer: S05-5 Kramar, Peter: S12-3
Johnson, Samuel: PB-32 Kranjc, Matej: PB-74, S07-6
Joosten, Stephan: S16-5 Kraus, Thomas: FA-5, PA-131, PA-41, PB-138
Joseph, Wout: S11-5, S14-4, S14-6, S14-3, PA-11, PB-44, Kreis, Christian: S05-5
FA-3, S14-2, PA-23 Kromhout, Hans: S04-1
Joshi, Paramananda: PB-26 Kubinyi, Györgyi: PA-89
Joyner, Ken: PW1-1 Kuehn, Sven: S11-5, S04-4
Juutilainen, Jukka: S06-2 Kuhne, Jens: S13-4
Juárez-Castillo, Alberto: PA-43 Kulbacka, Julita: PB-76
Kadoya, Toshihisa: PA-139 Kundi, Michael: W1-4
Kafozoff, Nicoll-Ellen: PA-117 Kurosaki, Shota: PB-48, FB-3
Kainz, Wolfgang: S04-6 Kuster, Niels: S11-5, S15-1, S09-3, S04-6, S04-5, PA-113,
Kandala, Sahithi: FB-11, PB-118, S06-1 S16-6, S13-2, S09-4, S04-4, PW1-10
Kang, Byeong-Nam: PA-133 Kwon, Ho-Jang: PB-78
Kappas, C: PB-146 Kwon, Yongki: PA-31
Karpowicz, Jolanta: PA-127, PB-134, PB-28 Laakso, Ilkka: PB-20, S15-3
Kattnig, Daniel R.: P1-2 Lagroye, Isabelle: PB-128, W2-2, PA-81, S08-4, S03-5
Kavet, Robert: S16-4 Lajevardipour, Alireza: PA-13, S07-4
Kawahira, Hiroshi: FB-1, PB-10 Lambrozo, Jacques: S01-4, S10-2, FB-6, PB-84 Lameth,
Kawase, Kodo: S06-5 Julie: S08-1
Kawshalya, M.D Rajitha: PA-135 Langus, Janez: PB-74
Kayser, Thorsten: PA-141 Last, David: S07-2
Kesari, Kavindra: S06-2 Lastauskienė, Eglė: S12-5
Kesselring, Jürg: S16-3, W1-7, W1-6 Lau, Edmund: PB-80
Kheifets, Leeka: S05-6, S11-1, PB-80, S05-5 Le Brusquet, Laurent: PB-82
Kim, DongGeon: PA-19 Le Dréan, Yves: PA-91, FA-9, S08-4
Kim, Geojung: PB-46 Le Page, Yann: FA-9, PA-91
Kim, Hakrim: S08-2 le Roux, Christiaan: PB-64
Kim, Hong Lae: PB-52, PA-137 Leblanc, Normand: S12-6, PB-104, PW2-8
Kim, Hye Sun: PA-111 Ledent, Maryse: S10-3, PA-83
Kim, Hyung-Gun: S08-2 Lee, Ae-kyoung: PB-18, PB-78
Kim, Ju-Hwan: S08-2 Lee, Byeongyoon: PB-50
Kim, Kyuhyeon: PB-46 Lee, Hae June: FB-10, PB-114
Kim, Nam: FB-13, PB-136, PB-144, PB-116, PA-85, PA-111, Lee, Hae-June: PB-116, FA-11, PA-109
FB-14, FB-10, PB-114, S13-1, PA-109, FA-11, PA-133 Lee, Haeng-Seon: PA-15
Kim, Yangho: PB-78 Lee, Hye-Ryoung: S08-2

800
Lee, Jaesung: PA-47 Martens, Luc: S14-3, S14-2, S14-6, PA-11, PA-23, S11-5,
Lee, Jin-Koo: S08-2 S14-4, FA-3
Lee, Jongil: PA-31 Maslanyj, Myron: S11-3
Lee, Seungwoo: PA-137, PB-144, FB-13, FB-14, PB-52, Massa, Rita: S11-6
PB-136 Masuchi, Ayumi: PA-77
Lee, Yong-Han: PB-78 Matsuhara, Shohei: FB-1, PB-10
Lee, Young Seung: PA-15 Matsuura, Nobuaki: S10-4
Lee, Yun Jin: PA-135 Mattsson, Mats-Olof: PA-7, PA-103, S09-2, W2-3
Lee, Yun-Sil: PB-116, PB-114, FB-10, FA-11, PA-85, Mazloum, Taghrid: S14-5, S02-6
PA-109 Meder, Inger Kristine: S11-1
Legros, Alexandre: S01-4, S01-5, S16-4, S10-2, FB-6, Medina-Salazar, Idalia: PB-2
PB-84, S01-6 Merla, Caterina: PA-61, W3-5, W3-7, PB-72, W3-6, S03-2
Lehtinen, Tuukka: PB-20 Mevissen, Meike: S16-3
Leveque, Philippe: PB-128, S08-1 Mezei, Gabor: PB-80
Lewis, Noëlle: PA-81 Michel, Olga: PB-76
Li, Congsheng: S02-3 Miclaus, Simona: PB-16, PA-49
Li, De-Kun: PA-69 Migliore, Marco Donald: S11-6
Li, Haijuan: PB-112, PA-129 Miklavčič,Damijan: PA-65, S07-1, PB-66, PB-74, FA-6,
Li, Kun: PB-22 S12-3, S07-6, FA-6
Li, Wenchao: PA-123 Min, Kyeongrae: PA-85
Libal, Urszula: PA-1 Mir, Lluis M.: W3-8, W3-7, W3-6, W3-2, W3-5, FB-5, W3-1,
Liberti, Micaela: W3-7, W3-3, S15-5, W3-5, PA-63, PB-56, PB-70, PB-72, PA-73, FA-7
PB-72, PA-119 Miyakoshi, Junji: PA-87, S06-6
Liorni, Ilaria: S04-4 Mizuno, Maya: S15-6, S06-6
Liu, Chuan: PA-105, S03-4 Mizuno, Yukio: PA-139
Liu, Zhen: PA-67 MO, Mireku: S05-1
Lloyd, Bryn: S09-4, S15-1, S04-6 Modolo, Julien: PA-75, S01-6, FA-8
Lloyd, David: S06-4 Moisescu, Cristina: PA-49
Lopez, Borja: FA-7, PA-73 Moser, Mirjana: W2-1
Lopresto, Vanni: S03-3 Mpalantinos, Athanasio: PA-35
Loughran, Sarah: S01-1, S01-2, S10-1, S01-5, S16-1 Mueller, William: S05-1
Lucía, Óscar: FA-7, PA-73 Muheim, Rachel: P1-1
Luis, Pablo: PA-45 Murbach, Manuel: S13-2, S04-6
Luukkonen, Jukka: S06-2 Muscat, Adeline: FA-7, PB-72, PA-73
Ma, Qin-Long: PA-105, S03-4 Naarala, Jonne: S06-2
Maciá, Judith: PB-140 Nagaoka, Tomoaki: PA-17
Madjid Ansari, Alireza: FB-13, PB-126 Nagase, Keisuke: PB-6
Maestu, Ceferino: PA-51, PB-8 Nagy, Noemi: PA-29
Maeyatsu, Humio: S04-2 Nah, Kyungku: PA-137
Magne, Isabelle: PB-82 Nakane, Tatsuhito: S10-4
Maiolino, Paola: S11-6 Nakasono, Satoshi: S03-1
Maiz, Odei: S09-4 Nakatani-Enomoto, Setsu: PB-88
Makino, Kei: S03-1 Nanda, Ranjan: PB-122, FB-12
Malagoli, Carlotta: S05-5 Naranpanawe, Lakshitha: S07-5
Mallat, Michel: S08-1 Nardoni, Martina: W3-7
Mancuso, Mariateresa: PA-61, S03-3 Narita, Eijiro: S06-6, PA-87
Manios, G.: PB-146 Nath, Ravindra: PA-79
Mardor, Yael: S07-2 Naval, Alejandro: FA-7, PA-73
Marino, Carmela: S03-2, S03-3 Necz, Peter: PA-29
Markovskaja, Svetlana: S12-5 Nedachi, Takaki: PA-107
Marquant, Fabienne: S05-5 Nelson, Kaitlin: PB-32
Marracino, Paolo: PA-119, W3-3, PB-72 Nemeth, Zsuzsanna: PA-89
Martano, Manuela: S11-6 Nesterova, Maryna: PB-54

801
Neufeld, Esra: S16-6, S09-3, S13-2, S15-1, S04-4, Pereira de Vasconcelos, Anne: PA-113
S09-4, S04-6 Perrin, Marie-Laure: PB-120
Neumann, Claudine: S16-5 Petersen, Ron: S16-4
Neuschulz, Hannelore: S16-5 Pethig, Ronald: P3-1
Niederhauser, Joel J: S09-1 Petralito, Stefania: W3-7
Nirala, Jayprakash: FB-12, PB-122 Petri, Anne-Kathrin: PB-138
Nojima, Toshio: PB-14 Peyman, Azadeh: S11-3
Novickij, Jurij: S12-5 Pezić, Tamara: S07-6, FA-6, PA-65, PB-66
Novickij, Vitalij: S12-5, S07-6 Pierce, Sophia: S12-6
Nowakowski, Piotr: PA-1 Pillet, Flavien: PA-95
Nuccitelli, Richard: S12-1 Pinchera, Daniele: S11-6
Nájera, Alberto: PA-45, PA-43, PB-140 Pintar, Matevz: PB-74
O'Connor, Rodney P.: PB-94, FB-4, PB-68 Pinto, Rosanna: S03-3, PA-27
Occelli, Florian: S08-1 Pinto, Yenny: S14-5
Odouli, Roxana: PA-69 Piot, Olivier: W3-6
Oetiker, Tobias: S09-4 Pirard, Willy: S10-3
Ohkubo, Chiyoji: PA-139 Pirc, Eva: FA-6, S07-1
Okano, Hideyuki: S15-3 Plante, Michel: S10-2, FB-6, PB-84, S01-4
Oksuzyan, Sona: S05-6 Plaskota, Przemyslaw: PA-1
Oliveira, Paulo: PA-35 Podesser, Bruno: PB-98
Olsen, Jorn: S11-1 Podlaska, Joanna: PB-34
Onishi, Teruo: PB-14, PB-6, S02-2 Poljak, Dragan: FB-2, PB-30
Ootsubo, Shigeru: PA-139 Pophof, Blanka: S13-4
Opitz, Josef: PA-141 Poque-Haro, Emmanuelle: S08-4, PB-128, PA-81, S03-5
Orlacchio, Rosa: FA-9, PA-91 Porch, Adrian: S06-4
Osepchuk, John: S16-4 Potocnik, Tjasa: FA-6, PA-65
Ostiguy, Genevieve: PB-84, S01-4, S10-2, FB-6 Poulletier De Gannes, Florence: PB-128, PA-81, S08-4,
Otsuki, Sato: S06-5 S03-5
Ouadah, Nihal: PB-124 Pääkkönen, Rauno: PB-142
Ozdemir, Emre: S04-1 Pölzl-Viol, Christiane: S13-4
Ozgur, Elcin: PB-102, FB-8, PA-93 Raaschou-Nielsen, Ole: S05-5
Pack, Jeong-Ki: PA-111, PA-47, PB-116, FA-11, PA-109 Rajamani, Paulraj: FB-12, PB-122
Paffi, Alessandra: PB-56, PA-63 Ravazzani, Paolo: PB-82
Paik, Man-Jeong: PA-111 Reberšek, Matej: FA-6, PB-66, PA-65, S07-1, FA-6
Pakhomov, Andrei: PW2-2, PW2-10, S12-2, PW2-1, P4-1 Reilly, J. Patrick: S16-4
Pakhomova, Olga: PW2-7 Rembiałkowska, Nina: PB-76
Palego, Cristiano: PA-61 Renom, Rémy: PB-128, S03-5, PA-81, S08-4
Palmisano, Stephen: S01-5 Restrepo-Parra, Elisabeth: PB-110, FB-9
Panariello, Gaetano: S11-6 Ricevuti, Giovanni: PB-100
Parappurath Vasudevan , Prakash: S14-2 Rivera-Gonzalez, Marco: PA-51, PB-8
Parazzini, Marta: PB-82 Rodic, Tomaz: PB-74
Park, Hyesook: PB-78 Rols, Marie-Pierre: PA-95, S03-6
Park, Ji Woong: PB-144, S13-1, PB-136, FB-14, FB-13, Romanus, Erik: S16-5
PA-133 Romeo, Stefania: PB-100, PA-103, PB-108, PA-27, S07-3
Park, Jin Seo: S15-1 Rudno-Rudzinski, Krzysztof: PA-1
Park, SangWook: PA-19 Ruigrok, Hermanus: PB-68, PB-128, FB-4
Pasquali, Emanuela: S03-3 Röösli, Martin: PB-82, S11-1, PB-44, S16-3, S05-1, S05-5
Pauli, Mario: PA-141 Saczko, Jolanta: PB-76
Payne, Jason: PB-148, PB-32, S10-5 Saito, Atsushi: S03-1
Paškevičius, Algimantas: S12-5 Saito, Kazuyuki: PB-10, PA-5, FB-1
Pecchia, Leandro: FA-6, S07-1 Sakakibara, Keita: PB-58, PB-14
Pedersen, Camilla: S05-5 Salazar-Galvez, Yuridia: S13-3
Peng, Ruiyun: PA-123 Salomon, Sharona: S07-2
Percherancier, Yann: S08-4, PB-128, S03-5 Samaras, Theodoros: W1-2, PB-60, T2-1

802
Sammut, Charles: PA-49 Sukhraj, Ruby: PB-104
Sanchez-Montero, Rocio: PA-45 Sun, Chuan: PA-97, S06-3
Sannino, Anna: PA-103, PB-100 Sun, Xiaobang: S02-3
Santangeli, Stefania: S03-2 Sun, Yongyan: FA-1, PA-3
Santer, David: PB-98 Šušnjara, Anna: FB-2, PB-30
Sarnago, Héctor: FA-7, PA-73 Sustar, Tomaz: PB-74
Sarolic, Antonio: PA-53, PB-24 Suzuki, Hiroshi: PA-107
Sasaki, Hiroshi: S15-6 Suzuki, Yaeko: PW1-5
Sasaki, Kensuke: PB-22, S06-6, S15-6 Suzuki, Yukihisa: S15-6, S03-1
Sauleau, Ronan: PA-91, FA-9 Švedienė, Jurgita: S12-5
Sauter, Cornelia: S01-3 Swanson, John: S05-5
Scarfi, Maria: W2-4 Szewczyk, Anna: PB-76
Scarfi, Maria Rosaria: PB-100, PA-103 Taibi, Abdelkader: PB-120
Schaap, Kristel: S04-1 Tajima, Tomoyuki: FB-1, PB-10
Scherman, Michaël: W3-7 Takeshita, Nobuyoshi: FB-1, PB-10
Schettino, Fulvio: S11-6 Taki, Masao: S06-6, S15-6, FB-3, PB-48
Schiessl, Sascha: S11-2, S14-1 Tang, Chao: PB-132, FA-1, PA-3, PB-130
Schmid, Gernot: PB-60, S01-3 Tanghe, Emmeric: FA-3, PA-23
Schmidt, Janine-Alison: S13-4 Tani, Masahiko: S15-6
Schmiedchen, Kristina: PA-131 Tanori, Mirella: PA-61
Schneeweiss, Pia: PA-21, FA-5, FA-2, PA-41 Tarantini, Letizia: S03-2
Schneider, Fabian: PA-57, PA-55 Tarao, Hiroo: PA-99, FA-10
Schuermann, David: S16-3 Tarek, Mounir: W3-4, P4-2
Sekiguchi, Tetsuya: PB-58 Tasaki, Takafumi: S15-6
Sel, Tevhide: PA-93, FB-8, PB-102 Tatematsu, Yoshinori: S15-6
Selmaoui, Brahim: PA-89, FB-7, PB-90 Tell, Richard: S16-4
Senturk, Fatih: PA-93, PB-102, FB-8 Teplan, Michal: PA-125
Serša, Gregor: PB-74, S07-1, FA-6 Tfayli, Ali: W3-6
Setti, Stefania: PA-119, S15-5 Thansandote, Artnarong: S16-4
Sharabi, Shirley: S07-2 Thayer, Kristina: T1-1
Sharma, Radhey Shyam: PA-79 Theiler, Tobias: PA-21, FA-2
Sheppard, Asher: PB-56 Thielens, Arno: PA-23, S14-2, FA-3, PB-44, S14-3
Shikhantsov, Sergei: PA-23, FA-3 Thomas, Michael: S05-1
Shimizu, Yoko: PA-87, S06-6 Thors, Björn: PB-36, PB-26
Shinohara, Naoki: PA-87, S06-6 Thuroczy, György: PA-27, PB-42, PB-108, PA-89, PA-29
Shoaf, Adam: PB-104 Tittarelli, Andrea: S05-5
Silny, Jiri: PB-98 Tlacuilo-Parra, Alberto: S13-3
Simko, Myrtill: PA-7, S09-2, PA-103, PB-108 Togo, Hiroyoshi: S10-4
Simon, Ferenc: PB-42 Toledano, Mireille: S05-1
Singh, Rajeev: PA-79 Tominaga, Toshihiko: FA-10, PA-99
Snoj, Marko: S07-1, FA-6 Torfs, Guy: FA-3, PA-23, S11-5
Sol, Jerome: S14-2 Tornevik, Christer: PB-26, PB-36
Soldati, Marco: PB-20 Torres Osorio, Javier: PB-110, FB-9
Song, Kiwon: PA-85 Traikov, Lubomir: PA-117
Soto-Sumuano, Leonardo: S13-3 Trampel, Christopher: S10-6
Souques, Martine: S10-2, FB-6, PB-82, S01-4, PB-84 Trček, Tomaž: PB-62, S11-1, PA-39
Sözer, Esin B: PB-106 Tsitsia, V.: PB-146
Spycher, Ben: S05-5 Tuck, Deirdre: S05-5
Steiner, Edith: S16-3 Tynes, Tore: S05-5
Storm, Jan: S09-3 Tyrakis, Charilaos: PB-146
Struchen, Benjamin: S11-1 Überbacher, Richard: PB-86, PA-25
Stunder, Dominik: FA-5, PA-41, PA-131, PA-21, FA-2 Uecker, Darrin: S12-1
Su, Liling: S06-3 Ugawa, Yoshikazu: PB-88
Sudan, Madhuri: S05-5 Untereiner, Valérie: W3-6

803
Valič, Blaž: PB-62, S11-1, S02-4, PA-39 Wünsch-Filho, Victor: S05-5
Van Den Bossche, Matthias: S14-3, S14-6, PA-11 Xiong, Yanyu: S06-1
Van Holen, Roel: S14-2 Xu, Bo: PB-26, PB-36
Van Thillo, Wim: S11-5 Xu, Zhengping: S06-3, PA-97
Van Torre, Patrick: PB-44 Yaekashiwa, Noriko: S06-5
van Wel, Luuk: S11-1 Yahia-Cherif, Lydia: PB-90, FB-7
Van Wyk, Marthinus: PB-64 Yakubo, Takashi: PA-5
Vandana, Kumari: FB-12, PB-122 Yamaguchi-Sekino, Sachiko: S04-2
Vanhecke, Kris: PA-11 Yamaguchi, Shunsuke: PA-99, FA-10
Vatovez, Benjamin: S10-3 Yamazaki, Kenichi: S16-4
Venturini, Letizia: PB-100 Yamazaki, Madoka: PB-88
Verbridge, Scott S.: PA-121, FA-12 Yan, Keping: PA-67
Verdugo-Diaz, Leticia: PB-2 Yan, Tony: PA-101
Vergara, Ximena: S05-5, S05-6, S14-6 Yang, Chuanjun: FA-1, PA-3, PB-132
Verlaek, Mart: PA-11 Yang, Hee-Sun: PB-78
Verloock, Leen: PA-11, S14-6 Yang, Lisha: S12-6
Vermeeren, Gunter: S11-5, FA-3, PA-23 Yang, Xiaotian: S08-3
Vermeulen, Roel: PB-44, S11-1 Yao, Aiping: S04-5
Vernier, P. Thomas: S12-6, PB-106, PW2-4, S12-4, S12-3, Yardin, Catherine: PB-120
S07-3 Yebda, Thinhinane: PA-81
Verrender, Adam: S01-2, S10-1 Yeo, Geun Teak: PB-52, PA-137
Verschaeve, Luc: S10-3, PA-83 Yi, Sang-Hwa: PB-50
Vesselinova, Liubina: PA-117 Yimaer, Aziguli: PA-97, S06-3
Veyret, Bernard: S03-5, PB-128, S08-4, PA-81 Yin, Shengyong: PA-67
Vezinet, Rene: PA-95 Yoon, Kyeonghee: FB-10, PB-114
Vilagosh, Zoltan: PA-13, S07-4 Yoshie, Sachiko: S03-1
Villard, Sebastien: S01-4, PB-84, FB-6, S10-2 Yu, Hui: PB-132, PA-3, PB-130, FA-1
Villegier, Anne-Sophie: PB-124 Yu, Zheng-Ping: PA-105, S03-4, PA-59
Vinceti, Marco: S05-5 Yumoto, Hiromichi: FA-10, PA-99
Visser, Christo: PB-64 Zahner, Marco: PA-57, PA-55
von Spakovsky, Michael R.: PA-121, FA-12 Zambotti, Alessandro: PA-61
Vrijheid, Martine: S11-1 Zare, Farzan: S07-5
Wada, Keiji: S03-1 Zastrow, Earl : S04-5, S04-6
Wake, Kanako: FB-3, PB-48 Zeleke, Berihun: S05-3, PA-71
Wallace, Jasmina: FB-7, PB-90 Zeni, Olga: PA-7, PA-103, S07-3, PB-100
Wang, Lifeng: PB-112, PA-129 Zhadobov, Maxim: PA-91, FA-9, S14-2
Wang, Rui-Sheng: S04-2 Zhan, Tianying: PA-115
Wang, Xiaomei: PA-115, PB-4 Zhang, Hongying: PA-3, FA-1
Watanabe, Soichi: PB-22, PA-17, S15-6, S06-6 Zhang, Lei: PA-105, S03-4, PA-59
Watanuki, Keiichi: S15-3 Zhang, Xiaoyun: PB-4
Weaver, James: PW2-6 Zhang, Ziyan: PA-3, FA-1, PB-130
Wei, Xiaoxia: PA-97 Zhao, Li: PA-123
Wendling, Fabrice: FA-8, PA-75 Zhao, Yingying: PA-115
Whitmore, Jeffrey: PB-32, S10-5, PB-148 Zheng, Chao: PA-67
Wiart, Joe: PB-38, S02-6, S14-4, S11-4, S14-5, S02-5 Zheng, Feizhou: PA-59
Wiedemann, Peter: PA-71, S16-1 Zheng, Shusen: PA-67
Wiesbeck, Werner: PA-141 Zhi, Weijia: PB-112
Wilen, Jonna: FA-4, PA-37 Zhou, Zhou: PA-105, S03-4
Williams, Catrin: S06-4 Zhu, Siyi: S08-3
Williams, Rhodri: S06-4 Ziegelberger, Gunde: S13-4
Winter, Lukas: PA-131 Zinkevičienė, Auksė: S12-5
Wittlich, Marc: S04-3 Ziskin, Marvin: PB-36, S16-4
Wood, Andrew: PA-13, S16-2, S07-4 Zradziński, Patryk: PA-127, PB-134, PB-28
Wood, Michael: PW1-2 Zubrzak, Bartlomiej: PA-1
Wu, Tongning: S02-3
804