Comparison

See discussions, stats, and author profiles for this publication at: https://www.researchgate.
net/publication/315059917
Comparison of Heart Rate Variability Recording With Smart Phone

Photoplethysmographic, Polar H7 Chest Strap and Electrocardiogram
Methods
Article in International journal of sports physiology and performance · March 2017

DOI: 10.1123/ijspp.2016-0668
CITATIONS READS
91 16,684
6 authors, including:
Daniel J Plews Ben Scott

Auckland University of Technology Loughborough University
42 PUBLICATIONS 1,226 CITATIONS 2 PUBLICATIONS 106 CITATIONS
SEE PROFILE SEE PROFILE
Marco Altini Matthew R. Wood

HRV4Training Auckland University of Technology
51 PUBLICATIONS 719 CITATIONS 22 PUBLICATIONS 322 CITATIONS
SEE PROFILE SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Defining the role of anaerobic speed reserve in middle distance running View project
Paul's PhD View project
All content following this page was uploaded by Marco Altini on 27 March 2017.
The user has requested enhancement of the downloaded file.

“Comparison of Heart Rate Variability Recording With Smart Phone Photoplethysmographic, Polar H7 Chest Strap and
Electrocardiogram Methods” by Plews DJ et al.
International Journal of Sports Physiology and Performance
© 2017 Human Kinetics, Inc.
Article Type: Original report
Title: Comparison of heart rate variability recording with smart phone photoplethysmographic,
Polar H7 chest strap and electrocardiogram methods.
Author: Daniel J. Plews1, 2, 3, Ben Scott1,4, Marco Altini5, Matt Wood2, Andrew E. Kilding2
and Paul B. Laursen1, 2
Affiliations:
1. High Performance Sport New Zealand, Auckland, New Zealand
2. Sports Performance Research Institute New Zealand (SPRINZ), Auckland University of
Downloaded by Alderman Library on 03/14/17, Volume 0, Article Number 0
Technology, Auckland, New Zealand

3. University of Waikato, Hamilton, New Zealand
4. Loughborough University, Loughborough, United Kingdom
5. ACTLab, University of Passau, Germany
Contact Information:
Daniel Plews
High Performance Sport New Zealand
Millennium Institute of Sport & Health,
17 Antares Place,
Mairangi Bay, 0632, New Zealand
Ph: +64 21 250 9591
Fax: +64 9 479 1486
Corresponding Author: Daniel Plews: daniel.plews@hpsnz.org.nz
Abstract: 174
Main text: 2745
Figures and tables: 2 Figures, 1 Tables

Abstract
Purpose: To establish the validity of smartphone photoplethysmography (PPG) and heart rate
sensor in the measurement of heart rate variability (HRV). Methods: 29 healthy subjects were
measured at rest during 5 min of guided breathing (GB) and normal breathing (NB) using
Smartphone PPG, heart rate chest strap and electrocardiography (ECG). The root mean sum of
the squared differences between R–R intervals (rMSSD) was determined from each device.
Results: Compared to ECG, the technical error of estimate (TEE) was acceptable for all
conditions (average TEE CV% (90% CI) = 6.35 (5.13; 8.5)). When assessed as a standardised
difference, all differences were deemed “Trivial” (average std. diff (90% CI) = 0.10 (0.08;
0.13). Both PPG and HR sensor derived measures had almost perfect correlations with ECG
(R = 1.00 (0.99; 1:00). Conclusion: Both PPG and heart rate sensor provide an acceptable
agreement for the measurement of rMSSD when compared with ECG. Smartphone PPG
technology may be a preferred method of HRV data collection for athletes due to its practicality
and ease of use in the field.
Keywords: Cardiac parasympathetic, monitoring, athletic performance

Introduction
The ability to monitor human movement and physiological state has advanced rapidly
in recent years. As one example, “Smart” devices, which use technologies such as
accelerometry, actigraphy and photoplethysmography can measure many aspects of human
performance and movement.1 For athletes striving for peak performances, the need to
effectively monitor human movement and physiological state are important so that more
objective decisions around training can be made.2 The regular assessment of heart rate
variability (HRV) has immerged as one measure of “physiological state” that has grown in its
popularity and is used by many sporting teams and athletes on a day-to-day basis.3 HRV
involves measurement of the variation between individual heart beats across consecutive
cardiac cycles, and this variation can provide an estimate of a person’s autonomic nervous
system (ANS) activity.4
Several aspects can converge to reduce daily athlete measurement compliance,
including the convenience of having the appropriate equipment available each morning, a
consistent morning room temperature to enable ease of putting on a chest strap, and other
factors. Moreover, due to the natural relative variability or noise of daily HRV recordings,
multiple daily recordings are required, with weekly and rolling averages needed to gain a true
representation of an athlete’s physiological state.5-7 As such, ways by which HRV recording
can be improved would be advantageous to both coaches and practitioners wishing to use HRV
in the field.
Photoplethysmography (PPG) is one technological advancement that may allow HRV
to be measured simply via a smartphone device. PPG is measured via reflection through the
illumination of the skin using an LED (e.g. the smartphone’s flash) and through detection of
the amount of light that is reflected by a photodetector or a camera located next to the light
source. The resulting PPG signal is composed of a direct current (DC) component, which varies
slowly depending on tissue properties and blood volume. The alternating current (AC)
component varies more rapidly to detect the pulsatile factor. After cardiac systole, local blood
volume increases acutely, reducing the received light intensity. During diastole, blood volume
decreases, and light reflection increases.8 Compared with other HRV measurement devices
used by athletes (e.g. heart rate monitor sensors), PPG can then be considered a more user-
friendly model of HRV attainment, as no additional apparatus is required other than a
smartphone device that can easily transfer acquired data via Wi-Fi or 3/4G transfer to the
internet. Together, these combined innovations have the capacity to greatly improve athlete
compliance via enhanced ease of daily recording.
The aim of this study was to compare the accuracy and validity of HRV recordings
attained via a PPG smartphone application (HRV4Training), and via the Polar H7 (a device
more traditionally used by athletes to record HRV in a practical setting), alongside “gold
standard” electrocardiography (ECG).
Methods
Participants
Twenty-nine subjects were initially recruited for this study. From this data set 2 subjects
were removed, as they were unable to complete an entire 60 s of usable PPG data. Another
subject was removed due to a suspected heart arrhythmia. This left 26 complete data sets to be
used in the final analysis (♂ = 22, ♀= 7, age = 31 ± 10 years; Height = 175 ± 9 cm; weight =
73 ± 11; BMI = 23.7 ± 2.3). Of these 26 subjects, 3 were elite athletes, 13 were well-trained
athletes and 10 were recreationally-trained athletes. Prior to taking part in the study, all
participants completed a standardised medical screening form and provided written informed
consent. The study was approved by the Human Research Ethics Committee of AUT
University. Participants were provided with a demonstration of how to use the PPG smartphone
application before they completed 5 min of guided learning time where they could become
familiar with how to use the app, including how to apply appropriate finger pressure as well as
use an entrained breathing setting.
Data acquisition and processing
Camera and HR chest strap data were acquired and processed using an in-house built
smartphone application. This application could acquire simultaneous RR intervals from a Polar
H7 Bluetooth heart rate monitor and a phone camera.9 ECG data was acquired using a
diagnostic quality 12-lead system (Cosmed, Quark T12x, USA).

Prior to electrode placement, the skin of participants was prepped at the appropriate
sites by way of shaving, abrading and swabbing with alcohol wipes. A standard 12-lead
electrode placement was used for ECG recording. The six chest leads were placed as follows:
V1 in the fourth intercostal space to the right of the sternum, V2 in the fourth intercostal space
to the left of the sternum, V3 between V2 and V4, V4 in the fifth intercostal space in the
midclavicular line, V5 between V4 and V6 and V6 in the fifth intercostal space in the
midaxillary line. Finally, arm electrodes were placed 2 cm below the anterior deltoids in the
midclavicular line and leg electrodes were placed medially from the suprailiac crest in the
midclavicular line. Once the ECG had been attached, participants were given a Polar H7 heart
rate monitor, which was fitted just below V6.
Photoplethysmography
Photoplethysmography (PPG) was acquired via a commercially available smartphone
application known as “HRV4training” (see http://www.hrv4training.com/). Given the low
frame rate of mobile phone cameras, different signal processing techniques should be
employed to derive HRV from the phone video stream.10 HRV4Training acquires a video
stream at a frame rate of 30 Hz, where red, green, and blue (RGB) channels are averaged over
the entire frame, before converting between the RGB and the hue, saturation, and value (HSV)
colour space. The intensity component of the HSV colour space is filtered using a Butterworth
band pass filter of order 4 and frequency pass band between 0.1 and 10 Hz, to remove the DC
component of the signal, as well as any high-frequency noise while maintaining the AC
component. Finally, cubic spline interpolation is used to up-sample the signal between 30 and
180 Hz. Up-sampling of the data is a necessary requirement for sufficient resolution of HRV
feature computation.11
RR interval extraction, data synchronization, and features computation.

HRV4Training implements a peak detection algorithm to determine peak-to-peak
intervals from up-sampled PPG data. Peak detection is based on a slope inversion
algorithm,9 where peak-to-peak intervals are corrected for artefacts according to two criteria.
First, consecutive RR intervals extracted with PPG are removed when they differ by more than
75% from the previous one. Additionally, outliers are removed by including only RR intervals
that are within less than 25% of the 1st quartile and within more than 25% of the 3rd quartile.
This technique avoids over-correcting, a problem of the widely employed removal of
consecutive RR intervals differing by more than 25%12 for individuals with very high beat-to-
beat variability. Finally, the first or second minute of data were discarded when the PPG signal
was disrupted by excessive noise, e.g. due to the participant’s movement or other unidentified
causes beyond the scope of this comparison.
The Polar H7, as with other Bluetooth low energy chest straps, already provides RR
intervals, and therefore RR data does not require additional processing.
ECG data was exported from the Cosmed Quark T12x system. A continuous wavelet
transform based beat detection algorithm was used to extract RR intervals from lead 3 of
the ECG data. A custom software was then used to display ECG and detected peaks so that the
experimenter could manually edit detected peaks to ensure correctness of the algorithm output.
The same lead configuration and processing procedure was used for all subjects.
RR intervals for PPG and H7 data were acquired using an in-house built smartphone
application so that RR interval data could be almost perfectly synchronized. However, some
limitations did apply. First, Bluetooth low energy radio packets have priority over camera
acquisition and therefore could from time to time introduce small (order of milliseconds) delays
in PPG data acquisition. As a result, this setup is a worst case scenario for time sensitive
operations such as RR interval extraction from a camera-based data stream. Second, data could
not be synchronized automatically as the Bluetooth low energy protocol does not provide
timestamped RR intervals, but sends RR intervals appended to the average heart rate of the
past second. Hence, RR intervals gathered over the relevant 60 s window were appended and
visual synchronization was necessary before HRV computation. Similalry, manual
synchronization was necessary for ECG data, as these data was acquired from a separate system
(Cosmed Quark T12x). Manual synchronization was performed by visually aligning the RR
interval time series (see for example Fig. 1), as RR interval oscillations due to breathing allow
for visual synchronization regardless of small time delays due to the unlikelihood of starting
the different systems at exactly the same time.
Testing procedure
5-min recordings were taken under two conditions; sitting guided breathing (GB) and
sitting normal breathing (NB). Sitting was chosen to reduce any possible parasympathetic
saturation which is often observed in individuals with low resting heart rates.15 Recordings
were taken in the same order as listed before duplicate measures were taken in the identical
order. Participants rested in each position for 1 min before beginning a recording to reduce the
influence of movement on HRV. As 1-min HRV data has been shown to be as valid a measure
as longer time frames,16 data were measured during the first 1 min of recordings after discarding
the first 5 s. As such, data were included from 5 s to 1 min 5 s (60 s total duration), thereby
allowing for a 5 sec stabilization period. Furthermore, as an aim of this study was to investigate
ways by which to increase the practicality and ease of HRV data capturing, only 1-min
durations were investigated.
Importantly, erroneous data were discarded from any recording. The in-house built
data-capturing application was designed to inform the user whether data were of sufficient
quality or not. For this, the first minute of data was discarded in two circumstances for LB1
and in three circumstances for LNB1. Periods of high noise were identified by analyzing the
percentage of discarded RR intervals over a given time-period, as RR intervals are discarded
when timing differences are outside of expected or normal values, typically due to underlying
noise or ectopic beats. In cases where the rMSSD data attained were inappropriate due to user
error (e.g. movement of the finger over the camera), the subject would be informed and data
would be discarded. The subject would then be asked to make another recording until it was
deemed successful.
Statistical analysis
All data are presented as mean ± 90% confidence limits (CL) unless otherwise stated.
Comparisons to rMSSD values derived from ECG to Polar H7 and PPG were achieved using
a Pearson product-moment correlation analysis, standard linear regression, typical error of
estimate (TEE) and mean bias (%). Inspection of the slope and intercept of the linear regression
was examined to characterize the level of agreement between PPG to ECG and Polar H7 to
PPG. The TEE and mean bias between PPG and the other methods were determined using an
excel spreadsheet,17 with the TEE expressed both in raw units and as a percentage. To assess
differences between measures, standardised differences were also calculated using the same
spreadsheet. The following threshold values for standardized differences were ≤0.2 (trivial),
>0.2 (small), >0.6 (moderate), >1.2 (large), and >2.0 (very large). The magnitude of the
correlation between PPG/Polar H7 to ECG was assessed with the following thresholds <0.1,
trivial; <0.1–0.3, small; <0.3–0.5, moderate ;< 0.5–0.7, large ;< 0.7–0.9, very large; and <0.9–
1.0, almost perfect. If the 90 % confidence intervals (CI) overlapped small positive and
negative values, the magnitude of correlation was deemed ‘unclear’.
Results
An example of raw R-R data from one subject across the recording period for all capture
methods is shown in Figure 1. The TEE for all four conditions are presented in Table 1.
Compared with ECG, PPG GB had the lowest TEE (CV% (90% CI) = 3.8 (3.1; 5.0)) whereas
Polar H7 NB had the highest CV% (90% CI) = 8.6 (6.9; 11.6)). When assessed as a standardised
difference (PPG/Polar H7 vs. ECG), all differences were deemed “Trivial”.
The magnitudes of the correlation between PPG/Polar H7 and ECG are shown in Figure
2. All methods of HRV assessment displayed almost perfect correlations compared with ECG.
PPG vs. ECG GB displayed the clearest correlation (r = 1.00 (1.00; 1.00), whereas the Polar
H7 NB showed the slightly lower correlation (r = 0.99 (0.98; 1.00). However all correlations
were deemed “almost perfect”.
Discussion
We have previously shown that in order to effectively monitor an athlete’s HRV,
weekly rolling-averaged values tend to be more useful than values taken on an isolated day.5
As such, daily monitoring and “clean” data is paramount for practitioners and coaches to
effectively monitor an athlete’s training adaptation via the use of morning resting HRV3.
However, in some athletes, achieving daily compliance can often be difficult.5 The main
finding of this study is that both PPG recorded via smartphone technology and the Polar HR
sensor has acceptable levels of agreement with ECG for recording the rMSSD index of HRV.
The ability to effectively record HRV via an athlete’s own smartphone and PPG
technology is one method that would indeed simplify data acquisition. This method removes
the necessity to acquire and fit the HR strap (methods which have been traditionally used by
athletes collecting HRV5,13,18) with now just one device needed (i.e. a smartphone). This also
happens to be a device that is now traditionally positioned at bedside by most individuals and
used as their alarm clock, etc., which makes the early morning routine of HRV data capture
relatively seamless.
Although all methods of HRV assessment used in the present study were shown to be
acceptable, the PPG method, using GB, showed the lowest TEE (CV% = 3.8 (3.1; 5.0) and
standardized difference (Std diff = 0.06 (0.05; 0.08) “trivial”). Furthermore, the mean bias in
raw units was ≤ 2.0 ms (Table 1). Considering that rMSSD values typically range from ~50-
250 ms, this is a very small bias. When we contrast other studies that have compared HRV
values measured through PPG against ECG it is currently difficult, as differences between
experimental settings and/or methods of analysis are apparent.19 Furthermore, many of these
studies have been carried out using a variety of “clip-on” devices (e.g. devices clipped onto the
finger or earlobe) rather than the smartphone camera per se. For example, Esco et al20 recently
compared PPG smartphone with ECG and similarly found “trivial” HRV differences (Std diff
= 0.15). Interestingly, during their supine recordings without breathing control, these authors
found the same negligible differences (Std diff = 0.15 vs 0.14). Similarly, Esco et al.20 found
almost a perfect correlation between PPG and ECG HRV recordings.
A novel inclusion in the present study is that we also compared the Polar H7 to ECG as
this heart rate sensor is currently the method being adopted by most athletes when measuring
HRV in the field.5,13,18 Indeed, all methods of HRV assessment compared to ECG were
statistically the same. For example, a correlation of 0.97, 0.99 or 1 are all practically identical,
and any small discrepency of value is likely the result of a small differences in the rMSSD
value (Figure 2.). These correlations are indeed very high, but due to the wide range of rMSSD
values (20-300 ms), statistical artifacts may have pushed the correlations to extremely high
values. Conversely, if all the rMSSD values had been lower (e.g. 40-60 ms), correlations may
not have been as high. The equivalency of the results obtained via all methods examined in the
present study is further supported by the “trivial” standardised differences shown.
Practical Application
Daily athlete compliance to complete HRV recordings can often be difficult. Due to the
relative noise of HRV recordings, daily recordings are required, with weekly and rolling
averages needed to gain a true representation of an athlete’s physiological state.5-7 Although all
methods we compared to gold standard ECG were acceptable, HRV recorded via PPG
smartphone technology with guided breathing showed the strongest validity compared with
ECG measures. Given the ease and practicality of use, such a data-capturing and analysis
system may be more advantageous than other methods of daily HRV assessment as daily
compliance is likely to be enhanced.
Conclusion
Measures of rMSSD derived via PPG and Polar H7 during guided and normal breathing
both shared acceptable agreement to HRV recorded via ECG. Given the superior practicality
and strong validity of HRV recorded via PPG with guided breathing, this method may be the
most sensible choice to select when assessing HRV on athletes in the field.
Conflicts of interest
Marco Altini is the owner and developer of HRV4Training.

References
1. Appelboom G, Camacho E, Abraham ME, et al. Smart wearable body sensors for
patient self-assessment and monitoring. Archives of public health = Archives belges de
sante publique 2014;72:28.
2. The Age of the Analyzed Athlete: Big data helps high-performance sports stars improve
their game., 2016. (Accessed at http://www.ca.com/us/rewrite/articles/application-
economy/the-age-of-the-analyzed-athlete.html)
3. Buchheit M. Monitoring training status with HR measures: do all roads lead to Rome?
Front Physiol 2014;5:73.
4. TaskForce. Heart Rate Variability. European Heart Journal 1996:354-81.
5. Plews DJ, Laursen PB, Kilding AE, Buchheit M. Heart rate variability in elite
triathletes, is variation in variability the key to effective training? A case comparison.
Eur J Appl Physiol 2012;112:3729-41.

6. Plews DJ, Laursen PB, Kilding AE, Buchheit M. Evaluating training adaptation with
heart-rate measures: a methodological comparison. Int J Sports Physiol Perform
2013;8:688-91.
7. Plews DJ, Laursen PB, Le Meur Y, Hausswirth C, Kilding AE, Buchheit M. Monitoring
training with heart rate-variability: how much compliance is needed for valid
assessment? Int J Sports Physiol Perform 2014;9:783-90.
8. Bagha SaS, L. A real time analysis of ppg signal for measurement of spo2 and pulse
rate,. International journal of computer applications 2011;36:45-50.
9. Altini. M aAO. HRV4Training: Large-Scale Longitudinal Training Load Analysis in
Unconstrained Free-Living Settings Using a Smartphone Application. Accepted for
publication at EMBC 2016.
10. Lee J, Reyes BA, McManus DD, Mathias O, Chon KH. Atrial fibrillation detection
using a smart phone. Conf Proc IEEE Eng Med Biol Soc 2012; 2012:1177-80.
11. Voss A, Wessel, N., Sander, A., Malberg, H. and Dietz, R. Influence of low sampling
rate on heart rate variability analysis based on nonlinear dynamics. Computers in
Cardiology 1995:698-2.
12. Lippman N, Stein KM, Lerman BB. Comparison of methods for removal of ectopy in
measurement of heart rate variability. Am J Physiol 1994;267:H411-8.
13. Plews DJ, Laursen PB, Stanley J, Kilding AE, Buchheit M. Training adaptation and
heart rate variability in elite endurance athletes: opening the door to effective
monitoring. Sports Med 2013;43:773-81.
14. Penttila J, Helminen A, Jartti T, et al. Time domain, geometrical and frequency domain
analysis of cardiac vagal outflow: effects of various respiratory patterns. Clin Physiol
2001;21:365-76.
15. Sacknoff D, Gleim G, Stachenfeld N, Glace B, Coplan N. Suppression of high-

frequency power spectrum of heart rate variability in well-trained endurance athletes`.
Circulation 1992;86:I-658.
16. Flatt AA, Esco MR. Heart rate variability stabilization in athletes: towards more
convenient data acquisition. Clin Physiol Funct Imaging 2015: 36(5): 331-336
17. Hopkins WA. Analysis of validity by linear regression (Excel spreadsheet). A new view
of statistics sportsci org: Internet Society for Sport Science 2000.
18. Plews DJ, Laursen PB, Kilding AE, Buchheit M. Heart-rate variability and training-
intensity distribution in elite rowers. Int J Sports Physiol Perform 2014;9:1026-32.
19. Schafer A, Vagedes J. How accurate is pulse rate variability as an estimate of heart rate
variability? A review on studies comparing photoplethysmographic technology with an
electrocardiogram. Int J Cardiol 2013;166:15-29.
20. Esco MR, Flatt AA, Nakamura FY. Agreement between a smart-phone pulse sensor
application and ECG for determining lnRMSSD. J Strength Cond Res 2016. Epub
ahead of print
Figure 1: Simultaneous R-R interval of an individual subject during 60 seconds of recording

for photoplethysmographic (PPG), Polar chest strap (H7) and electrocardiogram (ECG).
Figure 2: Correlation plots (±90 % confidence intervals expressed by dashed lines) and linear
regression equations for photoplethysmography (PPG) and Polar H7 heart rate sensor during
guided and normal breathing. Solid black line represents line of equivalence (r = 1.0).
Table 1:
1-min TEE 90% CI Std. 90% CI Qualitativ Mean 90% CI

measure as a differen e inference bias
CV ce (ms)
%
PPG vs. ECG, 3.8 3.1; 5.0 0.06 0.05; Trivial 2.0 1.3; 2,7
GB 0.08
H7 vs. ECG, 6.1 4.9; 8.1 0.10 0.08; Trivial -0.4 -0.6; 1.4
GB 0.13
PPG vs. ECG, 6.9 5.6; 9.3 0.11 0.09; Trivial 1.4 0.2; 2.6
NB 0.15
H7 vs. ECG, 8.6 6.9; 11.6 0.14 0.11; Trivial -1.5 -3.3; 0.4
NB 0.18
PPG = Photoplethysmographic; ECG = Electrocardiogram; H7 =heart rate sensor; GB =
Guided breathing; NB = Normal breathing; TEE = Technical error of estimate; CV =

Coefficient variation; CI = Confidence interval.
View publication stats

Comparison

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Comparison

Uploaded by

Copyright:

Available Formats

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

Comparison of Heart Rate Variability Recording With Smart Phone

Article in International journal of sports physiology and performance · March 2017

Daniel J Plews Ben Scott

SEE PROFILE SEE PROFILE

Marco Altini Matthew R. Wood

SEE PROFILE SEE PROFILE

Paul's PhD View project

The user has requested enhancement of the downloaded file.

Article Type: Original report

Technology, Auckland, New Zealand

Main text: 2745

Figures and tables: 2 Figures, 1 Tables

and ease of use in the field.

Keywords: Cardiac parasympathetic, monitoring, athletic performance

accelerometry, actigraphy and photoplethysmography can measure many aspects of human

system (ANS) activity.4

Several aspects can converge to reduce daily athlete measurement compliance,

representation of an athlete’s physiological state.5-7 As such, ways by which HRV recording

Photoplethysmography (PPG) is one technological advancement that may allow HRV

friendly model of HRV attainment, as no additional apparatus is required other than a

compliance via enhanced ease of daily recording.

standard” electrocardiography (ECG).

use an entrained breathing setting.

Data acquisition and processing

diagnostic quality 12-lead system (Cosmed, Quark T12x, USA).

rate monitor, which was fitted just below V6.

Photoplethysmography (PPG) was acquired via a commercially available smartphone

application known as “HRV4training” (see http://www.hrv4training.com/). Given the low

RR interval extraction, data synchronization, and features computation.

HRV4Training implements a peak detection algorithm to determine peak-to-peak

This technique avoids over-correcting, a problem of the widely employed removal of

causes beyond the scope of this comparison.

intervals, and therefore RR data does not require additional processing.

visual synchronization was necessary before HRV computation. Similalry, manual

the different systems at exactly the same time.

durations were investigated.

percentage of discarded RR intervals over a given time-period, as RR intervals are discarded

a Pearson product-moment correlation analysis, standard linear regression, typical error of

negative values, the magnitude of correlation was deemed ‘unclear’.

difference (PPG/Polar H7 vs. ECG), all differences were deemed “Trivial”.

were deemed “almost perfect”.

We have previously shown that in order to effectively monitor an athlete’s HRV,

almost a perfect correlation between PPG and ECG HRV recordings.

present study is further supported by the “trivial” standardised differences shown.

compliance is likely to be enhanced.

Marco Altini is the owner and developer of HRV4Training.

Eur J Appl Physiol 2012;112:3729-41.

15. Sacknoff D, Gleim G, Stachenfeld N, Glace B, Coplan N. Suppression of high-

Figure 1: Simultaneous R-R interval of an individual subject during 60 seconds of recording

1-min TEE 90% CI Std. 90% CI Qualitativ Mean 90% CI

Guided breathing; NB = Normal breathing; TEE = Technical error of estimate; CV =

View publication stats

You might also like