sensors

Article
Design and Development of a Novel Invasive Blood
Pressure Simulator for Patient’s Monitor Testing
Daniele Bibbo 1, * , Jan Kijonka 2 , Petr Kudrna 3 , Marek Penhaker 2 , Petr Vavra 4 and
Pavel Zonca 4
1 Department of Engineering, University of Roma Tre, Via Vito Volterra, 62, 00146 Rome, Italy
2 Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering and Computer
Science, VSB-Technical University of Ostrava, 17. listopadu 2172/15, 708 00 Ostrava–Poruba, Czech Republic;
kijonka84@gmail.com (J.K.); marek.penhaker@vsb.cz (M.P.)
3 Department of Biomedical Technology, Faculty of Biomedical Engineering, Czech Technical University in
Prague, nam. Sitna 3105, 272 01 Kladno, Czech Republic; petr.kudrna@fbmi.cvut.cz
4 Department of Internal Medicine, University Hospital in Ostrava, 17. Listopadu 1790, 70852 Ostrava, Czech
Republic; petr.vavra@fno.cz (P.V.); pavel.zonca@hotmail.co.uk (P.Z.)
* Correspondence: daniele.bibbo@uniroma3.it; Tel.: +39-06-5733-7298




Received: 18 November 2019; Accepted: 30 December 2019; Published: 1 January 2020


Abstract: This paper presents a newly-designed and realized Invasive Blood Pressure (IBP) device
for the simulation on patient’s monitors. This device shows improvements and presents extended
features with respect to a first prototype presented by the authors and similar systems available in
the state-of-the-art. A peculiarity of the presented device is that all implemented features can be
customized from the developer and from the point of view of the end user. The realized device has
been tested, and its performances in terms of accuracy and of the back-loop measurement of the
output for the blood pressure regulation utilization have been described. In particular, an accuracy
of ±1 mmHg at 25 ◦ C, on a range from −30 to 300 mmHg, was evaluated under different test
conditions. The designed device is an ideal tool for testing IBP modules, for zero setting, and for
calibrations. The implemented extended features, like the generation of custom waveforms and the
Universal Serial Bus (USB) connectivity, allow use of this device in a wide range of applications, from
research to equipment maintenance in clinical environments to educational purposes. Moreover,
the presented device represents an innovation, both in terms of technology and methodologies: It
allows quick and efficient tests to verify the proper functioning of IBP module of patients’ monitors.
With this innovative device, tests can be performed directly in the field and faster procedures can be
implemented by the clinical maintenance personnel. This device is an open source project and all
materials, hardware, and software are fully available for interested developers or researchers.

Keywords: invasive blood pressure; medical devices; exciter voltage; waveform simulation

1. Introduction
Direct Blood Pressure (BP) measurement, achieved by invasive techniques, is used in a specific
context when a continuous and reliable monitoring is needed. This method of blood pressure
measurement can provide detailed results, and it is mainly used for patients that need to be continuously
monitored (e.g., hospitalization in intensive care unit). The most common technique to obtain the BP
measurement is based on the use of a special catheter inserted into an artery, on which a pressure
transducer is fixed, and then the signal is transmitted by a cable inserted into the catheter itself.
These systems are generally made by a primary mechanical transducer to which one or more strain
gauges are fixed, in order to transduce the mechanical deformation into an electrical quantity. Usually,

Sensors 2020, 20, 259; doi:10.3390/s20010259 www.mdpi.com/journal/sensors

Sensors 2020, 20, 259 2 of 19

these transducers are connected to a patient’s monitor, in order to provide invasive blood pressure
information together with a plethora of vital parameters. A patient’s monitor, in order to be compliant
with certified standards for clinical environment, has to be tested before its installation, and it has to
undergo periodic maintenance to validate the correct functioning of all available features. To this aim,
simulators able to easily reply physiological signals, such as electrocardiography or body temperature,
are available for most monitoring systems. Concerning Invasive Blood Pressure (IBP) measurement, to
the best of our knowledge, very few attempts have been realized for simulating this signal and the
corresponding sensor, as demonstrated by a low number of papers in the corresponding literature. This
aspect shows the need for such a device to improve the knowledge in this field. Anyway, the simulation
of IBP may be useful for testing the functionality of the patient’s monitor IBP module, its zero and
calibration function, and function of alarms. Moreover, the possibility of simulating many types of
waveforms allows the use of this device in a wide range of applications, from research to equipment
maintenance in a clinical environment to educational purposes, which is a growing strategical sector
for the users of this kind of instruments [1–4].
The simulation of IBP on a patient’s monitor can be performed in two ways: The first is to
use a standard IBP transducer connected to some hydraulic or pneumatic system, simulating blood
pressure. In a simple case, this system may consist of a manual pump [5]. The second way is to use an
electronic circuit instead of a pressure transducer that allows generation of an output signal similar
to the IBP transducer’s output. This method does not require any mechanical parts, and both the
static and dynamic pressures can be simulated with the required bandwidth and high accuracy can be
achieved [6,7].
In this framework, the IBP simulator presented in this paper represents an innovation, both in
terms of technology and methodologies: currently, it is very difficult to realize a quick and efficient test
to verify the proper functioning of a patient’s devices features related to IBP measures, and the common
solution is to send these devices to maintenance technical services. Using the device proposed in this
paper, these tests can be performed directly in the field and faster procedures can be implemented by
clinical maintenance personnel.
A simple approach to the realization of these devices’ has been exploited in previous works [8–11],
where the basic design of an electronic IBP simulator was discussed. Moreover, some tests with
a patient’s monitor and a procedure for the calibration of the controlling circuits of such a kind of
device was presented in [12].
In this paper, some improvements and extended features of a newly-designed device are reported,
including all the new capabilities that guarantee better accuracy. Moreover, this new prototype allows
regulation of its output on the basis of both the output and the input exciter voltage measurements,
together with an auto-calibration feature implemented after power-on. To validate this, the accuracy
assessment of the device under different conditions was performed and results are described in this
paper. This is a crucial aspect since the evaluation of the accuracy is necessary when dealing with
systems used to obtain information on human performances or on the health state of a patient [13].
Finally, it is worth highlighting that the possibility of sharing information with a community of
developers is an important feature that can improve the knowledge in the development of a system
or a device. This allows us to overcome problems and find smarter solutions. To this aim, this
device was designed as an open source project, in order to share materials with interested researchers
and developers.

2. Design and Implementation

The IBP Simulator (IBPS) can simulate a bridge pressure transducer, which consists of a two-port
network with input voltage VIN = (+IN ) − (−IN ) and output voltage VOUT = (+OUT ) − (−OUT ).
An example is reported in Figure 1, where a schematic diagram for a commercial pressure transducer
(i.e., NovaSensor NPC-100 series @Amphenol) is reported [14].
 Sensors 2020, 20, 259 3 of 19
Sensors2020,
Sensors 2020,20,
20,xx 33of
of19
19

Figure 1. NovaSensor NPC-100 pressure transducer schematic diagram.

Figure1.
Figure 1.NovaSensor
NovaSensorNPC-100
NPC-100pressure
pressuretransducer
transducerschematic
schematicdiagram.
diagram.
The device has been designed in order to simulate a wide range of IBP modules; this is possible
Thethe
The
because device
device hasinterface
has
analog beendesigned
been designed inorder
in
is driven order
by tosimulate
to
a digitalsimulate
system aawide
widecan
that range
range ofIBP
of
controlIBP modules;
modules;
a wide this
rangethis isispossible
possible
of parameters.
because
because theanalog
the
The general analog interface
schemeinterface isisdriven
of the designeddrivenIBPS
byaaisdigital
by digital system
system
reported thatcan
that
in Figure can control
control
2. The aawide
wide
quality range
range
of the ofparameters.
of parameters.
performed design
The
The general scheme of the designed IBPS is reported in Figure 2. The quality
is guaranteed by the adoption of well-established procedures for embedded applicationsdesign
general scheme of the designed IBPS is reported in Figure 2. The quality of
of the
the performed
performed design
[15,16]
isis guaranteed
inguaranteed by the
by
different fields, thesuch
adoption
adoption of well-established
of well-established
as electronic circuits [17], procedures
procedures for embedded
for
transmission embedded applications
applications
of data [18], [15,16] in
[15,16]
and consumption in
different
different
optimization fields,
fields, such as
[19].such as electronic
electronic circuits
circuits [17],
[17], transmission
transmission of of data
data [18],
[18], andand consumption
consumption
optimization[19].
optimization [19].

Figure 2. Invasive Blood Pressure Simulator (IBPS) scheme of functioning. ADC = analog to digital
Figure 2.
Figure 2. Invasive
Invasive Blood
Blood Pressure
Pressure Simulator
Simulator (IBPS)
(IBPS) scheme
scheme of
of functioning.
functioning. ADC
ADC == analog
analog to
to digital
digital
converter; LCD = liquid crystal display; MCU = microcontroller unit.
converter;LCD
converter; LCD==liquid
liquidcrystal
crystaldisplay;
display;MCU
MCU==microcontroller
microcontrollerunit.
unit.
Three controlling stages have been designed in order to drive the output for the simulation of
Three controlling
Three
the differentcontrolling stages
stages
IBP modules. have
have
This been
canbeen designed
designed
be obtained in
byin order to
order to drive
controlling drive the output
the
the range output for
and thefor the simulation
the
positive simulation of
of
or negative
the
the different
different IBP
IBP modules.
modules. This
This can
can be
be obtained
obtained by
by controlling
controlling the
the range
range and
and the
the positive
positive
offset, as reported in the following. In general, the input exciter voltage VIN can vary for different or
or negative
negative
offset,
offset, asreported
as
IBP modulesreported inthe
in thefollowing.
but should following.
be in the In Ingeneral,
general,
range the
1–10the input
input
VDC; excitervalue
exciter
a typical voltage
voltage for𝑉𝑉a wide
canvary
can vary
range for
for different
ofdifferent IBP
IBP
IBP module
modules
modules but
but should
should be
be in
in the
the range
range 1–10
1–10 VDC;
VDC; a
a typical
typical value
value for
for aa wide
wide
is VIN = 5 VDC. The output voltage VOUT depends linearly on VIN . Considering as an example range
range of
of IBP
IBP module
module isis
𝑉 = 5 VDC.
𝑉 = 5 VDC. The output voltage 𝑉
The output voltage 𝑉 depends linearly on 𝑉
depends linearly on 𝑉 . Considering as an example an
. Considering as an example an
Sensors 2020, 20, x 4 of 19
Sensors 2020, 20, 259 4 of 19
Sensors 2020, 20, x 4 of 19
exciter voltage 𝑉 = 1 VDC, a transducer sensitivity 𝑘 = 5 µV/V/mmHg, and a pressure applied to
the transducer
an exciter of 1VmmHg,
voltage an output
IN = 1 VDC, of 5 µV iskobtained.
voltage sensitivity
a transducer = 5 µV/V/mmHg, and a pressure applied
exciter voltage 𝑉 = 1 VDC, a transducer sensitivity 𝑘 = 5 µV/V/mmHg, and a pressure applied to
to theThe transducer
transducer mmHg, an 𝑘
of 1 sensitivity can vary
output voltage 5 µV is obtained.
forofdifferent pressure transducers, even if the most
the transducer of 1 mmHg, an output voltage of 5 µV is obtained.
common 5 µV/V/mmHg
values are sensitivity
The transducer k can or
vary40forµV/V/mmHg.
different pressure transducers, even if the most common
The transducer sensitivity 𝑘 can vary for different pressure transducers, even if the most
valuesThe 5 µV/V/mmHg
aresimulator or 40 µV/V/mmHg.
output circuit (Figure 3) is obtained by means of a Wheatstone bridge driven by
common values are 5 µV/V/mmHg or 40 µV/V/mmHg.
threeThe
independent
simulator controlling
output circuitstages. Since
(Figure 3) isthe Wheatstone
obtained bridge
by means of aoutput is usually
Wheatstone bridge characterized by
driven by three
The simulator output circuit (Figure 3) is obtained by means of a Wheatstone bridge driven by
very low voltage
independent values,stages.
controlling in eachSince
stagethea controllable
Wheatstone current sourceisisusually
bridge output used (Figure 4). The by
characterized output
very
three independent controlling stages. Since the Wheatstone bridge output is usually characterized by
current
low voltagefromvalues,
these type of sources
in each is linearly dependent
stage a controllable on the
current source input(Figure
is used exciter4).
voltage 𝑉 from
The output the
current
very low voltage values, in each stage a controllable current source is used (Figure 4). The output
IBP module. Moreover, two measuring stages have been designed to control 𝑉
from these type of sources is linearly dependent on the input exciter voltage VIN from the IBP module. and 𝑉 : the
current from these type of sources is linearly dependent on the input exciter voltage 𝑉 from the
measurement
Moreover, twoofmeasuring
these two stages
voltagehave
levels is necessary
been designed to improve
control Vthe accuracy
IN and VOUT :ofthe
themeasurement
device. In fact,
of
IBP module. Moreover, two measuring stages have been designed to control 𝑉 and 𝑉 : the
the regulation
these two voltageof the output
levels and its fine
is necessary tuning are
to improve theperformed through
accuracy of a feedback
the device. In fact,loop
the control.
regulation of
measurement of these two voltage levels is necessary to improve the accuracy of the device. In fact,
the output and its fine tuning are performed through a feedback loop control.
the regulation of the output and its fine tuning are performed V through a feedback loop control.
IN

VININ
R
CR
IN
R R
+VOUT A C B -VOUT
R R
+VOUT ROUT A B ROUT -VOUT
R R
ROUT D ROUT
R RINR
D
RIN

Figure 3. IBPS output circuit.

Figure 3. IBPS
Figure 3. IBPS output
output circuit.
circuit.

Figure
Figure 4.
4. Controllable
Controllable current
current source
source circuit.
circuit.

2.1. Controlling
Controlling Stage
Stage 11 Figure 4. Controllable current source circuit.
2.1.
This stage
This stage is
2.1. Controlling is essential
essential
Stage 1 for the
for the generation
generation of of dynamic
dynamic BPBP waveforms;
waveforms; it
it consists
consists of
of aa range
range setting
setting
block, a buffer, a modulation block, and a current source (Figure 4). The blood pressure
block, a buffer, a modulation block, and a current source (Figure 4). The blood pressure range can range can be
be
Thisfrom
adjusted stage0istoessential
330 mmHg for the generation
with 10-bit of dynamic
resolution. The BP waveforms;
waveform it consists
modulation is of a range setting
performed using
adjusted from 0 to 330 mmHg with 10-bit resolution. The waveform modulation is performed using
block, a buffer, a modulation block,
anyand a current source (Figure 4). The blood pressure range can be
aa 10-bitcontroller
10-bitcontroller that is
that is used
used forfor any pressure
pressure range
range setting.
setting.
adjusted from
The current 0 to 330
current source mmHg
source calculation with
calculation for 10-bit resolution. The waveform modulation is performed using
The for aa blood
blood pressure
pressure of:
of:
a 10-bitcontroller thatD is used for any pressure range setting.
𝐵𝑃∈ ∈0,〈0,calculation
The currentBPsource 330mmHgmmHg〉for considering V ==of:
a blood pressure 1 1VDC,
VDC,k𝑘== 5uV/V/mmHg
5µV/V/mmHg

330 considering VIN
It is obtained as: 𝐵𝑃 ∈ 〈0, 330 mmHg〉 considering V = 1 VDC, 𝑘 = 5µV/V/mmHg
It is obtained as:
It is obtained as: |V|VOUT| =| =
k k×× VVIN ×
× |BP MAX ||=
|BP =55××11××330
330==1650
1650µV,µV, (1)
(1)
|V current
so the maximum output | = k is:× V × |BP | = 5 × 1 × 330 = 1650 µV, (1)
so the maximum output current is:
Sensors 2020, 20, 259 5 of 19

so the maximum output current is:

2|VOUT | 2 × 1650 × 10−6

|IOUT | = = = 33 µA, (2)
R 100
where R is the Wheatstone bridge resistor.
The maximum reference voltage VREF for the selected R1 = 6.8 kΩ, where R1 is a current source
resistor, is given by:

VREF = |IOUT | × R1 = 33 × 10−6 × 6.8 × 103 = 0.2244 V. (3)

After these values are computed, it is possible to obtain the theoretical R3 value when the digital
potentiometer is set with R2 = 20 kΩ:

VIN − VREF 1 − 0.2244

R3TH = × R2 = × 20 × 103  69 kΩ. (4)
VREF 0.2244

This value is not available among commercial precision resistors. For this reason, the closest
available one R = 68 kΩ, was selected. Using this value, the maximum output pressure for R3 ≡
R3real = 68 kΩ (E12 series) is re-calculated considering the obtained parameters:

VREF, MAX = 0.2272 V, (5)

  VREF, MAX
IOUT, MAX  = = 33.4 uA, (6)
R1
 
  IOUT, MAX  × R
VOUT, MAX  = = 1670.5 uV (334 mmHg), (7)
2
which gives a maximum blood pressure correspondent to the maximum digital potentiometer output
VREF − LSB (where LSB is the Least Significant Bit, and its value is LSB = 0.326 mmHg)

BPMAX = 333.7 mmHg. (8)

The current source is connected to the B node of the Wheatstone bridge for the positive output
(Figures 2 and 3). VOUT is linearly dependent on any exciter voltage VIN according to the relation:

R × R2
VOUT = VIN . (9)
2R1 (R2 + R3 )

The range setting block controls the Reference Voltage in the range from 0 V to VREF (VREF = 0.2244 V
given by Equation (3)).

2.2. Controlling Stage 2

This stage is designed to obtain a positive static pressure and to adjust the diastolic pressure.
It consists of a positive offset block and a current source. The positive static blood pressure can be
adjusted in the range from 0 to 330 mmHg with 10-bit resolution. The calculation of the current source
and the realization of this stage is the same as the one for controlling stage 1.

2.3. Controlling Stage 3

Similarly to controlling stage 2, a further stage 3 is included to set a negative output. In this case,
the current source is connected to the A node of the Wheatstone bridge (Figures 2 and 3). The negative
static blood pressure can be adjusted in the range from −60 mmHg to 0 mmHg with 10-bit resolution.
Sensors 2020, 20, 259 6 of 19

2.4. Measuring Stage 1 and 2

The first measuring stage (i.e., measuring stage 1) was been designed and implemented to
continuously monitor the output voltage VOUT of the device. VOUT is amplified by an accurate
instrumentation amplifier INA101HP (Texas Instruments, USA) with gain G = 201. After amplification,
an offset voltage of 1 V is added in order to fit the signal into the A/D converter range, avoiding possible
saturation problems. This signal is digitalized by CH1 channel of the 16-bit analog to digital converter
(ADC) with the common reference ADCREF = 2.5 V.
A second measuring stage (i.e., measuring stage 2) has been included to monitor the input exciter
voltage VIN that differently from VOUT is a constant value. The blood pressure can be calculated from
VIN , VOUT and k as:
Sensors 2020, 20, x VOUT 6 of 19
BP = . (10)
k × VIN
A second measuring stage (i.e., measuring stage 2) has been included to monitor the input exciter
VIN is divided
voltage by
𝑉 athat
factor of 2.5 from
differently before
𝑉 its is
digitalization byThe
a constant value. CH0 channel
blood pressureofcan
thebeADC.
calculated from
𝑉 , 𝑉 and 𝑘 as:
2.5. Digital Control
𝐵𝑃 = . (10)
×
The management of the device is performed by a central microcontroller unit (MCU, Atmel
𝑉 is divided by a factor of 2.5 before its digitalization by CH0 channel of the ADC.
ATMEGA644A). It drives the digital potentiometers in the controlling stages and manages the
acquisitions of VINControl
2.5. Digital and VOUT . Both potentiometer sets and ADC are connected to the MCU via Serial
Peripheral Interface (SPI), which
The management of theallows easy
device is and reliable
performed communication
by a central among
microcontroller the different
unit (MCU, Atmel blocks
of the device. Moreover,Itthe
ATMEGA644A). device
drives is equipped
the digital with a Universal
potentiometers Serial Bus
in the controlling (USB)
stages and port thatthe
manages is used for
of 𝑉 and 𝑉
multipurpose tasks: to store custom waveforms, to perform calibration data storage, and to control the
acquisitions . Both potentiometer sets and ADC are connected to the MCU via Serial
Peripheral Finally,
device remotely. Interface (SPI), which
a flash allows easy
memory andand
thereliable communication
necessary among the
control circuits different
were blocks on the
embedded
of the device. Moreover, the device is equipped with a Universal Serial Bus (USB) port that is used
device to store data.
for multipurpose tasks: to store custom waveforms, to perform calibration data storage, and to
All these
controlfeatures
the device canremotely.
be managed directly
Finally, a flash onboard
memory andby means of fourcontrol
the necessary navigation
circuitsbuttons,
were while
commands and data
embedded aredevice
on the displayed
to storeon an embedded liquid crystal display (LCD) screen.
data.
All these features can be managed directly onboard by means of four navigation buttons, while
3. Features
commands and data are displayed on an embedded liquid crystal display (LCD) screen.

As mentioned
3. Features above, the IBPS (Figure 5) is equipped with a user interface implemented using
four control As buttons and a 4-row, 20-character LCD screen (no. 1 in Figure 5). The four control
mentioned above, the IBPS (Figure 5) is equipped with a user interface implemented using
buttons are divided in oneand
four control buttons navigation button (no.
a 4-row, 20-character LCD 2 screen
in Figure
(no. 15)
in used
Figureto
5).control
The fourthe cursor
control in different
buttons
navigationaremenu,
dividedtwo setting
in one buttons
navigation (no.(no.
button 3 in2Figure 5) 5)
in Figure forused
setting parameters,
to control andinone
the cursor enter button
different
navigation
(n.4 in Figure menu,
5) used to two
set setting buttonsfunction.
the selected (no. 3 in Figure 5) for setting parameters, and one enter button
(n.4 in Figure 5) used to set the selected function.

Figure 5.Figure
IBPS5.user
IBPS interface.
user interface.
On Onthe
thetop
top are
arereported
reported(1) an
(1)LCD display
an LCD (2) a navigation
display button, (3) button,
(2) a navigation
two setting buttons, and (4) an enter button.
(3) two setting buttons, and (4) an enter button.
On the front side of the IBPS there are two connectors (Figure 6). The 4-pin output connector
(no. 2 in Figure 6) is used for the connection to the IBP module input of the patient monitor. The
amplified output signal is also available on a BNC connector (no. 1 in Figure 6).
Sensors 2020, 20, 259 7 of 19

On the front side of the IBPS there are two connectors (Figure 6). The 4-pin output connector (no. 2
in Figure 6) is used for the connection to the IBP module input of the patient monitor. The amplified
output signal
Sensors 2020, 20, xis also available on a BNC connector (no. 1 in Figure 6). 7 of 19
Sensors 2020, 20, x 7 of 19

Figure 6.
Figure 6. (1) Scope connector for signal monitoring and
and (2)
(2) 4-pin
4-pin connector
connector for
for IBP
IBP monitor.
monitor.
Figure 6. (1) Scope connector for signal monitoring and (2) 4-pin connector for IBP monitor.
On
On the
the backside
backside of
of the
the IBPS,
IBPS, an
an embedded
embedded USB
USB Mini-B
Mini-B connector
connector isis provided
provided (no.
(no. 3 in Figure 7),
On the backside of the IBPS, an embedded USB Mini-B connector is provided (no. 3 in Figure 7),
and
and this is used to connect the device to a computer to exploit all software-driven features
this is used to connect the device to a computer to exploit all software-driven features (e.g.,
(e.g., to
to
and this is used to connect the device to a computer to exploit all software-driven features (e.g., to
upstream
upstream data of different waveforms),
waveforms), together
together with
withthe
thepower
powerswitch
switchandanda aconnector
connector(no.
(no.1,31,3
in
upstream data of different waveforms), together with the power switch and a connector (no. 1,3 in
in Figure
Figure 7) 7)
to to supply
supply thethe device
device with
with a standard
a standard 230V
230V alternating
alternating current
current (AC).
(AC).
Figure 7) to supply the device with a standard 230V alternating current (AC).

Figure 7. (1) Power switch, (2) alternating current (AC) power connector, and (3) Universal Serial Bus
Figure 7. (1) Power switch, (2) alternating current (AC) power connector, and (3) Universal
Universal Serial
Serial Bus
(USB) mini-B connector.
(USB) mini-B
(USB) mini-B connector.
connector.

3.1. Static
3.1. Static Blood
Blood Pressure Simulation
Simulation
3.1. Static Blood Pressure
Pressure Simulation
This feature can
This can be used
used for zero
zero setting and
and for calibrating
calibrating thethe patient’s
patient’s monitors
monitors IBP
IBP modules.
modules.
This feature
feature can bebe used forfor zero setting
setting and for
for calibrating the patient’s monitors IBP modules.
The output pressure can be regulated from a minimum negative value of −30 mmHg to a maximum
The
The output
output pressure
pressure can
can be
be regulated
regulated from
from aa minimum
minimum negative
negative value of−30
value of −30 mmHg
mmHg to to aa maximum
maximum
positivevalue
positive valueof of300
300 mmHg,
mmHg,with with increments
incrementsof of11 mmHg.
mmHg. The The output
output pressure
pressure accuracy
accuracy ofof the device
device
positive value of 300 mmHg, with increments of 1 mmHg. The output pressure accuracy of the the device
is less than
than ±1 mmHg
mmHg at at 25
25 ◦°C on the
the fixed value.
value. The
The measured
measured output voltagevoltage and input
input excitation
is less than ±1
is less ±1 mmHg at 25 °C C on
on the fixed
fixed value. The measured output
output voltage and and input excitation
excitation
voltage
voltage is also displayed on the LCD to allow the users to easily manage these parameters. The four
voltage is also displayed on the LCD to allow the users to easily manage these parameters. The
is also displayed on the LCD to allow the users to easily manage these parameters. The four
four
push
push buttons are used for the menu navigation, blood pressure adjustment, and confirmation of the
push buttons
buttons areare used
used for
for the
the menu
menu navigation,
navigation, blood
blood pressure
pressure adjustment,
adjustment, andand confirmation
confirmation of of the
the
selected value.
selected
selected value.
value.
3.2. Dynamic Blood Pressure
3.2. Pressure Simulation
Simulation
3.2. Dynamic Blood Pressure Simulation
The device
The device can
can simulate
simulate dynamic
dynamic BP BP waveforms;
waveforms; thisthis feature
feature isis suitable
suitable for
for many
many physiological
physiological
The device can simulate dynamic BP waveforms; this feature is suitable for many physiological
and pathological BP waveforms
and pathological BP waveforms simulation (e.g., arterial, left and right ventricles, pulmonary artery,
and pathological BP waveforms simulation (e.g., arterial, left and right ventricles, pulmonary artery,
rightatrium).
right atrium). To
Toextend
extendthe
thepossibilities
possibilitiesofofthe
thesimulation,
simulation, there
there is is a USB
a USB upstream
upstream port,
port, allowing
allowing us us
to
right atrium). To extend the possibilities of the simulation, there is a USB upstream port, allowing us
to store
store up toup15to 15 user-defined
user-defined blood pressure
blood pressure waveforms.waveforms.
Moreover, Moreover, it istopossible
it is possible simulateto simulate
respiration
to store up to 15 user-defined blood pressure waveforms. Moreover, it is possible to simulate
respiration
and and other physiological
other physiological parameters. parameters.
respiration and other physiological parameters.
To implement these features, the user can select one of the waveforms displayed in a list on the
To implement these features, the user can select one of the waveforms displayed in a list on the
LCD screen. Thus, the waveform details are reported, and the user can accept them or set custom
LCD screen. Thus, the waveform details are reported, and the user can accept them or set custom
waveform properties. The systolic, diastolic, and mean pressures can be independently set with a
waveform properties. The systolic, diastolic, and mean pressures can be independently set with a
resolution of 1 mmHg for each step. In addition, the simulation of the heart beats per minute can be
resolution of 1 mmHg for each step. In addition, the simulation of the heart beats per minute can be
regulated with a resolution of 1 BPM. The required BP value and the actual BP value are both visible
regulated with a resolution of 1 BPM. The required BP value and the actual BP value are both visible
Sensors 2020, 20, 259 8 of 19

To implement these features, the user can select one of the waveforms displayed in a list on the
LCD screen. Thus, the waveform details are reported, and the user can accept them or set custom
waveform properties. The systolic, diastolic, and mean pressures can be independently set with
a resolution of 1 mmHg for each step. In addition, the simulation of the heart beats per minute can be
regulated with a resolution of 1 BPM. The required BP value and the actual BP value are both visible
on the2020,
Sensors LCD. 20, x 8 of 19

3.3. Custom Waveforms

3.3. Custom Waveforms and and Usage
Usage ofof Built-in
Built-in Memory
Memory
Custom waveforms can
Custom waveforms can be
be uploaded
uploaded to to the
the device via USB
device via USB port,
port, considering one period
considering one period of of the
the
desired signal. The IBPS update tool, shown in Figure 8, allows to program
desired signal. The IBPS update tool, shown in Figure 8, allows to program a waveform in one of the a waveform in one of
the device’s empty positions in the memory. Each waveform is characterized
device’s empty positions in the memory. Each waveform is characterized by a specific length (as by a specific length
(as number
number of samples)
of samples) and and sampling
sampling period.period. The parameters
The parameters of theof the uploaded
uploaded waveform,
waveform, such assuch as
name,
name, BPM, and systolic, diastolic, and mean pressure, must be specified by
BPM, and systolic, diastolic, and mean pressure, must be specified by the user. These values are the user. These values are
included
included at at the
the beginning
beginning of of the
the data
data waveform
waveform file
file as
as initialization
initialization bytes. The waveform
bytes. The waveform data
data can
can be
be
imported
imported fromfromdatadatafilefile
format
format(e.g.,(e.g.,
CSV,CSV,
TXT, MAT). Each data
TXT, MAT). Eachwaveform is automatically
data waveform converted
is automatically
to 10-bit resolution
converted and normalized
to 10-bit resolution in the range
and normalized infrom 0 to 1023.
the range from 0 to 1023.
The update tool also provides to program raw data
The update tool also provides to program raw data in a specified in a specified memory
memory location
location of
of the
the device.
device.
This
This feature can be useful for future memory usage, such as for calibration data. Backup memory to
feature can be useful for future memory usage, such as for calibration data. Backup memory to
aa data
data file
file and
and erase
erase operations
operations are are also
also possible.
possible.

(a) (b)
Figure 8.
Figure 8. (a)
(a)The
TheIBPS update
IBPS tool,
update showing
tool, IBP IBP
showing waveform parameters
waveform (for one
parameters (forperiod) such assuch
one period) name,as
beats per
name, minute
beats (BPM), (BPM),
per minute and systolic, diastolic,
and systolic, and mean
diastolic, pressure.
and (b) An example
mean pressure. of the obtained
(b) An example of the
waveform.
obtained waveform.

3.4. Remote-Control
3.4. Remote-Control
The
The device
device can
can be controlled remotely
be controlled remotely by
by aa PC
PC through
through the
the USB
USB port. Even if
port. Even the device
if the device has
has been
been
designed
designed andandrealized for afor
realized stand-alone use, anuse,
a stand-alone additional remote-control
an additional option has been
remote-control implemented
option has been
to allow the device integration in a computer-controlled environmental. The remote
implemented to allow the device integration in a computer-controlled environmental. The remotecontrol allows
the same operation as in stand-alone mode. Moreover, waveforms of any length can be
control allows the same operation as in stand-alone mode. Moreover, waveforms of any length can simulated in
real-time via serial
be simulated data communication.
in real-time via serial data communication.
3.5. Power-on Self-Test and Auto Calibration
3.5. Power-on Self-Test and Auto Calibration
After connecting the IBP module to the device, the device provides the necessary input excitation
After connecting the IBP module to the device, the device provides the necessary input excitation
voltage VIN to generate an appropriate output voltage VOUT according to the Equation (10). When
voltage V to generate an appropriate output voltage V according to the Equation (10). When
the device is powered on, VIN and VOUT are measured automatically by the measuring stages 1 and
the device is powered on, V and V are measured automatically by the measuring stages 1 and
2 (Figure 2) to allow different settings of the control stages.
2 (Figure 2) to allow different settings of the control stages.
The following parameters are evaluated by the auto calibration process:
The following parameters are evaluated by the auto calibration process:
• BP offset (controlling stages powered down);
• BP offset (controlling stages powered down);
• BP range (controlling stage 3 at its maximum);
• BP positive range (controlling stage 2 at its maximum);
• BP range offset (controlling stage 1 range at its maximum, modulation at its minimum);
• BP modulation minimum range (controlling stage 1 range at its minimum, modulation at its
Sensors 2020, 20, 259 9 of 19

• BP range (controlling stage 3 at its maximum);

• BP positive range (controlling stage 2 at its maximum);
• BP range offset (controlling stage 1 range at its maximum, modulation at its minimum);
• BP modulation minimum range (controlling stage 1 range at its minimum, modulation at its
maximum);
• BP modulation maximum range (controlling stage 1 range at its maximum, modulation at
its maximum);
• BP modulation half range (controlling stage 1 range at the half, modulation at its maximum).

All these parameters are evaluated and displayed on LCD at power-on. If one or more of these
are not in the desired interval, an error symbol indicates that the device did not pass the power-on
self-test. These parameters, which represent the input values for calculation of the measured BP, are
used to remove the BP offset caused by the measurement stages.

3.6. Programming the Device

The firmware can be upgraded using an In-System Programming (ISP) feature by a firmware
upload connector available inside the device. The program is written in AVR GCC language for
ATMEL MCU’s and consists of several files.

3.7. Technical and Simulation Parameters

The technical parameters of the IBPS, obtained on the basis of the performed tests on the device,
are summarized in Table 1.

Table 1. IBPS Technical Parameters.

Technical Parameter Value Unit Description

Input impedance 3700 Ω
Output impedance 300 Ω
Exciter Input voltage VIN range 1 to 10 VDC
Output Transducer Sensitivity 5 µV/V/mmHg Constant parameter
Output pressure range −30 to 300 mmHg
Output Offset Range −25 to 25 mmHg
−150 to 1500 µV
Output voltage VOUT range Equivalent to −30 to
−750 to 7500 µV
(VIN = {1; 5; 10} VDC) 300 mmHg
−1500 to 15,000 µV
Output voltage offset range Equivalent to −25 to
−625 to 625 µV
at VIN = 5 VDC 25 mmHg
Maximum difference between the desired 0.4 Displayed BP (BP errors
mmHg
and actual BP (0.1 for static BP) not included)
Test at the normal
Accuracy for ambient temperature 1 25 ◦ C 0.4 mmHg
ambient temperature
Maximum value in the
Maximum accuracy for ambient
1.4 mmHg complete range
temperature 1 from 0 ◦ C to 50 ◦ C
of temperature
BP offset for ambient temperature 1 25 ◦ C 0.2 mmHg Compensated by calibration
Maximum value in the
BP offset for ambient temperature 1 from
12 mmHg complete range
0 ◦ C to 50 ◦ C
of temperature
1. For the full input voltage range (Vin ={1; 5; 10} VDC) and all modes of blood pressure generating.

In particular, output and input impedance, excitation voltage VIN , transducer sensitivity and the
corresponding output voltage VOUT match the standard IBP transducer (as reported in [12]).
The maximum difference between the required and measured values of BP is given by the
controlling stages resolution and the noise (see Tables 3, 4 and 5). The accuracy of the device was
evaluated through experimental tests that are described in the following. More specifically, tests
Sensors 2020, 20, 259 10 of 19

were performed with the calibrated device (after offset and gain errors compensation) for ambient
temperature 25 ◦ C and for ambient temperature ranging from 0 ◦ C to 50 ◦ C when the offset was not
compensated. The reported data show that the accuracy is not constant over the entire temperature
range, and it is mainly influenced by non-linearities and gain errors, as reported in the test section of
this work. However, a temperature change of 25 ◦ C (i.e., from 25 ◦ C to 50 ◦ C) causes a voltage offset and
the minimum value is obtained for Vin = 1VDC, and this corresponds to an error of EoffsetTEMP = 0.84 ±
0.3 mmHg, that is EoffsetTEMPMAX = 1.14 mmHg. Moreover, the maximum input voltage measurement
error is 0.6 mV, which gives a further error of EoffsetIV = 0.18 mmHg (computed by Equation (10)), so
a total estimate of the error is Eoffset = EoffsetTEMPMAX + EoffsetIV = 1.32 mmHg (approximated in Table 1
as Eoffset = 1.4 mmHg). Anyway, it is important to highlight that errors connected to temperature drift
can be compensated adding temperature sensors, as it will be discussed in the final part of this work.
Since higher values of Vin produce lower values of BP error readings, a better accuracy in these cases is
obtained. The simulation parameters of the IBPS are summarized in Table 2. In particular, the possible
pressure simulation waveforms are shown together with the simulation features described above.

Table 2. IBPS simulation parameters.

Simulation Value Unit Description

Static Pressure −30 to 300 mmHg step 1 mmHg
Static Offset −25 to 25 mmHg step 1 mmHg
Sine 0–100 Hz other signal shapes optional
Sample Rate
250–20,000 Hz /
(waveform streaming)
Output Resolution 10 bit higher resolution (16-bit optional)
a real biological waveform with
BioSim Mode / /
adjustable SYS/DIA and heart rate
Biological Waveforms,
user programmable Biological
with embedded samples / /
Waveforms via USB interface optional
from Physionet website

4. Test and Validation

An accuracy analysis was carried out for evaluating the performances of the output circuit for
each of the controlling stages and of the measuring stages. The measurement of the device’s output
was achieved using a precision Instrumentation Amplifier (IA) INA131 (Texas Instruments, Dallas,
TX, USA) with a fixed gain G = 100 for a low level bridge VOUT output voltage measurement and
a precision digital multimeter (DMM) 34401A (Agilent Technologies, Inc., Santa Clara, CA, USA).
The measurement errors were evaluated for all the range of VIN and all the range of VOUT for the
ambient temperature of 25 ◦ C and of 50 ◦ C. The maximum instrument errors (INA131 + DMM) were
computed and included in the time trends reported in the following. The output noise measurement
was performed using INA131 and U2702A scope (Agilent Technologies, Inc., Santa Clara, CA, USA).

4.1. The Accuracy of Output Circuit

4.1.1. Unbalance
The accuracy of the device’s output circuit depends on the Wheatstone bridge unbalance and the
values of its resistors. Considering a tolerance of 0.1% for the used resistors, the maximum Wheatstone
bridge unbalance is 15 µV/V. This value corresponds to 3 mmHg and can be easily compensated.
The comparison of the calculated and the measured unbalance is reported in Table 3.
Sensors 2020, 20, 259 11 of 19

Sensors 2020, 20, x 11 of 19

Table 3. Output circuit accuracy.
4.1.2. Drift and Interference
Component Maximum Unbalance Unbalance Measured Noise Measured
The temperature
Wheatstone bridge drift is in 15 general
µV/V compensated by2.15 theµV/V
Wheatstone bridge circuit; 6 µV although its
resistors have the same initial resistance values and are at the same temperature, small temperature
variations of individual resistors or similar interferences can cause large changes in the output. To
4.1.2. Drift and Interference
minimize the drift and errors due to interference, resistors with a low temperature drift in the
The temperature
Wheatstone drift isbe
bridge should in used,
generalandcompensated by theshould
the output circuit Wheatstone bridge circuit;
be separated from thealthough
main board its ,
resistors
also by have
usingthe same initial
a specific resistance
shield, as should values andfor
be done areallatsensitive
the samecircuits
temperature, small temperature
of the IBPS.
variations of individual resistors or similar interferences can cause large changes in the output.
To4.1.3. Noise the drift and errors due to interference, resistors with a low temperature drift in the
minimize
Wheatstone bridge should be used, and the output circuit should be separated from the main board,
The results of tests for the evaluation of the output noise, including the measurement noise
also by using a specific shield, as should be done for all sensitive circuits of the IBPS.
(Instrumentation Amplifier + scope), for all input voltages V on a single device are reported in
Table
4.1.3. 3. The maximum calculated value of the Wheatstone bridge unbalance, considering a tolerance
Noise
of 0.1% for the bridge resistors as reported above, gives a theoretical value of 15 µV/V. The real
The results
measured valueofistests
2.15 for theand
µV/V, evaluation
it resultsoftothe output noise,
be smaller includinglimit;
than theoretical the measurement noise
this is due to the fact
(Instrumentation Amplifier + scope), for all input voltages V on a single device
that the declared tolerance of resistors is always the maximalINtheoretical, but in real applications,are reported in it
Table
can 3.
beThe maximum
lower, calculated
thus obtaining value of
benefits in the
the Wheatstone
performances. bridge unbalance, considering a tolerance of
0.1% for the bridge resistors as reported above, gives a theoretical value of 15 µV/V. The real measured
value is 2.15 µV/V, and it results to be smaller
Table thancircuit
3. Output theoretical limit; this is due to the fact that the
accuracy.
declared tolerance of resistors is always the maximal theoretical, but in real applications, it can be
Component Maximum Unbalance Unbalance Measured Noise Measured
lower, thus obtaining benefits in the performances.
Wheatstone bridge 15 µV/V 2.15 µV/V 6 µV
4.2. The Accuracy of Controlling Stages
4.2. The Accuracy of Controlling Stages
4.2.1. Nonlinearity
4.2.1. Nonlinearity
Considering the control stage 1, a linear output is strongly desirable, but the output linearity
is related to the digital
Considering potentiometer
the control stage 1, linearity. In the isnonlinearity
a linear output evaluation,
strongly desirable, other
but the causes
output can be is
linearity
neglected
related toas the
a general
digitalhypothesis.
potentiometer The integral
linearity.nonlinearity of the digital
In the nonlinearity potentiometer
evaluation, used for
other causes canthebe
modulation
neglected asis ±1,5 LSB. The
a general blood pressure
hypothesis. nonlinearity
The integral error of
nonlinearity calculated forpotentiometer
the digital any VIN is given
used by:for the
modulation is ±1,5 LSB. The blood pressure nonlinearity error calculated for any V is given by:
POTNONLINEARITY ±1.5
BPNONLINEARITY = =± . = 0.147% of range. (11)
𝐵𝑃 = POTRES = 210= 0.147% of range. (11)

The
Theoutput
outputnonlinearity
nonlinearitywas wasalso
alsomeasured
measuredfor foreacheachstep
stepofofthe
thecontrolling
controllingstage
stage1 1digital
digital
potentiometer; for each of its possible values (i.e., n = 2 10 ), the BP nonlinearity was calculated,
potentiometer; for each of its possible values (i.e., n = 2 ), the BP nonlinearity was calculated, and
10 and the
the
possible
possiblevalues distributions
values distributionsareare
reported in Figure
reported 9. The
in Figure obtained
9. The maximum
obtained maximum valuevalue
corresponds to 0.1to
corresponds
mmHg that represents
0.1 mmHg a maximum
that represents nonlinearity
a maximum of 0.03%
nonlinearity on theon
of 0.03% full
therange of simulated
full range BP. BP.
of simulated

Figure
Figure 9. 9. IBPS
IBPS controlling
controlling stage
stage 1 output
1 output nonlinearity
nonlinearity measurement.
measurement.

4.2.2. Drift
Although the output voltage is continuously measured and it is regulated to the required value
with the feedback loop control, it is important to achieve a good temperature stability for all control
Sensors 2020, 20, 259 12 of 19

4.2.2. Drift
Although the output voltage is continuously measured and it is regulated to the required value with
the feedback loop control, it is important to achieve a good temperature stability for all control stages, to
avoid further re-regulation and re-calibration of the device after small temperature variations. For this
purpose, resistors with small temperature coefficients in the controllable current sources (Figure 4) have
been used. Considering the sum of the effect of the maximum temperature drifts of each component
for temperature changes of ±25 ◦ C, an overall maximum drift error of 0.2% for each controlling stage
has been obtained, thus demonstrating a very low dependency from the temperature variations.

4.2.3. Resolution
The blood pressure output resolution can be calculated considering the digital potentiometers
resolution (POTRES ) and the blood pressure ranges of all individual controlling stages. The waveforms
are generated by the controlling stage 1, where the blood pressure resolution is linearly dependent
from the selected range. For the maximum range (8), the minimum resolution for the controlling
stage 1 is obtained as:
BPMAX 333.9
BPRES_MIN = = 10  0.33 mmHg. (12)
POTRES 2
The resolution of controlling stage 2 corresponds to the relation (12), with the same values of
BPMAX and POTRES . For the controlling stage 3, the blood pressure range is about −60 mmHg and the
resolution is 0.06 mmHg. The control stage 3 is, therefore, suitable for regulating the static pressure.
The controlling stages accuracy is summarized in Table 4.

Table 4. Controlling stages BP accuracy summary.

Parameter Value Unit Description

Minimum BP resolution
0.33 mmHg Full range
for the controlling stage 1
BP resolution for the
0.33 mmHg
controlling stage 2
BP resolution for the
0.06 mmHg
controlling stage 3
Dynamic BP The percentage of
0.15 %
nonlinearity error selected range
The percentage of output
BP drift error for
0.2 % for individual
∆t = 25 ◦ C
controlling stage

4.3. The Accuracy of Measuring Stages

The accuracy of the device can be evaluated considering the feedback measurement accuracy.
Each measuring stage, used to evaluate the output voltage VOUT and the input voltage VIN , consists of
some active components: an instrumentation amplifier, an operational amplifiers, a voltage reference
and an analog to digital converter (ADC). Moreover, passive components (e.g., resistors) have been
used to realize both stages, but some of these negatively affect measurement results. In the following,
a direct evaluation of the accuracy and of the requirements of the individual components is reported.
The measurement errors can be divided in different categories: voltage offset errors, common mode
(CM) voltage errors referred to 1 V of VIN , noise, nonlinearity, gain errors, and temperature drift for
∆t = 25 ◦ C, resolution. The errors due to instrumentation and operational amplifiers are referred to
the output (RTO) for possibility of aggregation. Errors caused by the bias current and the noise current
can be neglected when compared to input offset voltage errors.
Sensors 2020, 20, 259 13 of 19
Sensors 2020, 20, x 13 of 19

In Table
Most 5, the measuring
errors stage 1to
can be associated accuracy results
static errors are shown.
(offset voltage,The
CMabsolute
error). Invalue errors of VOUT
this application, where
canabe calculated from the relation:
calibration is available, these errors can be removed by resetting the output to zero after power-on.
Verrorout
The measurement noise can be compensated ∆VOUT = by normalizing the ADC conversion results by a(13) factor
GAIN1
of 1/√𝑛, where 𝑛 is the number of conversions.
where the Verrorout is the voltage error of the source (see Table 5), and the GAIN1 = 201 is the IA gain.
The corresponding BP error can Table
be computed fromstage
5. Measuring Equation (10). summary.
1 accuracy

Common stage 1 accuracy summary.

Table 5. Measuring
Offset Non- Gain Drift Resoluti
Component Mode (CM) Noise
OffsetVoltage
Common Mode (CM) Linearity Error DriftError on
Component Noise Non-Linearity Gain Error Resolution
Voltage Error Error Error
Instrumentation
25 mV 0.6 mV/V 160 µV negligible 0.24%
10 mV
10 mV + 0.055% -
Instrumentation
Amplifier (IA)
Voltage 25 mV - 0.6 mV/V 6 µV 160 µV- negligible 0.24%0.5 mV + - -
Amplifier (IA)1 mV
ref. -
Inv. amp. 5 mV - - - - 0.1 mV0.055% -
Sum. amp. 1 mV - - - 0.2% 0.005% -
Voltage
ADC CH1 ref. - 1 mV - - 20 µV 6 µV 0.15 mV - 0.024% - - 0.5 mV 76 µV-
OverallInv. amp. 32 mV
values 5 mV 0.6 mV/V - 186 µV - 0.15 mV - 0.464% - mV +0.1
10.6 mV 76 µV-
0.06%

Sum. amp. 1 mV - - - 0.2% 0.005% -

ADC CH1 - - 20 µV 0.15 mV 0.024%
Most errors can be associated to static errors (offset voltage, CM error). In this application, where - 76 µV
a calibration is available, these errors can be removed by resetting the output to zero 10.6 after power-on.
Overall values 32 mV 0.6 mV/V 186 µV 0.15 mV
The measurement noise can be compensated by normalizing the ADC conversion results by 0.464% mV + a factor
76 µV of
√
1/ n, where n is the number of conversions. 0.06%
The nonlinearity cannot be easily compensated but due to its small influence it can be neglected.
Thereported
The values nonlinearity cannot
in Table be easily compensated
5 correspond to a pressure of but duemmHg
0.15 to its small
for VINinfluence
= 1 VDC it can be neglected.
(0.015 mmHg
for The
VIN =values reported
10 VDC, in Tableobtained
respectively), 5 correspond to a pressure
from Equations (13).mmHg for 𝑉 = 1 VDC (0.015
of 0.15
(10) and
mmHg
The gain 𝑉 = can
for error 10 VDC, respectively),considering
be compensated obtained from the Equations
performed(10) testand (13).
measurements reported
in Figure 10. In this graph the mean values of ∆Vout are reported in blue, while the dispersion reported
The gain error can be compensated considering the performed test measurements for each in
BP Figure
value is10. In thisby
denoted graph the mean
the green values and
(maximum) of ΔV out are
grey reportedcurves.
(minimum) in blue,Forwhile
the the
testeddispersion
device, afor each
gain
error of 0.064% based on linear regression of the ∆VOUT was evaluated.
BP value is denoted by the green (maximum) and grey (minimum) curves. For the tested device, a
gain error of 0.064% based on linear regression of the Δ𝑉 was evaluated.

Figure 10. Gain error compensation, 𝑉 = 5 VDC: the mean values of ΔVout are reported in the blue
Figure 10. Gain error compensation, VIN = 5 VDC: the mean values of ∆Vout are reported in the blue
solid curve, while the dispersion values are denoted by the green dashed (maximum) and grey dotted
solid curve, while the dispersion values are denoted by the green dashed (maximum) and grey dotted
(minimum) curves.
(minimum) curves.

TheThe calibration
calibration could
could be integrated
be integrated in the
in the device
device firmware
firmware in order
in order to compensate
to compensate bothboth drift
drift andand
static
static errors,
errors, considering
considering thethe ambient
ambient temperature
temperature measurement.
measurement. Unfortunately,
Unfortunately, thethe designed
designed device
device
does not allow any drift compensation, since it is assumed that it operates at 25
◦
does not allow any drift compensation, since it is assumed that it operates at 25 C, but this feature °C, but this feature
could
could be implemented
be implemented as aasfuture
a future development.
development. TheThe values
values reported
reported in Table
in Table 5 show
5 show a drift
a drift error
error on on
the the pressure
pressure of 10.6ofmmHg
10.6 mmHg
+ 0.06% +in testing
0.06% conditions
in testing ofconditions
VIN = 1 VDC 𝑉 =∆t1=VDC
of and (1.06∆𝑡
25 ◦ Cand = 25 °C
mmHg
(1.06 mmHg + 0.06% for 𝑉 =
+ 0.06% for VIN = 10 VDC and ∆t = 25 C). 10 VDC
◦ and ∆𝑡 = 25 °C).
The BP resolution for 𝑉 is 0.076 mmHg for 𝑉 = 1 VDC (0.0076 mmHg for 𝑉 = 10 VDC),
as it can be computed from Equations (10) and (13).
Sensors 2020, 20, 259 14 of 19

The
Sensors BP20,
2020, resolution
x for VOUT is 0.076 mmHg for VIN = 1 VDC (0.0076 mmHg for VIN = 10 VDC), 14 of 19
Sensors 2020, 20, x 14 of 19
as it can be computed from Equations (10) and (13).
4.4.The
4.4. TheOutput
OutputAccuracy
Accuracy
4.4. The Output Accuracy
Themeasurement
The measurementofofΔ𝑉 Δ𝑉 has hasbeen
beenperformed
performedforfordifferent
different values
values of ofVINV. INIn
. In particular,
particular, 3 3
The measurement
different values of ∆V
have OUT has
been been performed
selected: 𝑉 = 1 for
VDC different
, 𝑉 =values
5 VDCof V, INand
. In particular,
𝑉 = 10 3VDC
different
(see
different values have been selected: 𝑉 = 1 VDC , 𝑉 = 5 VDC , and 𝑉 = 10 VDC (see
values
Figures have been selected:
11–13). For every
every = 1 VDC, IN
VINvalue = 5 VDC,
V,INtests have andbeenVIN = 10 VDC for(seetwoFigures 11–13). For
Figures 11–13). For value ofofVV repeated
IN, tests have been repeated for two different ambient
different ambient
◦
every value of Vi.e.,
IN , tests have
25°C°Cbeen
andrepeated for two different ambient temperatures, i.e., t = 25theCmeanand
temperatures,
temperatures, i.e., 𝑡 𝑡==25 and 𝑡 𝑡==5050°C. °C.InIneach
each figure,
figure, thethe blue
blue curve
curve represents
represents the mean
= 50 ◦ C.
t values In each figure,
Δ𝑉 , ,while the blue curve represents the mean values of ∆V OUT , while the dispersion of
values ofof Δ𝑉 whilethethedispersion
dispersionofofmeasurement
measurement errors,
errors, which
which areare mainly
mainly duedue to to
thethe adopted
adopted
measurement errors,
measurementinstruments which performances,
instruments are mainly dueare to reported
the adopted usingmeasurement
green instruments
(maximum) and performances,
grey (minimum)
measurement performances, are reported using green (maximum) and grey (minimum)
are reported
curves. using green (maximum) and grey (minimum) curves.
curves.
The
Theresults
resultsof the three sets ofoftest are shown in Figures 11–13 for the values of VIN = 11VDC, of
The results ofofthe
thethree
threesets
setsof testare
test areshown
shown inin Figures
Figures 11–13
11–13 forfor
thethe values
values of of𝑉 𝑉= 1= VDC, VDC, of of
V𝑉IN = 1 VDC, and
andofofof𝑉V𝑉IN=== 11 VDC, respectively. Numeric
Numeric valuesvalues are
are reported
reported in in Table
Table 6.
𝑉 ==11 VDC, VDC,and VDC,respectively.
1 VDC, respectively. Numeric values are reported in Table 6. 6.
These measurements
Thesemeasurements
measurementsshow show a relatively
showa arelatively high
relativelyhigh error
higherror due
errordueto measuring
duetoto measuringinstruments
instruments(green(green
and grey)and
These measuring instruments (green and
for high
grey) BP
for values.
high BP values.
grey) for high BP values.

(a)
(a) (b)(b)
Figure 11.
Figure11.
Figure 𝑉𝑉OUT deviations
11.V deviationsfor
deviations for𝑡t =
for 𝑡==2525◦°CC°C(a)
25 (a)(a) and𝑡t =
and
and 𝑡=
= 50 ◦°C
5050 (b),
C°C V𝑉IN𝑉=
(b),
(b), =1=1 1VDC.
VDC. The
VDC. mean
The
The mean
mean values ofofof
values
values
ΔVout
∆Vout are
are reported
reported inin the
the blue
blue solid
solid curve,
curve, whilewhile
the the dispersion
dispersion valuesvalues
are are denoted
denoted
ΔVout are reported in the blue solid curve, while the dispersion values are denoted by the greenby by
the the
green green
dashed
dashed
dashed(maximum)
(maximum) and greyand
(maximum) grey
dotted
and dotted
dotted(minimum)
(minimum)
grey curves. curves.
(minimum) curves.

(a) (b)
(a) (b)
Figure 12. 𝑉 deviations for 𝑡 = 25 °C (a) and 𝑡 = 50 °C (b), 𝑉 = 5 VDC. The mean values of
Figure 12.V𝑉OUT deviations fort 𝑡==25
deviationsfor 25◦ C°C(a)(a)and
and =𝑡 =5050
◦ C°C
(b),(b), 𝑉 = 5VDC.
IN = 5 are
VDC.The
The meanvalues
values of
Figure
ΔVout 12.
are reported in the blue solid curve, while tthe dispersion Vvalues denotedmean
by the greenof
ΔVout are
∆Vout reported in the
blueblue
solidsolid curve,
whilewhile the dispersion
values values are denoted by the green
dashedare reported
(maximum) inand
thegrey dotted curve,
(minimum) the dispersion
curves. are denoted by the green dashed
dashed (maximum) and grey dotted (minimum)
(maximum) and grey dotted (minimum) curves. curves.
Sensors 2020, 20, 259 15 of 19
Sensors 2020, 20, x 15 of 19

(a) (b)
Figure 13.
Figure 𝑉OUT deviations
13. V deviationsforfor 𝑡t =
=2525 ◦°C and t𝑡 =
C (a) and = 50
50 ◦°C (b), V𝑉IN =
C (b), = 10 VDC. The
10 VDC. Themean
meanvalues
valuesofof
∆Vout
ΔVoutare
arereported
reportedinin
thethe blue
blue solid
solid curve,
curve, while
while the dispersion
the dispersion values values are denoted
are denoted by theby the green
green dashed
dashed (maximum)
(maximum) and greyand grey(minimum)
dotted dotted (minimum)
curves. curves.

Table 6.
Table Summary of
6. Summary of the
theoutput
outputaccuracy
accuracytests
testsresults.
results.

Test Test Vin

Vin[V[V
DC]
DC] t [ C]ΔVout MAX∆V
◦
t [°C] [µV]
out MAX [µV]
ΔBP ∆BPMAX [mmHg]
MAX [mmHg]

25 25 1.5 1.5 0.3 0.3

Test 1 Test 1 1 1
50 50 4.2 4.2 0.8 0.8
25 25 3 3 0.12 0.12
Test 2 Test 2 5 5
50 5.8 0.23
50 5.8 0.23
25 10 0.2
Test 3 10 25 50 10 0.2
Test 3 10 10 0.2
50 10 0.2

In
In Table
Table 7,7, the
the results
results of
of the
the accuracy evaluation of
accuracy evaluation of measuring
measuringstage
stage22are
areshown.
shown.The
Theabsolute
absolute
value
value of
of the
the errors
errors ofof the
the input
input exciter voltage V
exciter voltage 𝑉IN can
can be calculated from
be calculated from the
the relation:
relation:

𝑉𝑒rror
Verrorout
Δ𝑉IN ==
∆V (14)
(14)
GAIN2
GAIN2
where 𝑉𝑒rror isisthe
where Verror theabsolute
absolutevoltage
voltage error
error of
of the
the source,
source, and GAIN2 ==0.4
and GAIN2 0.4isisthe
thevoltage
voltagedivider
divider
out
gain. The corresponding BP error can be computed from the relation (10).
gain. The corresponding BP error can be computed from the relation (10).

Table
Table 7.
7. Measuring
Measuringstage
stage22accuracy
accuracysummary.
summary.

Component
Offset CM CM Error Drift
Component Offset Voltage Noise Noise Non-linearity
Non-linearity GainGain Error Drift Error
Error Resolution
Resolution
Voltage Error Error
Voltage div. - - - - 0.25% 0.12% -
Voltage
ADC CH0div. -- - - - 20 µV - 0.15 mV 0.25% 0.024% 0.12% - - 76 µV
Total CH0
ADC -- - - 20 µV negligible
0.15 mV0.15 mV 0.024% 0.274% - 0.12% 76 µV 76 µV
Total - - negligible 0.15 mV 0.274% 0.12% 76 µV
The nonlinearity in the simulation of the BP can cause an error of 0.04% for VIN = 1 VDC (0.004%
for VINThe=nonlinearity
10 VDC), asinit thecansimulation
be deduced of from
the BPrelations
can cause anand
(10) of 0.04% for 𝑉 = 1 VDC (0.004%
error(14).
for 𝑉From= 10 VDC), as it can be deduced from relations (10) and (14).
these results, a BP error of 0.12%, caused by the drift error for any VIN , is obtained; moreover,
From that
it is shown thesethe results, a BPoferror
resolution VIN isof190
0.12%, caused by
µV, obtained therelation
from drift error
(14), for
andanythat𝑉the, gain
is obtained;
error can
moreover, it is shown that the resolution of 𝑉 is 190 µV, obtained from relation
be compensated by test measurements. In Table 8, the measurement errors of VIN for different input (14), and that the
gain error can be compensated by test
voltages at ambient temperatures are shown. measurements. In Table 8, the measurement errors of 𝑉 for
different input voltages at ambient temperatures are shown.

Table 8. 𝑽𝐈𝐍 measurement errors.

𝑽𝐈𝐍 𝚫𝑽𝐈𝐍 at 25 °C 𝚫𝑽𝐈𝐍 at 50 °C

1V 0 mV 0.6 mV
5V 0.02 mV 0.1 mV
10 V 0 mV 0.4 mV
Sensors 2020, 20, 259 16 of 19

Table 8. VIN measurement errors.

V IN ∆V IN at 25 ◦ C ∆V IN at 50 ◦ C
1V 0 mV 0.6 mV
5V 0.02 mV 0.1 mV
10 V 0 mV 0.4 mV

5. Discussion
The results described above show that the IBPS presented in this paper is a valid and reliable
instrument for testing pressure monitors. The calculation and measurement of the accuracy show
that there are potential improvements for the measuring stages, even if the actual performances allow
the use of the instrument in most practical cases. Better accuracy could be achieved by replacing the
components that cause the largest errors with higher quality components: this operation of re-designing
can be easily achieved, but in this phase, it is has not been realized because the principal aim was
to demonstrate how useful can be such kind of instruments, while testing BP monitoring devices.
The minimum accuracy of the device is achieved for the lowest input exciter voltage VIN and for
the lowest sensitivity. The sensitivity obtained at this stage of the device development is a constant
parameter, but the simulation of 40 µV/V/mmHg is necessary only for a minority of the commercial
IBP transducers. The switchable sensitivity feature in the range 5/40 µV/V/mmHg could be added
by modifying the current sources (shown in Figure 4) by means of switchable R1 or R3 resistors.
Moreover, the adjustable gain of the measuring stage 1 could improve the accuracy for lower VIN
values. As explained above, the temperature drift can cause a maximum BP calculated offset of about
0.5 mmHg/◦ C: this is mainly due to the contribution of drift on the output voltage measurement
(i.e., 0.43 mmHg/◦ C, as results from Table 5 data show) and to the contribution of drift on the input
voltage measurement (i.e., 0.015 mmHg/◦ C, as results from Table 7 data show). According to this
dimensioning data, this drift contribution could be compensated by the integration of a temperature
sensor that can allow the implementation of an automatic calibration feature to compensate the drift.
From the measurements performed during tests, a maximum drift value of 1.32 mmHg was obtained
when considering a variation of the temperature from 25 to 50 ◦ C: this corresponds to a maximum BP
offset of 0.053 mmHg/◦ C due to the temperature drift that is mainly caused by the instrumentation
amplifier selected for the measurement stage 1. The drift value calculated according to the parameters
given in the components datasheets is higher than the measured one. It is important to highlight
that measured values of the instrument accuracy have been obtained as average of repeated tests
on different specimens of IBPS. The use of a higher quality model of Instrumentation Amplifier
(i.e., with built-in gain resistor once the fixed gain is determined), the total calculated drift can be
theoretically reduced from 0.5 mmHg/◦ C to 0.05 mmHg/◦ C, thus reducing the instrument’s sensitivity
to temperature changes and increasing its accuracy for the range 0–50 ◦ C.
The results reported in Figures 11–13 and summarized in Table 6 show a very high accuracy:
the maximum value is about ±0,8 mmHg that corresponds to the 0.2% of setting when the Vin = 1 V
for tests conducted with a temperature of 50 ◦ C. This value increases in normal operative conditions
(i.e., less than 0.1% with a temperature of 25 ◦ C) and the accuracy increases for higher values of Vin .
The device presented in this work offers similar performances to more complex and expensive
systems used for the same purpose. There are few available device on the market available for IBP
simulation used to test BP monitors; among these, FLUKE Prosim 3 is one of the most complete
and performant system to simulate a wide range of biomedical signals, such as electrocardiogram
(ECG), respiration, temperature, cardiac output, etc., and IBP. This is a very complex and expensive
device, also considering the maintenance aspects, so the IBPS presented in this work can be a valid
alternative to such type of multi-simulators when simulations other than IBP are not needed. Moreover,
the performances of the two devices are very similar, in terms of sensitivity, pressure range simulation,
pressure accuracy, and input/output impedance, as reported in Table 9. The FLUKE Prosim presents
Sensors 2020, 20, 259 17 of 19

four different and independent channels, but only prefixed constant pressure values or pre-loaded
pressure waveforms can be simulated. The IBPS presented in this paper has only one channel, but
an extended version to four or more channels can be easily implemented. As described above,
the tuning of the static level of pressure simulation of the presented device is finer than FLUKE Prosim
3, and the possibility of simulating custom waveforms gives rise to a wider range of applications;
in fact, by using the USB connection, the presented IBPS can be programmed to simulate unlimited
and custom waveforms. The IBPS presented in this work can simulate a transducer sensitivity of
5 µV/V/mmHg, but in future developments, it will be easy implementable a second option for the
value of 40 µV/V/mmHg (it is HW reconfigurable ready). Finally, the pressure accuracy presented by
the IBPS described in this paper is considerably higher than the commercial one, also considering that
it can increase in normal operative conditions or with higher settings of Vin .

Table 9. Comparison between FLUKE Prosim 3 and the developed IBPS performances.

Fluke Prosim 3 IBPS

Channels 4 1
Input impedance 300 Ω 3700 Ω
Output impedance 300 Ω 300 Ω
Exciter input range 2.0 to 16.0 VDC 1.0 to 10.0 VDC
Exciter-input frequency range DC to 5000 Hz DC to 10,000 Hz
5 µV/V/mmHg
Transducer sensitivity 5 or 40 µV/V/mmHg (40 µV/V/mmHg HW
reconfigurable)
±2% of setting +2 mmHg (valid for dc ±0.2% of setting (maximum with
Pressure accuracy
excitation only) Vin = 1 V @50 ◦ C)
Static Levels 16 possible combining four channels −30 to 300 mmHg continuous
Dynamic waveform 10 predefined combining four channels User programmable via USB
Respiration Artifact BP delta changes from 3 to 16 mmHg User programmable via USB

6. Conclusions
The designed and developed IBPS presented in this paper is a low-cost single purpose device
compliant with the requirements for patient monitor IBP modules testing, zero setting, and calibration.
It is possible to simulate a wide range of situations by setting different options for static and dynamic
blood pressure waveform simulation.
The device allows setting a constant BP and to generate user-defined BP that can be adjusted by
setting the parameters of the waveforms. Both modes have an accuracy in simulating BP data that is
better than ±1 mmHg at 25 ◦ C, in which temperature represents the most common condition for the
environment in which this instrument is used. The controlling stages are sufficiently precise, allowing
to set the desired BP waveform in time and amplitude, with high sample rates and 10-bit resolution for
all the modulation ranges.
All described features can be easily implemented by a simplified user interface consisting in
a four-button navigation panel and a display. The USB connectivity increases the capabilities of the
device: It is possible to implement the remote control and to use the built-in memory to perform
operations, like uploading custom waveforms or generating real-time ones. The USB connection
can be used to upgrade the firmware in order to make the device hardware controllable. Moreover,
a MATLAB graphical user interface (GUI) is available for custom waveforms upstream in order to
make easy the use of the device for research and educational users.
The device has been fully tested and compared with one existing device, and it has proven to
have comparable and competitive features and performances; it can be used in research and clinical
environments and, after a proper certification process, it could be used as a medical device.
Finally, it is important to highlight that this device is an open source project, fully available for
interested developers and researchers. All the materials (i.e., schematics firmware, software, and user
manual) can be requested to the authors. In the future, it will be available on a web page, in order to
create a forum of developers where they can share information and contributions.
Sensors 2020, 20, 259 18 of 19

Author Contributions: Conceptualization, D.B. and J.K.; Methodology, J.K.; Software, J.K.; Validation, D.B.
and M.P.; Formal Analysis, P.K.; Investigation, P.V.; Data Curation, D.B., J.K. and P.Z.; Writing-Original Draft
Preparation, D.B., J.K. and P.K.; Visualization, D.B.; Supervision, P.K. and M.P. All authors have read and agreed to
the published version of the manuscript.
Funding: This research received no external funding.
Acknowledgments: The work and the contributions were supported by the project SV450994/2101 Biomedical
Engineering Systems XV’. This study was also supported by the research project The Czech Science Foundation
(TACR) ETA No. TL01000302 Medical devices development as an effective investment for public and private entities.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Eason, M.P. Simulation devices in cardiothoracic and vascular anesthesia. Semin. Cardiothorac. Vasc. Anesth.
2005, 9, 309–323. [CrossRef] [PubMed]
2. Green, M.; Tariq, R.; Green, P. Improving patient safety through simulation training in anesthesiology: Where
are we? Anesthesiol. Res. Pract. 2016, 2016, 12. [CrossRef] [PubMed]
3. Cerny, M.; Penhaker, M.; Gala, M.; Babusiak, B. Biomedical engineering education under european
union support. In 5th Kuala Lumpur International Conference on Biomedical Engineering; AbuOsman, N.A.,
Abas, W.A.W., AbdulWahab, A.K., Ting, H.N., Eds.; Springer: Berlin/Heidelberg, Germany, 2011; Volume 35,
pp. 16–20.
4. Johnson, S.M.; Owens, T.L.; O’Neil, J.N. Making the clinical connection from textbook to bedside during
MDY1: An integrative approach for medical physiology education employing human simulation. Adv.
Physiol. Educ. 2019, 43, 128–133. [CrossRef] [PubMed]
5. Fox, F. Method and Apparatus for Zeroing and Calibrating an Invasive Blood Pressure Monitoring System.
U.S. Patent No. 4,342,218, 3 August 1982.
6. Kim, C.H.; SaGong, G.; Nam, K.; Jeon, R. Blood pressure simulator using hybrid controller. J. Sens. Sci.
Technol. 2007, 16, 44–51.
7. Krejcar, O.; Slanina, Z.; Stambachr, J.; Silber, P.; Frischer, R. Noninvasive continuous blood pressure
measurement and GPS position monitoring of patients. In Proceedings of the 2009 IEEE 70th Vehicular
Technology Conference Fall, Anchorage, AK, USA, 20–23 September 2009; Volume 1–4, pp. 3–7. [CrossRef]
8. Kijonka, J.; Penhaker, M. Electronic invasive blood pressure simulator device for patient monitor testing.
Elektron. Elektrotechnika 2012, 122, 49–54. [CrossRef]
9. Penhaker, M.; Kijonka, J. Invasive blood pressure simulator electronics device bed side monitor testing. In
Electrical Power Systems and Computers; Springer: Berlin/Heidelberg, Germany, 2011; pp. 823–830.
10. Kijonka, J.; Penhaker, M.; Cernohorsky, J. Invasive blood pressure curves simulation device. In 2011 1st
Middle East Conference on Biomedical Engineering; IEEE: Piscataway, NJ, USA, 2011.
11. Kijonka, J.; Penhaker, M. Embedded programmable invasive blood pressure simulator. In XII Mediterranean
Conference on Medical and Biological Engineering and Computing; Springer: Berlin/Heidelberg, Germany, 2010.
12. Kijonka, J.; Penhaker, M. Invasive Blood Pressure Simulator, Testing, Adjusting and Calibration of the Device.
Master’s Thesis, Department of Cybernetics and Biomedical Engineering, Faculty of Electrical Engineering
and Computer Science, VSB-Technical University of Ostrava, Ostrava, Czech Republic, 2010.
13. Fida, B.; Bernabucci, I.; Bibbo, D.; Conforto, S.; Schmid, M. Pre-processing effect on the accuracy of event-based
activity segmentation and classification through inertial sensors. Sensors 2015, 15, 23095–23109. [CrossRef]
[PubMed]
14. NovaSensor Disposable Medical Pressure Sensor NPC-100 and NPC-120 Series. Available
online: http://www.amphenol-sensors.com/en/products/pressure-mems/mems-sensors/3150-npc-100-and-
120-series#features (accessed on 31 December 2019).
15. Garani, G.; Adam, G.K. Qualitative modeling at the design of concrete manufacturing machinery. Int. J.
Comput. Appl. 2008, 30, 325–330.
16. Proto, A.; Bibbo, D.; Conforto, S.; Schmid, M. A new microcontroller-based system to optimize the
digital conversion of signals originating from load cells built-in into pedals. In Proceedings of the IEEE
2014 Biomedical Circuits and Systems Conference (BioCAS), Lausanne, Switzerland, 22–24 October 2014;
pp. 300–303.
Sensors 2020, 20, 259 19 of 19

17. Babusiak, B.; Borik, S.; Gala, M. Bio-amplifier with programmable gain and adjustable leads for basic
measurement of bioelectric signals. In Information Technologies in Biomedicine; Pietka, E., Kawa, J.,
Wieclawek, W., Eds.; Springer: Cham, Germany, 2014; Volume 4, pp. 249–260.
18. Babusiak, B.; Borik, S. Bluetooth communication for battery powered medical devices. J. Electr. Eng.
Elektrotech. Cas. 2016, 67, 65–68. [CrossRef]
19. Babusiak, B.; Borik, S. Low energy wireless communication for medical devices. In Proceedings of the 38th
International Conference on Telecommunications and Signal Processing (TSP), Prague, Czech, 9–11 July 2015;
pp. 444–447.

© 2020 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access
article distributed under the terms and conditions of the Creative Commons Attribution
(CC BY) license (http://creativecommons.org/licenses/by/4.0/).