Download Impedance cardiography (ICG) is a non-invasive method to - IITB-EE

Internal report / Nov.2008
A microcontroller based thorax simulator for testing and calibration of impedance cardiographs
B. B. Patil, V. K. Pandey, and P. C. Pandey
Indian Institute of Technology Bombay, India
[email protected]
A thorax simulator is developed for testing and calibration of the instruments used for impedance
cardiography. Variations in the resistances and the bioelectric voltage sources in the thorax model are
realized using digital potentiometers and a microcontroller. It provides a small (0.1-1.2 %) periodic
resistance variation over the selected base resistance (20-200 Ω) for testing of the impedance detector
and the voltage signals for testing of the ECG amplifier.
1. Introduction
Sensing of the variation in the bioimpedance across a body segment can be used for noninvasive
monitoring of the changes in the fluid volume or underlying physiological events [1]-[10]. Impedance
glottography (also known as electroglottography or laryngography) senses the variation in the
impedance across the larynx for estimating the variation of the degree of contact between the vibrating
vocal folds during speech production [3]. Impedance plethysmography involves measuring impedance
changes in a body segment. Impedance cardiography can be used for estimating the stroke volume and
cardiac output by sensing the variation in thoracic impedance during the cardiac cycle [4]-[10]. The
bioimpedance is generally measured by passing an alternating current through a pair of surface
electrodes (spot or band electrodes). The voltage resulting across the body segment, sensed using the
same or another pair of electrodes, gets amplitude modulated due to the impedance variation. This
voltage is demodulated to get the impedance signal [4], [5]. In order to avoid any physiological effects,
a low-level current (<5 mA) in the frequency range of 20 kHz to 1 MHz is used. In this frequency
range, the tissues are not excitable (except possibly at very high current levels) and the impedance is
nearly resistive. The impedance measurement is carried out using two- or four-electrode configuration.
In the two-electrode configuration, the same set of electrodes is used for current injection and voltage
sensing. The current density is higher near the electrodes than elsewhere in the tissue and the measured
impedance generally becomes more dependent on the tissue near the electrodes. In the four-electrode
configuration, the current is passed between the two outer electrodes, while the voltage is sensed across
the inner electrodes. It generally provides a more uniform current density and reduces the effect of
skin-electrode impedance in the sensed voltage.
In impedance cardiography, the sensed impedance consists of basal impedance, a time-varying
component related to the physiological phenomenon of interest, and various physiological and nonphysiological artifacts. For satisfactory operation of the instrument, the impedance detector should have
appropriate sensitivity and the frequency response. The measurement of the variation in the impedance
should not be affected by variation in the basal impedance, the skin-electrode contact impedances, the
artifacts, and the external interferences. The cardiograph instruments generally have the facility for
connecting an internal resistance across the electrode terminals for calibration of the current source.
The current may be amplitude modulated to simulate the modulation of the sensed voltage due to time
varying impedance [11]. A thorax simulator can be used for testing and calibration of the instrument. A
simulator, developed earlier in our laboratory, used a model of the thorax with pulsatile change in the
impedance realized using analog switches [12]-[14]. The circuit also simulated a pulsatile voltage for
measuring the difference mode gain and CMRR of the ECG amplifier in the impedance cardiograph.
Here we present a thorax simulator for measuring various performance parameters of an
impedance cardiograph: the range of basal resistance, sensitivity, and frequency response of the
impedance detector, difference mode gain and CMRR of the ECG amplifier, and rejection of the
common mode interference from external sources. The variations in the resistance and voltage have
been realized using digital potentiometers. A microcontroller with keys and an LCD display provide
user interface for setting the simulation parameters.
2. A thorax model
For the purpose of impedance cardiography, the thorax can be modeled by a network as shown in Fig.1.
It represents the basal resistance, variation in the resistance related to the cardiac cycle, ECG voltage,
electrode-tissue contact resistances, and externally introduced common mode interference [12]-[14].
The terminals I1 and I2 are the current injection terminals and the voltage drop is sensed across the
terminals E1 and E2. The resistances Re’s represent tissue-electrode impedances, while Rs’s (Rs1, Rs2)



 
'
'
Vc  Vd / 2   R y 2 /  Rx 2  R y 2   Vx 2

 
R'y 2  R y 2 ||  Rv   Rx1 || R y1  
Re2
E1
+
Vd
2
Cardiac
Cycle
Ro
2
Vc _
+
Rs
Ro
2
Rp Vp
Int.
Com.
Ext.
Gnd
Re3
E2
Re4
I2
+

Vd
2
Rs2
'
'
Vc  Vd / 2  R y1 / Rx1  R y1 Vx1
where
I1
Rs1
_

R0 || Rs  R y1  R y 2 || Rv || Rx1  Rx2
Re1
_
represent the impedances outside the segment across which the
voltage is sensed. The impedance of the thoracic segment
contributing to the sensed voltage is Rs||Ro. The resistance Rs
varies with the cardiac cycle. The voltage sources Vd and Vc
represent the difference and the common mode voltages from
the internal bioelectric sources. In order to represent common
mode voltage with respect to an internal reference point, the
resistance Ro is represented as two resistances in series. The
voltage source Vp in series with resistance Rp represents the
common mode interference from external sources.
The thorax model of Fig. 1 has two voltage sources without
a common node. For realizing it as a circuit without using
transformers, the model can be modified to the schematic shown
in Fig. 2. The sources Vx1 and Vx2 are kept in phase and out of
phase for realizing Vc and Vd respectively. The relations
between the component values in the thorax model of Fig. 1 and
those of the schematic shown in Fig. 2 are as the following,
Rs1  Re1  Ra , Re 2  Rb , Re 3  Rc , Rs 2  Re 4  Rd
Fig. 1 Thorax model.
I1
Ra
R'y1  R y1 || Rv  Rx 2 || R y 2
Rx1
+
Cardiac
Cycle
Vx1
E1
Rb
Ry1
Rp
_
Rv
3. Simulator circuit
Vp
The variation in the thoracic resistance due to cardiovascular
_
Ry2
activity is a small percentage (< 2 %) of the basal resistance (20
Int.
Ext.
Com. Gnd
– 200 Ω). For digital control of the variation in the thoracic
Vx2
+
resistance, it is modeled as a fixed resistance in parallel with a
Rc
E2
Rx2
variable resistance with a value much larger than the onresistance of the analog switches (typically 50-200 Ω). For
Rd
I2
obtaining the frequency response of the impedance detector, a
sinusoidal variation in the resistance can be used. For a quick
Fig. 2 Schematic of the
estimation of the response, a square wave variation can be used
thorax simulator.
[15]. Use of a microcontroller and a digital potentiometer
permits us to select the frequency and waveshape of the
variation in the resistance.
The circuit of the thorax simulator is shown in Fig. 3. The resistance variation is realized using the
digital potentiometer IC5 (MCP 4105). Its resistance can be varied in 256 steps from 0 to 5 kΩ, with a
wiper resistance of approximately 75 Ω. The potentiometer is connected in series with R10 and R13 at
the two ends. The base resistance values can be selected through jumpers by connecting R14 and R15.
The resistance can be varied in accordance with the selected waveshape by providing appropriate
sequence of digital control words to the digital potentiometer. The relation between the values of
resistors in the simulator circuit in Fig. 3 and those in the schematic in Fig. 2 are as the following,
Ra  R6 , Rb  R7 , Rc  R8 , Rd  R9
 Ry1  Ry2  || Rv ||  Rx1  Rx2    R2  R3  ||  R10  RWB  R13  || R14 || R15 ||  R4  R5 
where RWB = resistance set by the digital potentiometer IC5, and the values R14 and R15 depend on the
jumper settings.
The voltage, representing internal bioelectric sources, with the selected waveform is generated by
digital potentiometer IC2. Its peak-to-peak amplitude is controlled by the digital potentiometer IC3.
Use of the two digital potentiometers permits to maintain the same number of quantization steps in the
C11 0.1µ
VDD
VCC+
Ex1
9
8 5
VSS 4
GND VCC A
E 1
CS
C 3
A 2
–
10
W 6
SDI
B
IC2
MCP4105
7
VCCVDD
C12 0.1µ
8 5 Ex
VSS 4
D 1 GND VCC A
CS
C 3
W 6
SDI
A
2
VDD
C10
SCKL
B
7
IC3
MCP4105
–
5
7 Ex2
+
C13 0.1µ
8
VSS 4
GND VCC
1
B
CS
C 3 SDI
5
7 VSS
VSS
EN
B
7
IC5
MCP4105
13
P3.3
R7
220
E1
JP6
JP5
R13
3.3K
R5
2.2K
IC8
CD4066
B
2
I1
W 6
R2 A 2
SCKL
100
IC4B
R15
150
R14
33
R6
220
A
1
A
14 VCC
0.1µ
VDD
R3
100
IC4C
R21
33K
6
R20
15K
R23
15K
R10
3.3K
8
+
Rpot
5K
R24
33K
SCKL
R4
2.2K
R8
220
E2
R9
220
I2
R16
2.2K
Ep
Fig. 3 Thorax simulator circuit
R1
Vss
VDD
C1 10µ
8.2k
31
9
RST EA
40
P2.7
VCC
P2.6
IC1
P2.5
AT89S52
P2.4
1
P3.6
P3.4
2
P3.7
P3.5
1
P1.0
2
P1.1
3
P1.2
4
P1.3
R
5
P1.4
10k
P0.1 38
20 GND
P0.0 39
28
DB7 14
27
DB6 13
26
DB5 12
11
25
DB4
6
14
E
4
15
RS
5 VDD
R/W
VCC 2
VEE 3
VSS
1
VSS
LCD1 VSS
C2
VSS
VDD
XTL1
18
33p
S1
0.1µ VSS
In
1µ
Out
Com
2
C4
9V
C5
0.1µ
0.1µ
VSS
Sync. Test
Outputs
A
B Digi. Pot.
C Control
D signals
E
VCC+
R19 VDD
4.7 k
IC7
1 7805 3
In Out
R11
4.7M
Com
2
R12
C16
R22
4.7 k
SW2
XTL2
SW1
19
C3
12MHz
1
C9
C14
IC6
7805 3
0.1µ
C6
0.1µ
2
–
4
IC4
1
3
4.7M
+ 11
IC4A
C7
LM324
0.1µ
AGnd
C8
0.1µ
VCC-
33p VSS VSS
Fig. 4 The controller and power supply circuit
waveform for different peak-to-peak amplitudes. The common and difference mode voltages in the
simulator circuit in Fig. 3 in terms of the voltages Ex1 and Ex2 are as the following,
 
'
'
Vc  Vd / 2  R3 / R3  R4
where
 Ex1 ,
 
'
'
Vc  Vd / 2  R2 / R2  R5
 Ex2
 R2 || R5    R10  RWB  R13  || R14 || R15 
R'2  R2 ||   R3 || R4     R10  RWB  R13  || R14 || R15  
'  R ||
R3
3
A polarity controlled unity gain amplifier, realized using op amp IC4B and analog switch IC8 is used
to control the phase relationship between the voltages Ex1 and Ex2. The port pin P3.3 of the
microcontroller is used to select the common or difference mode voltage. For testing the rejection of
external common mode interference by the impedance detector, the common mode interference can be
externally applied to Ep.
The controller and the power supply circuit is shown in Fig. 4. The microcontroller IC1 (AT
89S52) is used as the controller of the hardware. The three digital potentiometers are controlled via SPI
bus. The various wave shapes like sinusoidal, square, and a typical ECG have been realized using
different lookup tables in the microcontroller's memory. The synchronization test outputs are provided
on the microcontroller's port pins P3.6 and P3.7 for testing synchronization of the outputs of the ECG
amplifier and the impedance detector. The LCD display LCD1 and the two soft keys SW1 and SW2
provide the user interface. The circuit is powered by a single 9 V supply, with two fixed voltage
regulators (IC6, IC7) for the digital and the analog parts of the circuit.
The various parameters for simulation, type of waveshape (square, sine), the ECG mode (common,
differential), frequency (1-250 Hz), the amplitude of ECG voltages (0-100 mV), and variation in the
resistance (0.1 - 1.2 %) are selected using LCD and the two soft keys. The simulator can be used for
simultaneous testing of the impedance detector and ECG part of the impedance cardiograph, and also
for estimating the effect of external common mode interference on the measurements.
5. Results
In the common mode, the voltage can be varied in the range of 50-100 mV (p-p), with less than 200 μV
(p-p) of differential voltage. In the difference mode, the voltage can be varied in the range of 0-50 mV,
with less than 0.9 mV (p-p) common mode voltage. The base resistances for different jumper settings
are found to be 23.77 Ω, 28.24 Ω, 84.98 Ω, and 196.07 Ω. The resistance can be varied in the range of
0.1 - 1.2 % of the selected base resistance.
6. Conclusion
A thorax simulator has been developed for testing and calibration of the instruments for impedance
cardiography. The features includes settable values of heart beat rate, basal resistance, % variation in
resistance, the ECG voltage sources, and the various wave shapes (square, sinusoidal, and typical ECG)
for simulating the bioimpedance signal and the ECG.
References
[1] J. Nyboer, Electrical Impedance Plethysmography, 2nd ed., Charles C. Thomas, Springfield,
Massachusetts, 1970.
[2] L. Cromwell, F. J. Weibell, and E. A. Pfeiffer, Biomedical Instrumentation and Measurements,
2nd ed., Prentice Hall, New Delhi, 1990.
[3] A. K. Krishnamurthy and D. C. Childers, "Two-channel speech analysis", IEEE Trans.
Acoustics, Speech Signal Proc., vol. ASSP - 34, pp. 730 - 743, 1986.
[4] L. E. Baker, "Principles of impedance technique", IEEE Eng. Med. Biol. Mag., vol. 8, pp. 11 15, 1989.
[5] R. P. Patterson, "Fundamentals of impedance cardiography", IEEE Eng. Med. Biol. Mag., vol. 8,
pp. 35 - 38, 1989.
[6] J. Wtorek and A. Polinski, "Multifrequency impedance plethysmograph", in Proc. IEEE Inst.
Meas. Tech. Conf., Brussels, Belgium, 1996, pp. 1452 - 1455.
[7] J. Fortin, W. Habenbacher, and A. Heller, "Non-invasive beat-to-beat cardiac output monitoring
by an improved method of transthoracic bioimpedance measurement", Comp. Bio. Med., vol. 36,
pp. 1185 - 1203, 2006.
[8] F. Mora, G. Villegas, A. Hernandez, W. Coronado, R. Justiniano, and G. Passariello, "Design of
an instrumentation system for a neurocardiology laboratory", in Proc. 2nd Int. Conf. IEEE Dev.,
Cir. and Sys., Carcas, Venezuela, 1998, pp. 272 - 277.
[9] V. Vondra, J. Halamek, I. Viscor, and P. Jurak, "Two channel bioimpedance monitor for
impedance cardiography", in Proc. 28th Annu. Int. Conf. IEEE Eng. Med. Biol. Soc., 2006,
pp. 6061 - 6063.
[10] G. Panfili, L. Piccini, L. Maggi, S. Parini, and G. Andreoni, "A wearable device for continuous
monitoring of heart mechanical function based on Impedance Cardiography", in Proc. 28th Annu.
Int. Conf. IEEE Eng. Med. Biol. Soc., 2006, pp. 5968 - 5971.
[11] G. D. Jindal and J. P. Babu, "Calibration of dz/dt in impedance plethysmography", Med. Biol.
Eng. Comp., vol. 23, pp. 279 - 280, 1985.
[12] V. K. Pandey, P. C. Pandey, and J. N. Sarvaiya, "Impedance simulator for testing of instruments
for bioimpedance sensing", IETE Journal of Research, vol. 54, no. 3, pp. 203 - 207, 2008.
[13] N. S. Manigandan, V. K. Pandey, and P. C. Pandey, "Thoracic simulator for impedance cardiography", in Proc. National Symposium on Instrumentation (NSI-28), Pantnagar, India, 2003.
[14] L. Venkatchalam / P. C. Pandey (Supervisor), “Development of hardware for impedance cardiography”, M.Tech. Dissertation, Biomedical Engineering Group, School of Biosciences and
Bioengineering, Indian Institute of Technology Bombay, 2006.
[15] B. M. Oliver, Square-wave and pulse testing of linear systems, in B. M. Oliver and J. M. Cage
(Eds.), Electronic Measurements and Instrumentation, McGraw Hill, Singapore 1975.
Keywords: Bioimpedance simulator, Thoracic simulator, Impedance cardiography, Plethysmography.