Development of instrumentation for 12-Lead impedance cardiography

P. Kauppinen, A. Haapalainen, J. Hyttinen, J. Malmivuo
Ragnar Granit Institute, Tampere University of Technology,
P.O. Box 692, 33101 Tampere, FINLAND

Abstract: Cardiac output (CO) is one of the core parameters of cardiac function. Measurement of CO by impedance cardiography (ICG) has not achieved wide clinical acceptance due to its lack of reliability. We have been investigating multichannel ICG measurements based on computer simulations of sensitivity distribution of ICG applying the electrode system of 12-lead electrocardiography (ECG). In the present work, instrumentation for the purpose of multichannel 12-lead ICG is being presented along with initial test measurements of the device. For the first time, simultaneous multichannel ICG recording with all independent information is obtained. Future work will be aimed to discover the clinical usefulness of the 12-lead ICG.

INTRODUCTION

Measurement of cardiac output by impedance cardiography has not achieved wide clinical acceptance because the results obtained in comparison with accepted standard methods often prove discouraging [1]. Conventional ICG methods utilize a single waveform, which is difficult to apply in extracting reliable information. One potential reason may be the multiple sampling sensitivity of various ICG systems [2]. Since several sources contribute to impedance waveform simultaneously, an approach where a quantity of signals is recorded might be useful in estimating cardiac parameters [3].

It has been shown, based on computerized application of the lead field theory and volume conductor modeling, that by appropriately selecting the electrode configuration it is possible to arrive at measurement imposing increased sensitivity to particular regions [4]. The idea in 12-lead ICG is to record multiple signals with more specific recording properties and to combine these data for estimating cardiovascular function. Comparison of simulations with parameters derived from initial clinical experiments correlated favorably, indicating the potential of the theoretical method in developing new ICG [4].

Preliminary clinical experience has been acquired with an experimental single channel ICG system [3]. An important limitation in this line of research was, that only one channel could be recorded at a time, being time consuming and impractical. Requirements imposed by the theoretical aspects of achieving 12-lead ICG measurements led to the development of multi-channel ICG instrumentation, which is here preliminarily reported along with data from initial testing of the instrumentation.

METHODS

In 12-lead ICG all independent tetrapolar measurements are needed, i.e. measurements using two electrodes for delivering current and other two for sensing potential. Number of independent measurements is thus (by adding another reference electrode e.g. on the ankle) 8 * 8, from which half are reciprocal, resulting to 32 independent measurements [5]. From these data all possible configurations can be computationally derived, independent of the number of electrodes applied in the measurement configuration. Although the information is in the set of independent measurements, it has been shown, that by making measurements utilizing e.g. 8 electrodes simultaneously, clearly more selective presentation of the impedance changes can be obtained [4]. Compared to impedance tomography, also time-varying signal is required in ICG in addition to the baseline impedance.

Figure 1. Schematic view of the system and accompanying connections between subject and instrumentation.

Fig. 1 presents a schematic view of the system. The main components are:

Z/Multiplexing-board consists of a signal generator, voltage to current converter, multiplexers for both current and voltage circuits, measuring electronics and isolation circuit. The amplitude of the signal generator is determined by the supply voltages, providing well-regulated output. Low impedance multiplexers are used to change the current injecting and voltage measurement pairs so that the current and voltage channels are selectable independently. The peak value of the measured signal is compared to the reference signal and if basal impedance is measured, the reference signal is set to zero. After obtaining the basal value, the reference signal is set to be equivalent to this value to cancel basal component and only the time-varying component of the signal can be amplified. The obtained signal has to be amplified to the range suitable for A/D conversion. This is done by an inverting op-amp circuit, which also includes initial noise filtering. Finally, optocouplers are used to isolate both the digital and the analog signals.

Measurement Card. DAP 4000-series 14-bit Data Acquisition Processor board (Microstar Laboratories Inc., USA) with digital and analog I/O lines is used to collect data from the Z-board and to control its multiplexers. The digital output is connected to control lines and two analog inputs are connected to the basal impedance and time-varying impedance terminals.

Control Software was implemented with Matlab® (MathWorks Inc., USA) using a driver package ACCEL32 for DAP card. DAP has an onboard operating system which uses a special programming language DAPL. Matlab is used to initialize the card and then collect the data to the computer and to feed the DAP card with proper output signals in a constant rate. As explained earlier, the basal impedance is measured by setting the reference signal to zero and sampling corresponding the voltages from different channels.

RESULTS

Figure 2. Example four-channel test recording of time-varying impedance. Refer to the text for explanation of the recording configurations. Signal with the highest amplitude is obtained between the arms, two signals closely identical are normal and reciprocal measurements (channels 2 and 3) and the smallest amplitude reflects noise, measured over 400 Ohm R.

Example simultaneous four-channel impedance signals taken with the implemented system are shown in Fig. 2. Two ECG spot electrodes were attached to both arms and one on the chest. Current injection frequency was 100 kHz and data sampling rate 20 Hz. Configurations were following: Channel 1: Current between the arms, voltage between the arms, Channel 2: Current from the arm to the chest, voltage between the arms, Channel 3: Reciprocal of Channel 2 (Voltage from the arm to the chest, current between the arms) and Channel 4: Current and voltage over a 400 Ohm resistor.

DISCUSSION

ICG could be made more reliable by sampling the target region with several different electrode settings giving special emphasis to the particular region. The developed multi-channel ICG apparatus allow the recording of all independent impedance information from the 12-lead electrode system simultaneously. Derivation of complex multi electrode signals can be post-processed off-line.

At the present time, the system is being improved to be able to sample at higher sampling rate with less noise prior to practical measurements in clinical projects related to cardiac function and fluid dynamics. Effects of posture, ventilation, electrode misplacements, physiological response tests such as tilting test and pharmacological tests will be conducted. Investigations with the 12-lead ICG in clinical environment will include measurements during dialysis to monitor the progress of treatment in pulmonary oedema, measurements distinguishing cardiac abnormalities which disturb the CO estimation with conventional ICGs and, finally, measurements of quantity of signals closely related to cardiac function for deriving CO based on biophysical function.

Acknowledgments: Work was supported financially by the Ragnar Granit Foundation, the Finnish Cultural Foundation. and the Medical Research Fund of Tampere University Hospital.

REFERENCES

[1]  H.D. Fuller. "The validity of cardiac output measurement by thoracic impedance: a meta-analysis," Clin Invest Med, vol. 15, pp. 103-112, 1992.

[2]  P.K. Kauppinen, J.A. Hyttinen, J.A. Malmivuo. "Sensitivity distributions of impedance cardiography using band and spot electrodes analysed by a three-dimensional computer model," Ann Biomed Eng, vol. 26, pp. 694-702, 1998.

[3] P.K. Kauppinen, J.A.K. Hyttinen, T. Kööbi, J. Malmivuo. "Multiple lead recordings improve accuracy of bio-impedance plethysmographic technique," Med Eng Phys, vol. 21, pp. 371-375, 1999.

[4]  P. Kauppinen, T. Kööbi, S. Kaukinen, et al. "Application of computer modelling and lead field theory in developing multiple aimed impedance cardiography measurements," J Med Eng Technol, vol. 23, pp. 169-177, 1999.

[5]  J. Malmivuo and R. Plonsey, Bioelectromagnetism: principles and application of bioelectric and biomagnetic fields. New York: Oxford University Press, 1995.