We have modified impedance cardiography for monitoring cardiac output during stress tests. Employing an off-line microcomputer, our instrument ensemble averaged impedance signals to minimize the effect of motion artifacts. We proposed a new four-spot electrode array and replaced the usual encircling band electrode array with it. We tested ten normal subjects and compared the signal-to-noise ratio (SNR) from our spot electrode array to that from a typical band electrode array at rest and during four levels of exercise on a treadmill. The average of the sighal-to-noise ratios for ten subjects from our spot electrode array was 13.6-45.5 percent larger than that from the band electrode array at rest and during four levels of exercise. Thus, it is desirable to replace band electrodes with spot electrodes in impedance cardiography for exercising subjects.

We have modified impedance cardiography for monitoring cardiac output during stress tests. We adapted an ensemble-averaging technique for eliminating motion artifacts. We applied an array consisting of four spot ECG electrodes for impedance cardiogram (ICG) monitoring and compared it to conventional encircling band electrodes. We tested ten normal adults, compared the cardiac output obtained by our ICG monitoring system to that simultaneously obtained by the carbon dioxide (CO2) rebreathing method at rest and during three levels of treadmill exercise. The results show that the correlation coefficient between the spot electrode ICG and the CO2 rebreathing method is r = 0.90, and between the band electrode ICG and the CO2 rebreathing method is r = 0.96. If we use the peak-to-valley height of dz/dt instead of the peak height of dZ/dt in computing cardiac output, the correlation coefficient between the spot electrode ICG and the CO2 rebreathing method can be improved to r = 0.95.