Summary: | A method of analysis of the electrocardiogram (ECG) and phonocardiogram (PCG) designed to detect cardiac disorders with increased sensitivity over conventional ECG and PCG analysis is developed here. The method involves the use of computerized statistical signal analysis techniques to study the random components of the ECG and PCG, with the processing done in a manner that is phase-locked to the cardiac cycle.
This method of cardiac signal analysis is called Phase Invariant Signature Analysis or PISA. The method was implemented on a hybrid computer· system
and tested using simulations of the ECG and PCG. Further testing was done by analysing the ECG from dogs in which varying degrees of ischemia were induced and detected using PISA method.
The ECG and PCG of a group of normal human subjects were analysed by PISA. By using a quantification process a PISA cardiac index was defined and the normal ranges of this index were found for both the ECG and PCG. Clinical tests were carried out using a group of control subjects and a group of subjects
with ischemic heart disease. The ECG of both groups were analysed by the PISA method during rest and exercise. A third group of combined normal and abnormal subjects was analysed in the form of a blind study to compare PISA analysis
to conventional analysis of the ECG and PCG. It was found that'through"'PISA analysis of the ECG and PCG disorders in the cardiac system could be detected with a greater sensitivity than conventional EeG and PCG analysis.
'The greater sensitivity is due, in part, to the fact that in P!SA analysis an intrinsic component of the signal (ECG or PCG) is measured directly from the signal itself rather than comparing the ECG and PCG to external standard references
as is done in conventional analysis. A second reason for the increased sensitivity is that a much longer record of the ECG and PCG are used in PISA than are used conventionally. Although PISA analysis has been found useful in the detection
of cardiac disorders, further studies are required to develop a means of classifying the disorders once they have been detected.
|