ABSTRACT
The segregation of seemingly similar electrocardiogram data into two mutually exclusive classes can be achieved with a non-invasive procedure. Specifically, the problem of separating the electrocardiograms of preclinical-coronary subjects from those who are truly normal has been studied. In the current approach, the standard EKG waveform in its conventional linear format is transformed into a non-linear closed display, greatly improving the degree of visual perceptibility. In addition, specific non-dimensionalized parameters of the EKG waveform are extracted to produce a multivector spatial representation. The analysis of 129 cases indicates that this new technique results in a significantly higher degree of detection of preclinical coronary artery disease than current clinical methods. A prototype of a clinical system utilizing the output of an electrocardiograph has been developed for performing this analysis.
