Noninvasive method for identifying coronary artery disease utilizing electrocardiography derived data

摘要

A method of analyzing empirically derived electrocardiograph (ECG) data which allows surprisingly accurate catagorization of subjects into various abnormal and normal classifications is disclosed. The presently preferred embodiment of the present invention applies an algorithm which compares root-mean-square (RMS) mean values derived from analysis of a representative composite of selected portions of a number of ECG PQRST waveforms obtained from (ECG) investigation of a subject, to similarly derived RMS mean and RMS standard deviation values present in a compiled data bank derived from (ECG) investigation of numerous normals, in each of a plurality of frequency range bands. A highly diagnostic numerical "Score" is calculated by addition of "Score" components found to be acceptable under certain mathematical criteria, and provided by the algorithm. Visually interpretable power spectral density plots supplement the method.