Eigenvector analysis in body surface potential maps for evaluating right ventricular hypertrophy in primary pulmonary hypertension]

This study related the eigenvectors derived from the QRS complex in body surface isopotential maps of normal subjects to regional myocardial excitation, with special emphasis on the right ventricle. The subjects consisted of 120 normal healthy adults and eight patients with primary pulmonary hypertension and right ventricular hypertrophy (PPH). According to the Karhunen-Loève theory, eigenvectors were derived from the normal group and the eigenvector coefficient of each subject was obtained against time for the first three components. The cumulative proportion of the first three eigenvectors was 90.3% in normals and 80.2% in PPH. The first eigenvector had a peak of eigenvector coefficients early in the QRS and had a bottom late in the QRS. The peak of the second was in the middle of the QRS. The coefficients of the third eigenvector were low compared to the former two eigenvectors, and had no characteristic time pattern. PPH had a similar time pattern of coefficients in the first eigenvector, reduced coefficients in the second, and a definite peak in the mid-QRS in the third. The average eigenvector coefficient of the third eigenvector in PPH was significantly higher than those in normal subjects (p less than 0.01). We conclude that the third eignevector derived from normal subjects strongly reflects right ventricular excitation.