[New index for prediction of the patients prone to ventricular arrhythmias after acute myocardial infarction].

Recently late potential (LP) has been evaluated using signal averaged electrograms in order to predict the patients prone to the ventricular arrhythmias such as ventricular tachycardia (VT) or ventricular fibrillation (VF). However, the predictive accuracy was not satisfactory in clinical use. Therefore some new approach was expected. We tried to find a new index to predict the susceptibility to ventricular arrhythmias using frequency analysis by fast Fourier transformation technique (FFT). Fragmented electrograms are elucidated to be present through QRS complex beside the terminal portion of QRS and ST segment by Cain et al. Eighty-seven body surface electrograms were analyzed by FFT, and the power spectral distribution of specific band width of 25-50 Hz was evaluated. Late potential was also obtained by ordinary technique. The patients after acute myocardial infarction were classified into 5 groups according to Lown's grade; group 1 (grade 0.1), group 2 (grade 2), group 3 (grade 3,4A) and group 4 (grade 4B, non sustained VT), and group 5 (grade 4B, sustained VT/VF). The body surface distribution of 25-50 Hz were classified into 3 types; type A (almost equal two maxima), type B (unequal two maxima) and type C (single maximum). Types A & B were frequently observed, but type C was never seen in high risk group (group 4 and 5). Predictive accuracy of VT/VF was 89.2% by type A and 66.7% by LP. The combination of the two criteria improved the predictive accuracy of VT/VF up to 92.9%.(ABSTRACT TRUNCATED AT 250 WORDS)