[New approaches to the analysis of ECG data]

Recently, techniques for analyzing computerized ECG data have been developed in time domain, frequency domain and micro level analysis of signals. In body surface ECG mapping, reverse problem analysis from body surface to cardiac surface was investigated, but findings were not sufficient to be clinically useful. Heart rate variability analysis from Holter ECG tapes showed the utility for determination of autonomic nervous function by time and frequency domain analysis. Late ventricular potentials recorded by Signal Averaged ECG have been beneficial for predicting malignant ventricular tachyarrhythmias, as well as delayed potentials in atrial activity. QT dispersion from 12-lead ECG has been studied as one of the indicators of ventricular instability and there was some difficulty in detecting the end points of QT intervals such as that in patients with flat T waves. The recently developed automatic analysis of the Least square fit method for detecting terminal points of T wave might be optimal compared with the other methods. Micro volt T wave alternans by spectrum frequency analysis of T wave in consecutive heart beats is a unique technique for predicting cardiac sudden death, however the mechanism remains obscure.