Classification of ST-segments in ECG using ICA and Triangle method

The ischemia, which occurs because of blockage in cardiac electrical conduction path and cardiac injury, is detected with ST-Segment variations such as elevation and depressions. The frequency analysis and time analysis, which mostly make use of linear filter techniques and wavelet techniques, do not remove noise to large extent and particularly from the passband of ECG. In this paper Independent Component Analysis (ICA) is used to extract ECG and Trinagular Method is used to extract ST-segments. As ICA can transform noisy ECG into several components with more independence and maximum separation, the desired cleaned ECG can accurately be extracted and the K-point can be identified more accurately with Triangular Method. In earlier methods the end point of ST-Segment (K-point) is identified by using some fixed time backward from T-wave peak. The proposed method gives accurate measurement of ST-Segment and its elevation or depression. Finally ST-Segments are classified according to their shapes.