This paper presents a technique that integrates independent component analysis (ICA) and Wavelet packet decomposition (WPD) for feature extraction from ECG for detecting Myocardial Ischemic beats. In the proposed work, the denoised ECG beat segments are projected on the bases to create the independent component (IC) vectors. Further, these IC vectors are disintegrated by WPD. The feature set for distinguishing ischemic beats is extracted by calculating entropy, mean and standard deviation from wavelet coefficients. The extracted features are used as inputs for SVM, ANN and KNN based classifiers for uncovering ischemic beats from normal beats. The performance of classifiers is cross-validated with ECG signal obtained from physiobank database in terms of accuracy, sensitivity and PPA. The results have revealed a maximum detection accuracy of 96.85% from ANN classifier which implies that the ANN classifier has huge potential in medical decision making.
Volume 11 | 01-Special Issue