Biosignal Analysis Using Independent Components with Intelligent Systems

Abstract

This paper proposes early detection of myocardial infarct and heart arrhythmias from the characteristic pattern of the ECG waveform, and signal processing techniques for analysis of the biosignals along with the feature extraction and classification technique. Independent component analysis (ICA) is considered as a new technique suitable for the separation and removal of assorted noises independent of ECG signals. ECG Feature Extraction plays a major role in analyzing most of the cardiac diseases. This scheme determines the intervals and amplitudes in the ECG signal for succeeding analysis. The amplitudes and intervals value of P-QRS-T segment defines the functioning of the human heart. Artificial intelligence improves the biosignals’ monitoring efficiency and helps serious caretakers to get a faster prior diagnosis. In our current work, we incorporated machine learning and different architecture of artificial neural network (ANN). The annotated standard samples from MIT-BIH arrhythmia database are used for experiments. Results attained using the proposed algorithm using MIT-BIH and PTB database illustrates that the neural network classifiers demonstrate high classification accuracies of over 98.96% should help cardiologist for early diagnostic of arrhythmias.