Principal Component and Independent Component Calculation of ECG Signal in Different Posture

A correct diagnosis of the problem can only lead to the correct treatment. Use of various signals (biomedical signals) produced by the human body for the purpose of diagnosis is not new. It is well known that a Proper analysis of biomedical signals leads to a correct diagnosis of the problem. Presently the biomedical signals for ECG are collected in supine position which may not be comfortable for all the patients. In this paper, a method of analysing the ECG signals collected in various postures i.e. sitting, standing and supine has been presented. ECG signals of 16 patients have been collected in all the three postures and analysed by using Principal Component Analysis (PCA) and Independent Component Analysis (ICA). PCA and ICA together are suitable for signal analysis because PCA reduces dimension of the acquired data whereas ICA reduces noise of the dimension reduced signal. Eigen value variance for all three postures has been calculated and found the first principal component Eigen value variance is 99.95% in standing posture, 99.94% in sitting posture, 99.66% in supine posture. The result indicates a high scale reduction of the data with considerable accuracy.