Energy-Efficient and High-Speed Robust Channel Identification Methodology to Solve Permutation Indeterminacy in ICA for Artifacts Removal from ECG in Remote Healthcare

In this paper we propose an energy-efficient and high-speed methodology for solving the permutation indeterminacy encountered in Independent Component Analysis (ICA) for artifacts removal from recorded 3-channel ECG signals within the remote health monitoring environment for cardiovascular patients. Unlike the existing approaches, the proposed algorithm takes into account the non-stationarity of the ECG signal thereby making this methodology more realistic. Since the proposed methodology is based on the adaptive template matching scheme, once the channel containing the intended ECG signal is detected, there is no need to continue this proposed methodology for the rest of the channels. Thus it saves computation time as well as minimizes the overall energy consumption of the system. The proposed methodology is tested using 275 ECG signal recordings of patients obtained from clinically well-known MIT-BIH PTB databases and real-life ECG data collected from University of Southampton hospital corrupted by an exhaustive set of 5 bodily artifacts and noise. The proposed methodology is found to save on the 55.55% of the otherwise incurred energy and 33.33% speed improvement as compared to the existing approaches.