Ballistocardiogram Signal Denoising Using Independent Component Analysis

Abstract

Ballistocardiogram (BCG) signal provides measurement of body’s reaction force for cardiac blood ejection. BCG signal acquisition and processing techniques have attracted researchers due to its huge applications in biomedical field. These techniques comprise of detection of BCG signals which includes weighing scales, chairs, tables, or beds. There are various signal acquisition processes present, i.e., EEG, fMRI, etc. In last decade, simultaneous recording to EEG and fMRI signal has grown as promising technique in this field of biomedical where brain monitoring or other physical activities need to be monitored. During signal acquisition, original signal gets corrupted due to various conditions which affect the signal. This unwanted signal causes artifact in the original signal, which makes difficult for researchers to analyze and monitor the activities. In order to address this, we propose independent component analysis-based approach for artifact removal. Proposed approach is implemented using MATLAB tool, and experimental study is carried out for various user ‘data’set. Outcomes of the filtering method show better filtering or denoising performance.