Designing Efficient Optimum Reduced Order IIR Filter for Smoothening EEG Motion Artifacts Signals

Abstract

Electro-Encephalography (EEG) signals are used for recording the human brain activities. As multi channel sensor nodes are used for recording these EEG signals thus recorded signals are usually suffers from the muscular motion artifacts or motions due to eye blinking. Thus it is highly required to smooth these motions artifacts form EEG signals. The main purpose of this paper is to test and design the effectiveness of reduced order IIR filters to smooth the motion artefact EEG signals. In the proposed method initially basic IIR filter is designed using band pass-band stop filter. Single EEG channel is considered at a time for smoothening. The parameters of IIR filter are tuned for improving the performance of artefact removal. In order to smooth the motion artefacts, IIR filter with Min-Max optimization is designed. The proposed technique is evaluated in terms of the quality of the artefact removal. Performance is also tested with parameters viz. Noise measurement (SNR) signal. Results show that the proposed algorithm provides improvements in SNR and various other performance parameters.