Evolutionary Algorithm-Based Optimal Wiener-Adaptive Filter Design: An Application on EEG Noise Mitigation

Abstract

Objective: Electroencephalogram (EEG) signals are well-known non-stationary brain signals of lower strength. Due to their small amplitude, they attract other biomedical artefacts from the surroundings. This research mainly focuses on removing artefacts from the EEG. Method: The presented work uses the recent metaheuristics to efficiently design a Wiener-based adaptive noise mitigation structure (WANMS). Many powerful evolutionary optimisation algorithms (EOAs) such as particle swarm optimisation algorithm (PSOA), moth flame optimisation algorithm (MFOA), symbiotic organism search optimisation algorithm (SOSOA), and the student psychology-based optimisation algorithm (SPBOA) have been applied for the optimal design of WANMS. The proposed structure is analysed with various noisy signals such as electrooculogram (EOG) and electrocardiogram (ECG) with white Gaussian noise (WGN). Results: Among all the metaheuristic algorithms applied to WANMS, the SPBOA-based WANMS has performed better with improved Signal-to-Noise-Ratio (SNR) and minimal Mean-Squared-Error (MSE) values. Conclusion: The results obtained through the proposed technique ensure its supremacy compared to other state-of-the-art techniques. Significance: Hence, the proposed method can be utilised for EEG signal enhancement.