Optimized Digital Notch Filter Design for Enhanced EEG Medical Diagnostics Using Modified Particle Swarm Optimization（MPSO）

Abstract

Electroencephalography (EEG) signals are critical for understanding brain activity, diagnosing neurological disorders, and developing brain-computer interface systems. However, these signals are highly susceptible to powerline noise, including its harmonics, which compromises the accuracy of diagnosis and analysis. This paper presents a novel approach for designing digital notch filters using a Modified Particle Swarm Optimization (MPSO) algorithm. The proposed method enhances stopband attenuation, reduces transition width, and minimizes ripples in passband and stopband, making it ideal for applications requiring high precision. Simulation results demonstrate the superiority of MPSO compared to conventional PSO and other established methods like Parks-McClellan and Real-Coded Genetic Algorithms. The proposed filter effectively removes 60 Hz noise and its harmonics, ensuring cleaner EEG signals for better diagnostic and analytical reliability.