Efficient EEG Channel-and-Frequency-Band Selection for Epileptic Seizure Classification Using Multi-Objective Optimization

This study optimizes EEG signal acquisition by jointly selecting the most relevant channels and frequency bands for seizure classification while reducing computational costs, using a genetic algorithm approach tested on the CHB-MIT database. Results identified specific channels (P3-O1, P4-O2, CZ-PZ) and frequency bands (gamma and alpha) as most relevant for accurate seizure detection in resource-constrained clinical settings.