Signal Decomposition and Noise Reduction in Single-Channel EEG: A Morphological Component Analysis (MCA) Approach
Kosei Shibata, Yide Yang, Rena Kato, Hendry Ferreira Chame, Laurent Bougrain, Hiroaki Wagatsuma · Proceedings of International Conference on Artificial Life and Robotics · 2025
This study applies the Morphological Component Analysis (MCA) to single-channel EEG data obtained during human-to-human interactions in a board game (Hex-game).MCA, a dictionary-based signal decomposition method, separates signals into distinct morphological components.It enables the extraction of plausible brain activity and the removal of noise, such as ocular and muscular artifacts.By focusing on neural dynamics in interactive settings, this approach highlights the relationship between cognitive processes and social behavior.The approach suggests that MCA offers a promising framework for EEG analysis in complex, dynamic environments, combining effective feature extraction with robust artifact removal.