Multimodal EEG-fNIRS Fusion for Passive BCI-based Depressive State Classification
Riki Sakurai, Simon Kojima, Mihoko Otake, Shin’ichiro Kanoh, Tomasz M. Rutkowski
This paper develops a multimodal EEG-fNIRS pBCI system to classify depressive states using BDI-II scores, achieving 90.9% balanced accuracy and 0.867 F1-score through SincShallowNet.
bioRxiv · Apr 8, 2026 · PreprintNeuroscience