Hi all, my team at MindsApplied recently developed a physics-based filter package that improved our downstream AI classification more than alternatives like PCA or ASR, and performed particularly well on dynamic artifacts in real-time. We wanted to share it with the community for feedback and to advance real-time commercial BCI development!
Preprint Link: A lightweight, physics-based, sensor-fusion filter for real-time EEG denoising and improved downstream AI classification
Code: GitHub Package and Test App (core.py of the file ‘mindsai_filter_python’)