SENSI EEG PREPROC — Bad-Channel Detection Module

Version v1.0.0 — Initial Public Release

This is the first public release of the Bad-Channel Detection Module.

This version corresponds to the codebase described in the accompanying preprint:

Amilcar J. Malave and Blair Kaneshiro (2026). EEG Bad-Channel Detection Using Multi-Feature Thresholding and Co-Occurrence of High-Amplitude Transients. bioRxiv. https://doi.org/10.64898/2026.02.04.703874

Dataset:

Amilcar J. Malave and Blair Kaneshiro (2025). Example EEG data for the SENSI EEG PREPROC Bad-Channel Detection Module [Data set]. Stanford Digital Repository. https://doi.org/10.25740/dg856vy8753

Included in this Release

• markSusChs.m (main user entry point)
• Multi-feature suspiciousness scoring:
  • Neighbor dissimilarity
  • Amplitude screening
  • Variance-based measures
• High-amplitude transient clustering via Jaccard-like similarity
• Interactive review interface (reviewBadChsUI)
• Example workflow (example.m)
• User Manual (PDF)

Intended Usage

This Module is designed as a quality-control step prior to ICA and downstream EEG analyses.
It emphasizes interpretability and human-in-the-loop validation rather than fully automated rejection.

MATLAB Requirements

• MATLAB R2024b (tested)
• Statistics and Machine Learning Toolbox

Notes

This release represents the first stable public version of the Module.
Future releases may refine clustering behavior, visualization outputs, and parameter defaults.

This version should be cited when referencing the v1.0.0 implementation.