eeg-microstate

You can quantify EEG microstate dynamics by fitting modified k-means models to GFP peaks in cleaned resting-state EEG, which outputs coverage, GEV, mean duration, occurrence, and transition probability matrices automatically.

Standard EEG microstate segmentation analysis includes coverage, global explained variance (GEV), mean duration, occurrence, and a transition probability matrix to characterize scalp topography temporal dynamics across conditions.

Run modified k-means microstate segmentation by configuring the pycrostates backend and fitting models to GFP peaks in your continuous EEG data, using configurable K values like the default Koenig-style 4 canonical templates.

Yes, computing microstate parameters requires pre-cleaned or continuous EEG data available in clean-stage or epochs-stage formats, plus a frozen ANALYSIS_PLAN.md and a resolved pycrostates backend configured via ENVIRONMENT.json.

The best way to compare EEG microstate transitions across conditions is to fit and back-fit microstate templates per subject, generating transition probability matrices and group summary CSV files for direct statistical comparison.

Your pycrostates EEG microstate analysis may fail to execute if the backend is unresolved in ENVIRONMENT.json, or if the ANALYSIS_PLAN.md is not frozen and cleaned continuous EEG data is absent from clean-stage or epochs-stage directories.