Version 0.4.2🔗

Enhancements🔗

Add native 3D PyVista scraper support in Sphinx-Gallery for headless 3D sensor coregistration and cortical source plots (#157 by Tomas Ros).
Introduce comprehensive Optically Pumped Magnetometer (OPM) preprocessing tutorial comparing the standard MNE-Python workflow to single-step GEDAI (#157 by Tomas Ros).
Automatically pre-fetch UCL OPM auditory dataset in documentation CI workflow to ensure reliable builds (#157 by Tomas Ros).
Add Seymour et al. (2022) citation on OPM MEG movement and speech to the bibliography (#157 by Tomas Ros).

Fix n_pc propagation in fit_epochs() so user-specified component counts are forwarded to both broadband and per-band models (#157 by Tomas Ros).
Validate n_pc before scoring in Gedai and compute_composite_sensai to reject non-positive or over-channel component counts (#157 by Tomas Ros).
Treat ref_meg channels as MEG in _detect_signal_type for correct signal classification (#157 by Tomas Ros).
Widen reference_cov type annotations in AdaptiveMultibandGedai and MultibandGedai to support mne.Covariance and mne.Forward (#157 by Tomas Ros).
Align primary auditory cortex (Heschl’s gyrus / STG) dipole peak latency and time courses in OPM tutorial to target N100m latency (#157 by Tomas Ros).
Fix typo in multiband and adaptive tutorial callouts (#157 by Tomas Ros).