Line noise

Dealing with line noise in MEG data requires effective suppression methods to avoid compromising data quality. While traditional low-pass or notch filters are commonly used, they can eliminate valuable information in the line noise frequency range, limiting their applicability. Advanced techniques like frequency-domain regression (e.g., CleanLine) offer alternatives but are less effective in handling amplitude and phase fluctuations. Recent methods like Zapline and its enhanced version, ZapLine-Plus, provide more refined approaches, addressing non-stationary noise and automating parameter selection. MEGAP integrates ZapLine-Plus using MATLAB-Engine for processing. To further suppress residual noise, a regression-based method is applied post-ZapLine-Plus, optimized for speed and relevance by focusing on the same frequency range.

Comparison between the effect of different stages of line noise removal (A) before removing any line noise, (B) after removing line noise with ZapLine_plus and (C) using regression-based method in addition to ZapLine_plus