Signal Processing¶

Coming soon

This page is currently under construction. The signal processing modules provide:

Extraction:

• extract_with_mask: Flatten spatial dimensions into a voxel array using a boolean mask.
• extract_with_labels: Extract region-averaged signals using an integer label map.
• unmask: Reconstruct a full spatial volume from a flat voxel array.

Preprocessing:

• censor_samples / interpolate_samples: Mark or interpolate corrupted samples (e.g., motion outliers).
• regress_confounds: Remove nuisance signals via least-squares regression.
• compute_compcor_confounds: Compute CompCor confounds from high-variance voxels.
• filter_butterworth: Apply low-pass, high-pass, or band-pass Butterworth filters.
• standardize: Scale signals to unit variance or percent signal change.
• detrend: Remove linear or polynomial trends from time series.
• clean: Apply full preprocessing pipeline in one step.

Please refer to the API Reference and Roadmap for more information.