CorpusCallosum.paint_cc_into_pred

CorpusCallosum.paint_cc_into_pred.argument_parse()

Create a command line interface and return command line options.

CorpusCallosum.paint_cc_into_pred.correct_wm_ventricles(aseg_cc, fornix_mask, voxel_size, close_gap_size_mm=3.0)

Fill small gaps between corpus callosum, ventricles, and background.

This function performs two gap-filling operations: 1. Fills WM and background gaps between CC and ventricles with ventricle labels 2. Fills WM gaps between CC and background with background label

Note: Fornix and non-CC-connected WM component removal are intentionally not implemented in this function as they have been removed from the processing pipeline.

Parameters:

aseg_ccnp.ndarray

Aseg segmentation with CC already painted in.

fornix_masknp.ndarray

Mask of the fornix. Not currently used (kept for interface compatibility).

voxel_sizetuple[float, float, float]

Voxel size of the aseg image in mm.

close_gap_size_mmfloat, default=3.0

Maximum size of the gap to fill in millimeters.

Returns:

np.ndarray

Corrected segmentation map with filled gaps.

CorpusCallosum.paint_cc_into_pred.correct_wm_ventricles_in_lia(aseg_cc, fornix_mask, affine, voxel_size, close_gap_size_mm=3.0)

Run WM/ventricle gap correction in LIA orientation and map back.

Parameters:

aseg_ccnp.ndarray

Segmentation with CC already painted in, in native orientation.

fornix_masknp.ndarray

Fornix mask in native orientation.

affinenp.ndarray

Native image affine.

voxel_sizetuple[float, float, float]

Native voxel sizes.

close_gap_size_mmfloat, default=3.0

Maximum size of the gap to fill in millimeters.

Returns:

np.ndarray

Corrected segmentation map in the original/native orientation.

CorpusCallosum.paint_cc_into_pred.paint_in_cc(pred, aseg_cc)

Paint corpus callosum segmentation into aseg+dkt segmentation map.

Parameters:

prednp.ndarray

Deep-learning segmentation map.

aseg_ccnp.ndarray

Aseg segmentation with CC.

Returns:

np.ndarray

Segmentation map with added CC.