Add group-averaged segstats maps to template space by contrast

[Copilot]

Extends group analysis to generate volumetric maps of averaged segmentation metrics (fieldfrac, density, count, volume) per contrast group, analogous to existing voxel-wise count maps from parquet tables.

Changes

• New script concat_subj_segstats_contrast.py: Concatenates subject-level segstats.tsv files filtered by contrast, computes region-wise group averages

• New rule concat_subj_segstats_contrast: Collects mergedsegstats.tsv across subjects, filters by contrast column/value, outputs averaged TSV per region

• New rule map_groupavg_segstats_to_template_nii: Maps averaged TSV to template NIfTI using existing map_tsv_dseg_to_nii.py

• Updated all_group_stats target: Generates group-averaged NIfTI maps for all atlas segs × stain metrics × contrast values

Outputs

Per contrast group:

• *_contrast-<column>+<value>_groupavgsegstats.tsv: Averaged metrics per brain region
• *_contrast-<column>+<value>_metric-<stain>+<metric>_groupavg.nii.gz: 3D maps of averaged metrics in template space

Complements existing statistical comparison maps (t-stats, p-values) with actual group-averaged values for visualization.

Original prompt

right now we have voxel-wise group-means created from parquet tables e.g.:

    counts_nii=bids(
        root=root,
        datatype="group",
        space="{template}",
        level="{level}",
        desc="{desc}",
        contrast="{contrast_column}+{contrast_value}",
        suffix="coloccount.nii",
    ),

but I would also like segstats.tsv tables to be averaged over groups in the same way, using the script "../scripts/map_tsv_dseg_to_nii.py" to write group avg maps in template space, for each atlas seg. add new rules to the groupstats smk file.

---

✨ Let Copilot coding agent set things up for you — coding agent works faster and does higher quality work when set up for your repo.