Outputs

AutoHS writes BIDS derivatives under <output_dir>/autohs/. For interpretation of the metrics and thresholds in each report, see Background and theory.

Directory layout

output_dir/
├── work/                          # intermediate files (optional -w path)
└── autohs/
    ├── dataset_description.json   # BIDS derivative metadata
    ├── autohs_run.json            # run summary
    └── sub-<label>/
        └── ses-<label>/           # omitted when no session
            ├── sub-<label>_ses-<label>_desc-autohs_metrics.json
            ├── sub-<label>_ses-<label>_desc-autohs_summary.txt
            ├── sub-<label>_ses-<label>_desc-autohs_provenance.json
            ├── report.json        # legacy alias of metrics JSON
            └── figures/
                ├── sub-<label>_ses-<label>_desc-autohs_report.pdf
                └── sub-<label>_ses-<label>_desc-autohs_hippocampus-coronal*.png

Participant reports

*_desc-autohs_metrics.json

BIDS-style metrics file with Sources, SpatialReference, and GeneratedBy metadata plus hippocampal volumes, asymmetry index, and HS classification.

*_desc-autohs_summary.txt

Human-readable one-page summary suitable for clinical review.

figures/*_desc-autohs_report.pdf

Formatted PDF report with metrics and interpretation.

*_desc-autohs_provenance.json

AI-compute provenance sidecar with pipeline outputs and source T1w URI.

Quality control

Check autohs_run.json for the number of processed scans, pipeline backend (FreeSurfer vs FastSurfer), and completion timestamp.

Hippocampal sclerosis classification

Reports include two related layers (see Background and theory):

Volume laterality (±0.05): descriptive left/right imbalance.

HS screening (Brain Communications, in press): pathology-oriented thresholds —

Condition

Interpretation

AI > 0.046915816971433

Left-dominant (Right HS suspected)

AI < −0.070839747728063

Right-dominant (Left HS suspected)

otherwise

Balanced (No HS)

Note

“Left HS suspected” appears when volumes are right-dominant (smaller left hippocampus, more negative AI). See Background and theory for the clinical naming convention.

Volume laterality uses ±0.05 (AutoHS defaults).

Provenance

dataset_description.json records AutoHS version and citation text in HowToAcknowledge.