---
license: cc-by-nc-sa-4.0
tags:
  - medical-imaging
  - mri
  - segmentation
  - neuroscience
  - unetr
---

# GRACE — Whole-Head MRI Segmentation

GRACE is a UNETR-based model for automated whole-head MRI segmentation into 12 tissue classes, developed by the SMILE Lab.

## Variants

- `grace_native.pth` — operates in native MRI space
- `grace_fs.pth` — operates in FreeSurfer conformed space (256³ @ 1mm isotropic)

## Tissue Classes

See `labels.json` for the full 12-class label map.

## Usage

Install the CROWN CLI and download this model:

```bash
pip install crown-cli
crown models download grace-native
crown segment input.nii.gz --model grace-native
```

---
Citation
Stolte, S. E., Indahlastari, A., Chen, J., Albizu, A., Dunn, A., Pedersen, S., See, K. B., Woods, A. J., & Fang, R. (2024). Precise and Rapid Whole-Head Segmentation from Magnetic Resonance Images of Older Adults using Deep Learning. Imaging neuroscience (Cambridge, Mass.), 2, imag-2-00090. https://doi.org/10.1162/imag_a_00090

--- 
Terms
By downloading these weights you agree to use them for non-commercial research purposes only. Redistribution of the weights is not permitted.

---