---
license: apache-2.0
base_model: blueyo0/SAM-Med3D
datasets:
  - pcarnahan/MVSeg2023
pipeline_tag: image-segmentation
library_name: pytorch
tags:
  - medical-imaging
  - ultrasound
  - echocardiography
  - segmentation
  - sam-med3d
  - segment-anything
  - 3d
  - mitral-valve
model-index:
  - name: UltraSAM-3D-Ultrasound-MVSeg
    results:
      - task:
          type: image-segmentation
        dataset:
          name: MVSeg 2023
          type: pcarnahan/MVSeg2023
        metrics:
          - type: dice
            value: 0.87
            name: Validation Dice
---

# UltraSAM — SAM-Med3D fine-tuned for 3D ultrasound (mitral valve)

A fine-tuned [SAM-Med3D](https://github.com/uni-medical/SAM-Med3D) checkpoint for
**promptable segmentation of the mitral valve in 3D echocardiography**, trained on
the MVSeg 2023 dataset with a boundary-aware objective.

🏆 Part of the **UltraSAM** project — 4th place at **AI for Life 2026** Hackathon.

- **Code, training & interactive MONAI Label app:** https://github.com/thelet/UltraSAM-3D-Ultrasound-Segmentation
- **Base model:** SAM-Med3D (`vit_b_ori`, 128³ input)
- **Task:** interactive (click-prompted) binary segmentation
- **Modality:** 3D ultrasound / echocardiography

## Results

Predicted mitral-valve segmentation on a held-out test case, shown in 3D Slicer.
**Pink = model prediction, yellow = ground truth.** Top: 3D surface renderings;
bottom: axial / sagittal / coronal slices with overlays.

<p align="center">
  <img src="https://cdn-uploads.huggingface.co/production/uploads/6967bc554f6963a866d40495/m6rAelYl1mB66vZqk07z7.jpeg" width="700" alt="Qualitative result: prediction vs. ground truth">
</p>

## Usage
```python
import medim
model = medim.create_model(
    "SAM-Med3D", pretrained=True,
    checkpoint_path="sam_model_dice_best.pth",
)
```
See the [GitHub repo](https://github.com/thelet/UltraSAM-3D-Ultrasound-Segmentation)
for the full inference pipeline (`run_infer_no_baseline.py`) and the interactive
3D Slicer workflow.

## Training

Fine-tuned for 200 epochs on MVSeg 2023 with a combined **DiceCE + 0.5·Boundary**
objective. Validation Dice plateaus around **~0.87**.

| Train Dice | Validation Dice |
|---|---|
| ![Train Dice](https://cdn-uploads.huggingface.co/production/uploads/6967bc554f6963a866d40495/0BxDZ-Quxtef-v0ihVBA2.png) | ![Validation Dice](https://cdn-uploads.huggingface.co/production/uploads/6967bc554f6963a866d40495/ysZLpgPeQVWgXo4vIafBu.png) |


## Attribution & license

This model is a fine-tuned derivative of **SAM-Med3D** (Apache-2.0) and is released
under the same license. The base weights and architecture are the work of the
SAM-Med3D authors; this repository modifies them by fine-tuning on 3D
echocardiography (MVSeg 2023) with a boundary-aware objective.

- Base model: [SAM-Med3D](https://github.com/uni-medical/SAM-Med3D) (Apache-2.0)
- SAM-Med3D builds on Meta AI's [Segment Anything](https://github.com/facebookresearch/segment-anything) (Apache-2.0)
- Dataset: [MVSeg 2023](https://www.synapse.org/mvseg2023)

### Citation
```bibtex
@article{wang2023sammed3d,
  title={SAM-Med3D},
  author={Wang, Haoyu and Guo, Sizheng and Ye, Jin and Deng, Zhongying and
          Cheng, Junlong and Li, Tianbin and Chen, Jianpin and Su, Yanzhou and
          Huang, Ziyan and Shen, Yiqing and Fu, Bin and Zhang, Shaoting and
          He, Junjun and Qiao, Yu},
  journal={arXiv preprint arXiv:2310.15161},
  year={2023}
}
```