README.md

---
license: cc-by-nc-sa-4.0
task_categories:
  - image-segmentation
tags:
  - medical-imaging
  - ultrasound
  - fetal-imaging
  - segmentation
  - miccai-2024
pretty_name: ACOUSLIC-AI
size_categories:
  - n<1K
---

# ACOUSLIC-AI Dataset

## Dataset Description

ACOUSLIC-AI (Abdominal Circumference Operator-agnostic UltraSound measurement in Low-Income Countries) is a dataset for the MICCAI 2024 challenge focused on fetal abdominal circumference measurement from ultrasound images.

### Dataset Summary

- **Source**: [Zenodo](https://zenodo.org/records/12697994)
- **Challenge**: [ACOUSLIC-AI Grand Challenge](https://acouslic-ai.grand-challenge.org/)
- **Version**: 1.1
- **License**: CC-BY-NC-SA 4.0

### Supported Tasks

- Fetal abdomen segmentation
- Abdominal circumference measurement
- Ultrasound image analysis

## Dataset Structure

```
.
├── images/
│   └── stacked_fetal_ultrasound/
│       └── *.mha (300 files, ~335MB each)
├── masks/
│   └── stacked_fetal_abdomen/
│       └── *.mha (300 files, ~335MB each)
└── circumferences/
    └── fetal_abdominal_circumferences_per_sweep.csv
```

### Data Fields

**Images (.mha)**:
- 3D volumetric ultrasound stacks
- Format: MetaImage (.mha)
- Each file: ~335MB

**Masks (.mha)**:
- Segmentation masks with pixel values:
  - 0: Background
  - 1: Optimal plane (fetal abdomen)
  - 2: Suboptimal plane

**Circumferences (CSV)**:
- `uuid`: Unique identifier matching image/mask filenames
- `subject_id`: Subject identifier
- `sweep_1_ac_mm` to `sweep_6_ac_mm`: Abdominal circumference measurements in mm

## Dataset Statistics

| Type | Count | Format | Size |
|------|-------|--------|------|
| Images | 300 | .mha | ~335MB each |
| Masks | 300 | .mha | ~335MB each |
| CSV | 1 | .csv | 24KB |

**Total size**: ~200GB

## Usage

```python
from huggingface_hub import snapshot_download
import SimpleITK as sitk

# Download dataset
local_dir = snapshot_download(
    repo_id="Angelou0516/ACOUSLIC-AI",
    repo_type="dataset"
)

# Load an image
image_path = f"{local_dir}/images/stacked_fetal_ultrasound/0199616b-bdeb-4119-97a3-a5a3571bd641.mha"
image = sitk.ReadImage(image_path)
array = sitk.GetArrayFromImage(image)  # (D, H, W) numpy array

# Load corresponding mask
mask_path = f"{local_dir}/masks/stacked_fetal_abdomen/0199616b-bdeb-4119-97a3-a5a3571bd641.mha"
mask = sitk.ReadImage(mask_path)
mask_array = sitk.GetArrayFromImage(mask)
```

## Citation

If you use this dataset, please cite the original source:

```bibtex
@misc{acouslic_ai_2024,
  title={ACOUSLIC-AI: Abdominal Circumference Operator-agnostic UltraSound measurement in Low-Income Countries},
  year={2024},
  publisher={Zenodo},
  doi={10.5281/zenodo.12697994},
  url={https://zenodo.org/records/12697994}
}
```

## License

This dataset is licensed under [CC-BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/).

## Acknowledgements

This dataset was created for the MICCAI 2024 ACOUSLIC-AI Challenge.