README.md

---
dataset_info:
  - config_name: syntax
    features:
      - name: image
        dtype: image
      - name: mask
        dtype: image
      - name: image_id
        dtype: int64
      - name: file_name
        dtype: string
    splits:
      - name: train
        num_examples: 1000
      - name: val
        num_examples: 200
      - name: test
        num_examples: 300
  - config_name: stenosis
    features:
      - name: image
        dtype: image
      - name: mask
        dtype: image
      - name: image_id
        dtype: int64
      - name: file_name
        dtype: string
    splits:
      - name: train
        num_examples: 1000
      - name: val
        num_examples: 200
      - name: test
        num_examples: 300
configs:
  - config_name: syntax
    data_files:
      - split: train
        path: syntax/train-*
      - split: val
        path: syntax/val-*
      - split: test
        path: syntax/test-*
  - config_name: stenosis
    data_files:
      - split: train
        path: stenosis/train-*
      - split: val
        path: stenosis/val-*
      - split: test
        path: stenosis/test-*
license: cc0-1.0
task_categories:
  - image-segmentation
tags:
  - medical-imaging
  - coronary-artery
  - x-ray-angiography
  - SYNTAX-score
pretty_name: ARCADE
size_categories:
  - 1K<n<10K
---

# ARCADE

Automatic Region-based Coronary Artery Disease diagnostics using X-ray angiography imagEs.

## Description

X-ray coronary angiography (XCA) dataset for coronary artery segmentation and stenosis detection. Contains 3,000 grayscale 512x512 images across two tasks.

## Subsets

- **syntax**: Coronary artery segment classification into 26 anatomical regions per SYNTAX Score methodology (1,500 images)
- **stenosis**: Atherosclerotic plaque / stenotic lesion detection and segmentation (1,500 images)

## Splits

| Split | Images per subset |
|-------|-------------------|
| Train | 1,000 |
| Val   | 200 |
| Test  | 300 |

## Mask Labels

**Syntax** (26 classes): Coronary artery segments 1-16c per SYNTAX Score methodology.

**Stenosis** (26 classes): Same artery segments (1-25) plus class 26 = stenosis lesion.

Pixel value 0 = background.

## Source

- **Paper**: Popov et al., "Dataset for Automatic Region-based Coronary Artery Disease Diagnostics Using X-Ray Angiography Images", *Scientific Data* 11, 20 (2024). DOI: 10.1038/s41597-023-02871-z
- **Zenodo**: https://zenodo.org/records/10390295
- **License**: CC0 1.0 Universal