# ISLES 2022: ischemic stroke lesion segmentation MRI (DWI and ADC)

## Facts

- Provider: Moritz Hernandez Petzsche et al., ISLES 2022 challenge (converted by MedOtter)
- License: CC-BY-4.0 (https://zenodo.org/records/7153326)
- Size: 753 files
- Format: NIfTI, JSON Lines (in sampled files)
- Stored at: Hugging Face, https://huggingface.co/datasets/MedOtter/isles2022
- Access: public, readable without an account
- DOI: https://doi.org/10.5281/zenodo.7153326
- Category: Neuroscience
- Source: https://doi.org/10.1038/s41597-022-01875-5
- Page: https://scigantic.com/archive/73c9e4f3-cf05-5c14-a011-40e55cbb6566
- Keywords: Neuroscience, MRI, image segmentation, NIfTI, JSON Lines, Hugging Face

## About

ISLES 2022 is the multi-centre magnetic resonance imaging stroke lesion segmentation dataset of the MICCAI 2022 Ischemic Stroke Lesion Segmentation challenge, published in Scientific Data (Hernandez Petzsche et al. 2022). It contains acute to subacute ischemic stroke MRI from several centres with expert manual infarct masks. This MedOtter conversion packages the diffusion sequences used for segmentation, apparent diffusion coefficient (ADC) and diffusion-weighted imaging (DWI), together with the lesion masks as NIfTI (.nii.gz) files, and a JSONL index giving image path, mask path, modality, official split and patient ID for each case. The FLAIR sequence of the original release is not part of this conversion. As measured 2026-11-14 the repository uses about 0.6 GB; read the index with datasets or pandas and the volumes with nibabel, SimpleITK or MONAI. Used for stroke lesion segmentation, infarct volume estimation and as one of the standard tasks in medical segmentation foundation model benchmarks. The ISLES organisers release the data on Zenodo under CC BY 4.0; cite Hernandez Petzsche et al. 2022.

## Load the data

```python
from datasets import load_dataset
ds = load_dataset("MedOtter/isles2022")
print(ds)
```

## Data structure

File types in sampled files: NIfTI, JSON Lines

Example file: `train.jsonl`

## Open it

- In a hosted notebook with the data mounted: https://scigantic.com/archive/73c9e4f3-cf05-5c14-a011-40e55cbb6566
- Full record as JSON (schema card included): https://api.scigantic.com/api/archives/73c9e4f3-cf05-5c14-a011-40e55cbb6566
- Over MCP: `get_schema_card` with id `73c9e4f3-cf05-5c14-a011-40e55cbb6566` at https://api.scigantic.com/mcp

## Related topics

- [Neuroscience datasets](https://scigantic.com/topics/neuroscience)
- [MRI datasets](https://scigantic.com/topics/mri)
- [Image segmentation datasets](https://scigantic.com/topics/image-segmentation)
- [JSON Lines datasets](https://scigantic.com/topics/json-lines)
- [NIfTI datasets](https://scigantic.com/topics/nifti)
