# facebookresearch/fastMRI

A large-scale dataset of both raw MRI measurements and clinical MRI images.

Repository: https://github.com/facebookresearch/fastMRI
Canonical: https://ross.abutalabs.com/products/fastmri
Homepage: https://fastmri.org
Language: Python
License: MIT
License Family: permissive
Topics: fastmri, mri, deep-learning, mri-reconstruction, convolutional-neural-networks, fastmri-dataset, fastmri-challenge, pytorch, medical-imaging
Archived: true
Last push: 2025-01-21T14:15:30+00:00

## Health v2 (maintenance only)
Score: 10/100 (v2, computed 2026-09-03T02:20:16.233290+00:00)
- activity 2, release rhythm 8, longevity 100
- inputs: {"age_days": 2868, "days_push": 589, "days_rel": null, "gap_med": null, "n_releases_24m": 0}
- flags: archived
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 1533, forks 422 (observed 2026-08-28T04:04:59.573409+00:00)

## What it is
fastMRI is a large-scale dataset of raw k-space MRI measurements and clinical MRI images released by Facebook AI Research and NYU Langone Health, along with PyTorch data loaders, subsampling functions, evaluation metrics, and baseline reconstruction models. It supports research into accelerated MRI reconstruction using machine learning.

## Use cases
- train deep learning models for MRI image reconstruction
- benchmark accelerated MRI reconstruction methods
- load raw k-space MRI data in PyTorch
- evaluate MRI reconstruction quality with standard metrics
- research undersampling strategies for MRI
- participate in the fastMRI challenge

## When to choose
- you need large-scale raw k-space and clinical MRI data for ML research
- you want reference implementations and baselines for MRI reconstruction
- you work with PyTorch and need ready-made MRI data loaders

## When to avoid
- you need a production clinical MRI reconstruction system
- you need other imaging modalities like CT or ultrasound
- you cannot accept the dataset access agreement from NYU

## Facets
- artifact type: dataset
- maturity: stable
- function: machine-learning, deep-learning, data-science, image-processing
- domain: machine-learning, healthcare, deep-learning, data-science
- platform: python
- tags: mri, medical-imaging, mri-reconstruction, pytorch, kspace, benchmark-dataset, convolutional-neural-networks

## Member repositories
- facebookresearch/fastMRI (main) score 10

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:04:59.573409+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T04:31:14.731911+00:00, confidence not recorded.
  - readme: https://github.com/facebookresearch/fastMRI (fetched 2026-08-28T04:04:59.573409+00:00, sha cb4b4b27b68a)
  - homepage: https://fastmri.org (fetched 2026-08-29T11:33:23.771657+00:00, sha 60c3ee0c41bf)
  - registry_pypi: https://pypi.org/pypi/fastmri/json (fetched 2026-08-29T11:33:23.781019+00:00, sha efccb69a7c75)
- Data as of 2026-08-30T08:39:29.467469+00:00.