fastMRI (knee and brain)
fastMRI: raw k-space and DICOM dataset of knee and brain MRI
Raw multi-coil k-space of 1,594 knee and 6,970 brain MRI scans from NYU Langone, plus about 10,000 clinical knee MRI exams as DICOM, released with Meta AI (FAIR) to benchmark machine-learning reconstruction of accelerated MRI.
Overview
fastMRI is a joint project of NYU Langone Health and Facebook AI Research (now Meta AI) that released one of the first large collections of raw MRI scanner data. It was built to train and compare machine-learning methods that reconstruct images from undersampled k-space, so that MRI exams can be made faster, and it powered the fastMRI reconstruction challenges.
Composition
The knee part holds fully sampled multi-coil k-space from 1,594 clinical knee scans, each a coronal proton-density volume with or without fat suppression, plus an emulated single-coil version. It is split into 973 training, 199 validation and 118 multi-coil test volumes, with further single-coil test and held-back challenge volumes; test and challenge volumes come undersampled and without ground truth. A separate DICOM collection adds 10,012 consecutive clinical knee exams from 9,290 patients with up to five standard sequences (coronal PD with and without fat suppression, sagittal PD, sagittal and axial fat-suppressed T2). The brain part holds raw k-space for 6,970 axial scans: T1-weighted with and without contrast agent, T2-weighted and FLAIR. Not every brain exam has every contrast. Prostate and breast sets released later are not covered here.
Acquisition
All raw data come from Siemens scanners at NYU: for the knee three 3 T systems (Skyra, Prisma, Biograph mMR) and a 1.5 T Aera with a 15-channel knee coil and a clinical 2D turbo spin echo protocol; for the brain 11 magnets at five locations, at 1.5 T (Avanto, Aera) and 3 T (Prisma, Skyra, Biograph, Tim Trio). Raw data were converted to the vendor-neutral ISMRMRD format and shipped as one HDF5 file per volume. The DICOM images come from a wider range of scanners and are mostly reconstructions of accelerated acquisitions.
Annotations
No clinical labels are included. The reconstruction targets are root-sum-of-squares images of the fully sampled data.
Known limitations
- Brain k-space slices near and below the orbits were zeroed for de-identification, and only axial 2D brain data were released.
- All raw data come from one vendor and one health system.
- No age, sex or diagnosis information is published for the cohort.
- Counts differ slightly between the paper and the website (for example 6,970 vs 7,002 brain scans).
Cohort
Aggregate numbers from the sources below. Bars are relative to the largest value.
Anatomy
subjects
- Knee 9,290
License and access
Our reading of the license, not legal advice. Before you use the data, read the original license and confirm that your use is allowed. We take no responsibility for how you use a dataset. Full disclaimer
Sign a data use agreement, often reviewed by the provider
NYU Langone Health fastMRI Dataset Sharing Agreement
NYU grants a royalty-free license for internal research or education only, after an online application with an electronic signature. No selling, commercial exploitation or passing on of the data or of files derived from it, except to people under your direct supervision and in cited academic publications. Copies must be destroyed when the work is done.
The website says each sub-dataset has its own Data Sharing Agreement; the agreement text shown on the page is the one summarised here. The prostate and breast sub-datasets are not part of this entry.
What you can do
- No
- Conditional
- Yes
- Conditional
What you can share
- No
- No
- Not stated
What you must do
- Yes
- Share alike No
- Yes
- Conditional
- Manuscript review No
- Release code No
- Return results No
- Yes
Limits
- Yes
- Conditional
Citation
Knoll F, Zbontar J, Sriram A, et al. fastMRI: A Publicly Available Raw k-Space and DICOM Dataset of Knee Images for Accelerated MR Image Reconstruction Using Machine Learning. Radiol Artif Intell 2(1):e190007 (2020). doi:10.1148/ryai.2020190007; and Zbontar J, Knoll F, Sriram A, et al. fastMRI: An Open Dataset and Benchmarks for Accelerated MRI. arXiv:1811.08839.
All numbers
Every number on this page, as stored in stats.csv, with its source.
| Measure | Breakdown | Value | Source |
|---|---|---|---|
| Scans | total raw volumes; the knee (1594) and brain (6970) sections add up to 8564 | 8,344 | zbontar2018 Section 1 (p. 2) |
| Scans | anatomy=knee fully sampled multi-coil raw knee scans (one volume each); DICOM exams are counted under studies | 1,594 | zbontar2018 Section 4.2 (p. 7) |
| Scans | anatomy=knee;contrast=PDw coronal PD with fat suppression (798) and without (796) | 1,594 | zbontar2018 Section 4.2 (p. 7) |
| Scans | anatomy=knee;field_strength=1.5 raw knee scans on Siemens Aera 1.5T; the rest on 3T Skyra (663) Prisma (83) and Biograph mMR (153) | 695 | zbontar2018 Table 2 (p. 8) |
| Scans | anatomy=knee;split=train raw knee volumes (multi-coil and single-coil) | 973 | zbontar2018 Table 4 (p. 10) |
| Scans | anatomy=knee;split=validation raw knee volumes (multi-coil and single-coil) | 199 | zbontar2018 Table 4 (p. 10) |
| Scans | anatomy=knee;split=test multi-coil test volumes; a further 108 single-coil test and 196 challenge volumes are held back | 118 | zbontar2018 Table 4 (p. 10) |
| Scans | anatomy=brain fully sampled multi-coil raw brain scans; the website table now lists 7002 | 6,970 | zbontar2018 Section 4.3 (p. 7) |
| Scans | anatomy=brain;field_strength=1.5 | 3,001 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;field_strength=3 | 3,969 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T1w;field_strength=1.5 T1 without contrast | 375 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T1w;field_strength=3 T1 without contrast | 407 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T1w_ce;field_strength=1.5 T1 POST | 849 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T1w_ce;field_strength=3 T1 POST | 641 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T2w;field_strength=1.5 | 1,651 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=T2w;field_strength=3 | 2,515 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=FLAIR;field_strength=1.5 | 126 | zbontar2018 Table 3 (p. 8) |
| Scans | anatomy=brain;contrast=FLAIR;field_strength=3 | 406 | zbontar2018 Table 3 (p. 8) |
| Studies | anatomy=knee consecutive clinical knee DICOM image datasets (the website rounds to 10,000) | 10,012 | knoll2020 Description of the Dataset, DICOM Dataset |
| Subjects | anatomy=knee patients in the knee DICOM set only; patient counts for the raw data are not reported | 9,290 | knoll2020 Description of the Dataset, DICOM Dataset |
Sources
The keys used in the table above.
- zbontar2018 Zbontar et al., fastMRI - An Open Dataset and Benchmarks for Accelerated MRI (arXiv:1811.08839v2, pages of that version) paper
- knoll2020 Knoll et al. 2020, Radiology - Artificial Intelligence (PMC6996599) paper
- fastmri-nyu fastMRI dataset website website