As a part of their FastMRI project to use artificial intelligence to speed up MRI procedures, New York University and Facebook have just introduced a first-of-its-kind massive open-source MRI dataset, which includes 1.5 million anonymous MR images of the knee.

Back in August, Facebook AI Research joined the Department of Radiology at NYU Langone Health’s FastMRI, a project to make MRI scans 10 times faster through using AI models, baselines, and evaluation metrics. The researchers are devising ways to create scans with minimal data and then using AI to fill in the gaps. The objective of the project is to expand AI resources for the medical imaging field and create opportunities for research reproducibility.

The knee images were generated from 10,000 scans and 1,600 scans with “raw measurement data.” Researchers are calling this the “largest public release of raw MRI data to date.” The knee data is a part of the first phase of the project, but sequential phases will feature data from liver and brain scans. The dataset is fully compliant with HIPAA standards. No personal Facebook information was used, and all data is anonymized and there are no markers of individual patients’ identity.

“We hope that the release of this landmark data set, the largest-ever collection of fully-sampled MRI raw data, will provide researchers with the tools necessary to overcome the challenges inherent in accelerating MR imaging,” said Michael P. Recht, MD, chair and the Louis Marx Professor of Radiology at NYU Langone Health. “This work has the potential to not only help increase access to MR imaging, but also improve patient care worldwide,”