Artificial intelligence can reconstruct rapid MRIs into higher-quality images than traditional scans, according to a recent study that came out of a partnership between New York City-based NYU Langone Health and Facebook parent company Meta.
Titled “Deep Learning Reconstruction Enables Prospectively Accelerated Clinical Knee MRI”, the study, published in the journal Radiology, suggests that using AI to "fill in" missing parts of the scans may expand the availability of the technology and lower appointment wait times. FastMRIs are four times faster than standard ones and take about 5 minutes compared to a half hour for a traditional one, the researchers say.
The study is a part of the fastMRI initiative established by NYU Langone Health and Meta AI Research in 2018. This initiative resulted in an AI model jointly developed by Meta AI researchers and NYU Langone imaging scientists and radiologists. It also produced the largest-ever publicly-available collection of raw MRI data, which has been used by scientists and engineers around the world.
The Study
In the study, a total of 170 participants received a diagnostic knee MRI using a conventional MRI protocol followed by an accelerated AI protocol between January 2020 and February 2021. Each examination was reviewed by six musculoskeletal radiologists, who looked for signs of meniscal or ligament tears and bone marrow or cartilage abnormalities.
The radiologists evaluating the scans were not told which images were reconstructed with AI, and to limit the potential for recall bias, the evaluations of the standard images and AI-accelerated images were spaced at least four weeks apart.
Results
The radiologists judged the AI-reconstructed images to be diagnostically equivalent to the conventional images for detecting tears or abnormalities, and found the overall image quality of the accelerated scans to be significantly better than the conventional images.
FastMRI, the researchers note, does not require special equipment. Technicians can program standard MRI machines to gather less data than is usually required, and the fastMRI initiative has made its data, models, and code available as an open-source project for other researchers, as well as manufacturers of commercial MRI systems.
Join the HealthXL Masterclass on ‘How to Make Value-Based Health Care (VBHC) work with Digital Health?’ on 2nd February. Click here to Request to Join.
Click here to read the original news story.