The goal of this research partnership is to see if using AI can make MRI scans up to 10 times faster. MRIs work by gathering data and turning it into cross-sectional images of internal body structures, like organs and blood vessels. But as the area that needs to be scanned gets larger, so does the scan time. That's where AI might be able to help. The NYU and FAIR researchers want to speed up scans by collecting less raw data and having trained neural networks fill in the gaps. "The key is to train artificial neural networks to recognize the underlying structure of the images in order to fill in views omitted from the accelerated scan," said researchers in a blog post about the work.
The project will use around 10,000 clinical cases and 3 million MRI images that have been stripped of all patient information. The team plans to make the work open-source, making its AI models, baselines, evaluation metrics and image data sets available to other researchers.
If scan times can be significantly reduced without sacrificing accuracy, it could particularly benefit areas that have a shortage of scanners to begin with. Shorter scans could allow more patients to be imaged and faster MRIs could also take the place of X-rays and CT scans for some applications. Additionally, the researchers believe this use of AI could be applied to CT scans as well. "Such improvements would not only help transform the experience and effectiveness of medical imaging, but they'd also help equalize access to an indispensable element of medical care," they wrote.