Facebook, NYU School of Medicine to use AI to make MRI scans 10 times faster

Social media giant Facebook and the NYU School of Medicine have started a new collaborative research project aimed at using artificial intelligence to make magnetic resonance imaging scans 10 times faster.

The effort, called fastMRI, is meant to speed up notoriously slow MRI machines that can take from 15 minutes to more than an hour to conduct scans, vs. less than a second or as long as a minute—respectively—for X-ray and CT scans.

“This project will initially focus on changing how MRI machines operate,” wrote Larry Zitnick from the Facebook Artificial Intelligence Research group and NYU School of Medicine’s Michael Recht, MD, and Daniel Sodickson, MD, in a blog announcing the initiative. “Using AI, it may be possible to capture less data and therefore scan faster, while preserving or even enhancing the rich information content of magnetic resonance images. 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.”

Recht is chair of the Department of Radiology at NYU School of Medicine and Sodickson is vice chair for research and director of the Center for Advanced Imaging Innovation and Research, which includes a multidisciplinary team focused on developing novel imaging technologies and translating them into clinical practice.

MRI-CROP.jpg

Also See: Faster CPU can accelerate ability to use AI on medical images

NYU’s Center for Advanced Imaging Innovation and Research has been working with the Facebook Artificial Intelligence Research group on the enabling technology.

While Facebook data will not be used as part of the project, fastMRI will leverage a de-identified imaging dataset collected by NYU that includes 10,000 clinical cases and about 3 million magnetic resonance images of the knee, brain and liver.

As the project progresses, researchers say Facebook will share the AI models, baselines and evaluation metrics with the broader research community, while the NYU School of Medicine will open-source the image dataset.

“We believe the fastMRI project will demonstrate how domain-specific experts from different fields and industries can work together to produce the kind of open research that will make a far-reaching and lasting positive impact in the world,” the researchers concluded, including ultra-low-dose CT scans suitable for vulnerable populations such as pediatric patients.

“With the goal of radically changing the way medical images are acquired in the first place, our aim is not simply enhanced data mining with AI, but rather the generation of fundamentally new capabilities for medical visualization to benefit human health,” they added.

For reprint and licensing requests for this article, click here.