The New York University School of Medicine is collaborating with an unlikely partner to make MRI scans up to 10 times faster: Facebook.
The medical school is less interested in Facebook’s timeline, and more focused on how advance algorithms and artificial intelligence could solve a nuanced but impactful issue in healthcare.
Researchers with NYU’s Department of Radiology announced the new initiative alongside Facebook's lead AI researcher to identify ways to improve the speed of MRIs, which can require patients to remain in one position for more than an hour.
In a blog post NYU School of Medicine’s Daniel Sodickson, M.D., Ph.D , and Michael Recht, M.D. and Larry Zinick, who leads the Facebook Artificial Intelligence Research (FAIR) group, said AI could speed up MRI procedures by training neural networks to recognize and fill in date omitted by a faster scan.
RELATED: NYU will offer medical students free tuition
Sodickson serves as the vice chair for research and director of the Center for Advanced Imaging Innovation and Research at NYU, which has been pursuing the idea of using AI for MRIs since 2016. Recht chairs the Department of Radiology.
The project leads acknowledged they are taking on “an exceedingly hard problem,” since inaccurate images could alter images, leading radiologists to miss vital medical issues.
“A few missing or incorrectly modeled pixels could mean the difference between an all-clear scan and one in which radiologists find a torn ligament or a possible tumor,” they wrote. “Conversely, capturing previously inaccessible information in an image can quite literally save lives.”
Facebook has been plotting an entrance into the healthcare industry for some time. In April, CNBC reported that Facebook asked several major U.S. hospitals to share anonymized data for a new research project.
RELATED: AI’s success in healthcare relies on quality data
Accelerating the time it takes to perform an MRI may seem insignificant, but the group argues it will assist rural providers where limited access creates a scheduling backlog. It could also improve patient experiences, reducing the time patients need to hold their breath during heart and lung images and filling the role of X-ray and CT scans that include ionized radiation.
The project will begin by sifting through 3 million MRIs of the knee, brain and liver from 10,000 deidentified clinical cases. Researchers were quick to point out that “no Facebook data of any kind will be used in the project.” The social media platform has been under fire lately for its user privacy failings.
“This collaboration focuses on applying the strengths of machine learning to reconstruct the most high-value images in entirely new ways,” Recht, Sodickson and Zinick wrote. “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.”