Breakthrough MRI technology uses AI for faster scans

Introduction

Nearly 40 million MRI scans are performed in the United States each year, according to the National Library of Medicine. The time patients spend in the enclosed machine is often cited as a source of anxiety. However, a new technology developed by NYU Langone Health and Meta AI Research aims to revolutionize this experience. By employing artificial intelligence (AI), the new MRI technology has dramatically reduced the time needed to complete scans without compromising image quality.

Dr. Michael Rex, chair of radiology at NYU Langone Health, has depended on traditional MRI technology for most of his career, acknowledging its effectiveness in imaging soft tissues. However, he noted that acquiring high-quality images has always required a significant amount of time. The collaboration between NYU Langone Health and Meta began in 2018 to create "Fast MRI," a technology that leverages AI to expedite the scanning process.

Fast MRI technology allows for scanning speeds to be increased by a factor of six. This is achieved by using a neural network to reconstruct images, resulting in image quality that is comparable to, or even surpasses, that of traditional scans. Patients like Jackie G. from Long Island, who undergoes regular scans to monitor her cancer remission, have experienced the benefits firsthand. Initially skeptical, she reports that scan times have reduced from about 45 minutes to just 20-30 minutes, creating a more comfortable experience.

The faster scanning process also enables Emily Duria, assistant director of MRI services at NYU Langone, and her team to accommodate more patients. She noted that patients often leave the facility even before their appointment times, as the faster technology calms their nerves. Fast MRI has now become standard practice for most patients at NYU Langone at no additional cost, as it is available on nearly all MRI machines.

Moreover, the Fast MRI technology is capable of producing more detailed images than traditional MRI. Dr. Rex showcased a knee scan acquired with eight times acceleration, emphasizing how AI reconstruction significantly improves image clarity compared to conventional techniques.

The success of Fast MRI has sparked hope for the improved accessibility of these scans worldwide. Both NYU Langone Health and Meta have committed to an open-source approach for the technology, allowing MRI manufacturers to incorporate variations of the technique into their equipment. This open-access model signals a promising future where AI may continue transforming medical practices.

Dr. Rex expressed optimism about AI’s potential in the healthcare sector, explaining that it is not intended to replace physicians or radiologists. Instead, it aims to assist and enhance their capabilities, ultimately leading to better patient care across the board.

Keywords

• MRI scans
• AI Technology
• Fast MRI
• NYU Langone Health
• Meta AI Research
• Image quality
• Patient comfort
• Open-source technology
• Healthcare innovation

FAQ

1. What is Fast MRI?
Fast MRI is a new technology developed by NYU Langone Health and Meta AI Research that uses artificial intelligence to significantly reduce the time required for MRI scans without compromising image quality.

2. How much faster are Fast MRI scans compared to traditional ones?
Fast MRI can increase the scanning speed by a factor of six.

3. Are there any additional costs for Fast MRI at NYU Langone Health?
No, Fast MRI has become standard practice for most patients at NYU Langone Health at no additional cost.

4. How does the image quality of Fast MRI compare to traditional MRI scans?
In many cases, Fast MRI produces more detailed images than traditional MRI scans.

5. Is Fast MRI technology usable by other MRI manufacturers?
Yes, the technology has been made open-source, allowing various MRI manufacturers to incorporate it into their equipment.