New system eliminates need for repeat MRI scans.

GE Health care has received FDA clearance to use deep learning to power imaging technology in MRI scans.

By combining the deep learning software with the scanning method, clinicians can diagnose faster, reduce MRI scan times and gain improved image quality, according to a GE announcement.

The software works across both 2D and 3D imaging sequences, making it unnecessary to take several 2D scans to diagnose specific cases.

The deep learning method helps generate crisper images—eliminating the need for repeat scans, which in turn provides optimal scheduling for an improved patient experience.

GE’s AIR Recon DL software is designed to work with the PROPELLER scanning technique. PROPELLER, or Periodically Rotated Overlapping ParallEL Lines with Enhanced Reconstruction, is a motion-insensitive imaging used in MRI scans.

The PROPELLER scanning method uses parallel, overlapping images to capture parts of the body that tend to move during MRI scans. Breathing lungs and beating hearts are a couple of examples. Other situations where patients tend to move include geriatric, neurodegenerative, claustrophobic and pediatric cases.

GE reported that its deep learning software can result in as much as a 50% decrease in patient exam times. The health care provider further suggested its software-powered system can better detect lesions.

Dr. Tiron Pechet, radiologist and assistant medical director at Shields Health Care Group in New England, said that the deep learning MRI offering could benefit scanning musculoskeletal issues as well as the brain.

“GE Health care has closed the gaps in our ability to provide improved image quality and patient experience to all our patients across exam types,” Dr. Pechet added.

Some 3.5 million patients globally have been scanned with GE’s deep learning powered software as of September 2022.

Image: Dennis Degnan/ Getty