New study on deep learning technology to improve MRI image quality published in European Radiology

SwiftMR™, an AI-assisted MRI reconstruction solution from AIRS Medical, has demonstrated its power to improve the image quality of high-resolution 3D MRIs

Seoul, South Korea, November 25, 2022 /PRNewswire/ — A study recently published in European Radiology demonstrated that SwiftMR, an AI-assisted MRI reconstruction solution from AIRS Medical, successfully denoises 3D MR images and improves their image quality using only routine clinical scans.

The study aimed to develop deep neural network (DNN)-based noise reduction and image quality enhancement using only routine clinical scans and evaluating their performance in high-resolution 3D MRI. The study was conducted by AIRS Medical and Dr. Jinhee Jang, MD, Ph.D. of the Seoul St. Mary’s Hospital. The retrospective study included T1-weighted magnetization-prepared MP-RAGE (rapid gradient echo) images from 185 clinical scans.

The qualitative evaluation between conventional MP-RAGE and DNN-based MP-RAGE was performed by two radiologists in terms of image quality, fine structure delineation, and lesion conspicuity. Using fully sampled data as a reference, the quantitative assessment was performed by measuring quantitative error metrics and volumetrics at seven different simulated noise levels. The DNN application at the VWI was evaluated by two radiologists in terms of image quality.

The study shows that DNN-based MP-RAGE outperformed conventional MP-RAGE in all image quality parameters (mean score = 3.7 vs. 4.9, p<0.001). In the quantitative evaluation, DNN showed better error metrics (p < 0,001) und vergleichbare (p > 0.09) or better (p<0.02) volumetric results than conventional MP-RAGE. DNN application in VWI also showed improved image quality (3.5 vs. 4.6, p<0.001).

“We are very pleased to have proven that SwiftMR™ not only helps reduce time, but also helps produce good to great images, ultimately helping radiologists read images with confidence.” noticed Hyeseong Lee, MD, CEO of AIRS Medical. “We anticipate further customizations and collaborations with radiologists around the world so we can grow together.” he added.

AIRS Medical recently announced its participation in the 108th RSNA Scientific Assembly and Annual Meeting 2022, hosted by November 27th to 30th in Chicago. During the event, AIRS Medical will showcase its award-winning SwiftMR™ MRI reconstruction solution and also make two oral presentations at a scientific session.

“We are proud of ourselves as it is quite unusual for a start-up to make oral presentations at an RSNA scientific meeting. To be accepted as an oral presentation is a recognition of academically important achievement,” he added.

About AIRS Medical

Founded in October 2018, AIRS Medical Inc. is a medical AI solutions provider headquartered in Seoul, Korea, developing innovative products and services to improve patient experience. AIRS Medical helps healthcare systems achieve greater institutional efficiencies in areas where lack of productivity limits clinical value.

For more information visit

Eutak Jin
[email protected]



Leave a Reply

Your email address will not be published. Required fields are marked *