Diagnostic Accuracy of 5-Minute Knee MRI Using AI Image Quality Enhancement

Despite advances in accelerating MRI scans, diagnostic knee MRI protocols typically require upwards of 30 minutes of scanner time, which fundamentally limits patient throughput. Moreover, current MRI protocols are qualitative and cannot provide precise quantitative information pertaining to the biochemical status of the tissues being interrogated. In this work, we develop a rapid 5-minute diagnostic knee MRI method that provides comparable accuracy to the longer standard-of-care techniques. In addition, our proposed technique also provides quantitative information which assists radiologists in outperforming the standard-of-care and has high correlation with surgical findings. We demonstrate these advances through one of the first prospective studies that utilizes artificial intelligence to improve MRI image quality.

Chaudhari AS, Grissom MJ, Fang Z, Sveinsson B, Lee JH, Gold GE, Hargreaves BA, Stevens KJ. Diagnostic accuracy of quantitative multi-contrast 5-minute knee MRI using prospective artificial intelligence image quality enhancement. AJR Am J Roentgenol. 2020 Aug 5. doi: 10.2214/AJR.20.24172. Online ahead of print.

Online Journal Article

Example images and quantitative T2 maps from the rapid 5-minute diagnostic knee MRI method used in this evaluation.

Akshay Chaudhari
Assistant Professor (Research) of Radiology (Integrative Biomedical Imaging Informatics at Stanford) and, by courtesy, of Biomedical Data Science
Jin Hyung Lee
Associate Professor of Neurology (Neurology Research Faculty), of Neurosurgery and of Bioengineering and, by courtesy, of Electrical Engineering
Garry Gold
Stanford Medicine Professor of Radiology and Biomedical Imaging
Brian A. Hargreaves
Professor of Radiology (Radiological Sciences Laboratory) and, by courtesy, of Electrical Engineering and of Bioengineering
Kate Stevens
Associate Professor of Radiology (Musculoskeletal Imaging)