Back to all papers

AI- assisted compressed sensing for accelerated 1.5T shoulder MRI.

November 14, 2026pubmed logopapers

Authors

Hu W,Wang Y,Dong H,Wang J,Wang L,Chen Z,Zheng Z

Affiliations (2)

  • School of Clinical Medicine, Tsinghua University, Beijing, China.
  • Department of Radiology, Beijing Tsinghua Changgung Hospital, School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Beijing, China.

Abstract

BackgroundShoulder MRI has relatively long acquisition times. The clinical performance of AI-assisted compressed sensing (ACS) acceleration levels in 1.5-T shoulder MRI remains unclear.ObjectiveTo compare the acquisition time, image quality, and diagnostic agreement of three clinical shoulder MRI protocols using parallel imaging (PI), ACS4, and ACS8 at 1.5 T.Materials and MethodsProspectively, 158 patients with shoulder discomfort underwent three scanning protocols-the conventional Parallel imaging (PI) protocol, the ACS with a nominal acceleration factor of 4 (ACS4), and the ACS with a nominal acceleration factor of 8 (ACS8), covering four common sequences. Acquisition time, quantitative image quality parameters including True SNR, Relative SNR, Relative CNR, and edge rising distance (ERD), subjective image quality scores from three radiologists, and diagnostic agreement in an exploratory subgroup of 25 patients were assessed.ResultsACS4 and ACS8 reduce the total acquisition time by approximately 29% and 45%, respectively, compared with the PI protocol while maintaining comparable image quality across evaluated sequences. ACS8 generally demonstrated higher quantitative Relative SNR/CNR values in several tissues and sequences, whereas ACS4 achieved the highest True SNR and subjective image quality scores. ERD differences were not statistically significant, and diagnostic agreement was generally high across protocols.ConclusionACS protocols reduced acquisition time in 1.5-T shoulder MRI, although ACS8 compromised subjective image quality. ACS4 offered a more favorable balance between scanning efficiency and image quality.

Topics

Journal Article

Ready to Sharpen Your Edge?

Subscribe to join 11k+ peers who rely on RadAISlice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.