biostudies-literatureUnknownToner BTechnology and Research Initiative Fund (TRIF) Improving Health InitiativeNational Cancer Institute (NCI) - NIHArizona Biomedical Research CenterNIBIB NIH HHSNational Institute of Biomedical Imaging and Bioengineering (NIBIB) - NIHResearch Training Group in Data Driven Discovery at the University of Arizona - NSFNCI NIH HHSNational Institute of Biomedical Imaging and Bioengineering<h4>Purpose</h4>To accelerate respiratory triggered free-breathing T2 mapping of the abdomen while maintaining high-quality anatomical images, accurate T2 maps, and fast reconstruction times.<h4>Methods</h4>We developed a flexible deep learning framework that can be trained in a fully supervised manner to improve T2-weighted images or in a self-supervised manner to reconstruct T2 maps.<h4>Results</h4>For retrospectively undersampled data, anatomical images and T2 maps reconstructed by the proposed deep learning method demonstrated reduced voxel-wise error compared to existing traditional and compressed sensing techniques. Reconstruction times were approximately 1 s per slice, significantly faster than existing compressed sensing techniques. Prospectively undersampled data were also acquired to assess the model.<h4>Conclusion</h4>The proposed deep-learning framework reconstructed high-quality anatomical images and accurate T2 maps from datasets undersampled to only 160 total radial views (5 views per echo time), enabling full liver coverage in under three minutes on average with per-slice reconstruction times of approximately one second.Magnetic resonance in medicinehttps://www.ebi.ac.uk/biostudies/studies/S-EPMC12396147biostudies-literatureAccelerated free-breathing abdominal T2 mapping with deep learning reconstruction of radial turbo spin-echo data.PMC12396147CTR056039DMS-1937229EB031894U01 EB031894CA245920R01 CA245920Johnson KCodreanu ISridhar SMartin DRZhang SBilgin ADeshpande VToner BNadar MAltbach MIArberet SAbouelfetouh ZArif-Tiwari HAhanonu EGoerke UHan FfalseAccelerated free-breathing abdominal T2 mapping with deep learning reconstruction of radial turbo spin-echo data.<h4>Purpose</h4>To accelerate respiratory triggered free-breathing T2 mapping of the abdomen while maintaining high-quality anatomical images, accurate T2 maps, and fast reconstruction times.<h4>Methods</h4>We developed a flexible deep learning framework that can be trained in a fully supervised manner to improve T2-weighted images or in a self-supervised manner to reconstruct T2 maps.<h4>Results</h4>For retrospectively undersampled data, anatomical images and T2 maps reconstructed by the proposed deep learning method demonstrated reduced voxel-wise error compared to existing traditional and compressed sensing techniques. Reconstruction times were approximately 1 s per slice, significantly faster than existing compressed sensing techniques. Prospectively undersampled data were also acquired to assess the model.<h4>Conclusion</h4>The proposed deep-learning framework reconstructed high-quality anatomical images and accurate T2 maps from datasets undersampled to only 160 total radial views (5 views per echo time), enabling full liver coverage in under three minutes on average with per-slice reconstruction times of approximately one second.2025-01-01T00:00:00Z2025 Aug2026-05-28T04:12:37.099Z2026-04-08T02:15:00.708ZS-EPMC123961474076214910.1002/mrm.70017