Abstract #1317

Rapid Dealiasing of Undersampled, Radial First-Pass Cardiac Perfusion MR Images using 3D Residual U-Net

Lexiaozi Fan^1,2, Daming Shen^1,2, Hassan Haji-Valizadeh³, Nivedita K Naresh⁴, James C. Carr¹, Benjamin H. Freed⁵, Daniel C. Lee⁵, and Daniel Kim^1,2

¹Department of Radiology, Northwestern University Feinberg School of Medicine, Chicago, IL, United States, ²Department of Biomedical Engineering, Northwestern University, Evanston, IL, United States, ³Department of Medicine (Cardiovascular Division), Beth Israel Deaconess Medical Center & Harvard Medical School, Boston, MA, United States, ⁴Department of Radiology, University of Colorado Denver, Denver, CO, United States, ⁵Division of Cardiology, Internal Medicine, Northwestern University Feinberg School of Medicine, Chicago, IL, United States

Compressed sensing (CS) is capable of accelerating cardiac perfusion MRI for achieving high spatial resolution (1.6 mm x 1.6 mm x 8 mm) and extensive spatial coverage (6+ slices per heartbeat), but the lengthy image reconstruction time (~8 min per slice with 64 frames using GPU) hinders its clinical translation. In this study, we sought to, for the first time, rapidly reconstruct accelerated cardiac perfusion data using a 3D residual U-net for clinical translation.

This abstract and the presentation materials are available to members only; a login is required.

Join Here