Meeting Banner
Abstract #2809

REconstruction of MR images acquired in highly inhOmogeneous fields using DEep Learning (REMODEL 2.0)

Marina Manso Jimeno1,2, John Thomas Vaughan Jr.1,2, and Sairam Geethanath2
1Department of Biomedical Engineering, Columbia University, New York, NY, United States, 2Columbia Magnetic Resonance Research Center, New York, NY, United States


The trade-off of a shorter MRI magnet design is having to compromise image quality due to reduced field homogeneity. Moreover, the sample induces unique field perturbations during the scan that can only be corrected if the exact field map is known. Here we proposed a DL model that synthesizes sample-specific field maps based on artifact-corrupted acquired images and the system’s field distribution. In the retrospective study, we obtained a mean NRMSE of 0.04 during testing between model output and the true field maps and found that using the model output for correction significantly reduced the artifacts on the images.

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

Join Here