Abstract #3196

SOJU-Net—Denoising MR phase images with physics-informed deep learning using artificial Rician noise augmentation

Thomas Jochmann¹, Nora Kuechler¹, Jens Haueisen¹, and Ferdinand Schweser^2,3

¹Department of Computer Science and Automation, Technische Universität Ilmenau, Ilmenau, Germany, ²Buffalo Neuroimaging Analysis Center, Dept. of Neurology, Jacobs School of Medicine and Biomedical Sciences, University at Buffalo, The State University of New York, Buffalo, NY, United States, ³Center for Biomedical Imaging, Clinical and Translational Science Institute, University at Buffalo, The State University of New York, Buffalo, NY, United States

Phase noise follows the Rician distribution, with a non-linear dependency of the phase noise on the local magnitude signal intensity. In this work, we present SOJU-Net, a deep-learning based denoising for MR phase images. SOJU-Net reduces Rician noise while preserving boundary contrast.

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

Join Here