Jo Schlemper1, Neel Dey2, Seyed Sadegh Mohseni Salehi1, Carole Lazarus1, Rafael O'Halloran1, Prantik Kundu1,3, and Michal Sofka1
1Hyperfine Research Inc., Guilford, CT, United States, 2New York University, New York, NY, United States, 3Icahn School of Medicine at Mount Sinai, New York, NY, United States
1Hyperfine Research Inc., Guilford, CT, United States, 2New York University, New York, NY, United States, 3Icahn School of Medicine at Mount Sinai, New York, NY, United States
An unsupervised deep learning framework is proposed for denoising low-field 64 mT diffusion-weighted MRI images (DWI), which enables denoising correlated noise without ground truth, leading to improved DWI quality as measured by expert evaluation
Figure 1: The summary of the proposed approach. (a) The training data is generated by creating a pair of original noisy and noisier images. The additional noise is simulated by simply feeding the raw data with higher noise to the reconstruction pipeline. (b) Once the pair is generated, the noisier images are fed to U-net like architecture to regress noisy images.
Figure 3: The reconstructions for DWI images for b=890 for a patient with pathology. The residual maps show the difference between noisy and the denoised images, highlighting the content that was removed. Both proposed and BM3D showed effectiveness, however, the proposed approach was able to preserve sharpness better.
