Meeting Banner
Abstract #2332

Deep Learning Denoising to Accelerate Diffusion-Weighted Imaging of Rectal Cancer

Mohaddese Mohammadi1, Elena Kayee1, Youngwook Kee1, Jennifer Golia Pernicka 2, Iva Petkovska2, and Ricardo Otazo 2
1Medical physics, Memorial Sloan Kettering Cancer Center, New York, NY, United States, 2Memorial Sloan Kettering Cancer Center, New York, NY, United States

Deep learning denoising using a convolutional neural network (DnCNN) is proposed to accelerate the acquisition of diffusion-weighted imaging (DWI) data in patients with rectal tumors. 4-fold acceleration was achieved by reducing the number of averages from 16 to 4 and applying the DnCNN to denoise the data. The DnCNN was trained using pairs of noisy (1 average) and reference (16 averages) images from 92 patients. The trained network was then tested with the data from 6 patients and the results were evaluated qualitatively by radiologists. DnCNN represents a powerful approach to improve efficiency of DWI in patients with rectal cancer.

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

Join Here