3433

View Abstract / View Presentation

Physically Motivated Deep-Neural Networks of the Intravoxel Incoherent Motion Signal Decay Model for Quantitative Diffusion-Weighted MRI

Shira Nemirovsky-Rotman¹, Elad Rotman¹, Onur Afacan², Sila Kurugol², Simon Warfield², and Moti Freiman¹
¹Biomedical Engineering, Technion, Haifa, Israel, ²Boston's Children's Hospital, Harvard Medical School, Boston, MA, United States

A deep learning model is introduced for Quantitative Diffusion-Weighted MRI with the Intra-Voxel Incoherent Motion model, which incorporates the acquisition protocol into the network input, thus providing improved robustness of parameters estimates to acquisition protocol variations.

Proposed method estimated parameters maps compared to Barbieri's, with the ground truth maps on the bottom row.

Proposed method architecture, where the acquisition parameters are added to the network's input.