Diffusion-Weighted MRI based Characterization of Bowel Inflammation in Crohn`s Disease

Non-invasive differentiation between active inflammation and fibrosis in Crohn`s disease imposes a challenging problem using commonly used MRI techniques. Quantitative Diffusion-weighted MRI has the potential to provide important new insights into micro-structural properties of the body that may differentiate between active inflammation and fibrosis. However, microstructural analysis of DW-MRI signal decay through commonly used signal decay model is limited in providing a detailed and precise characterization of the tissue microstructure. In this talk, I`ll present a novel algorithmic approach that enables a detailed and precise characterization of the tissue microstructure by utilized a motion-robust probabilistic signal decay model that explicitly characterizes the multi-scale nature of diffusion inside the tissue, combined with a powerful combinatorial solver to achieve robust estimates of the model parameters.

I’ll demonstrate the potential clinical impact of the proposed approach in both improving the quality of DW-MRI images for visual assessment of Crohn`s disease and in the non-invasive characterization of active inflammation and fibrosis in Crohn’s disease patients.

This work was done at Boston Children`s Hospital/Harvard Medical School.