Project Summary: Crohn's disease (CD) is a chronic inflammatory disease of the gastrointestinal tract that has a prevalence of over 800,000 in the US alone. The economic impact is disproportionally high because it affects primarily young individuals. The characteristic periods of remission and relapse necessitate frequent hospitalizations. There has been a recent shift in clinical practice from a reactive to a proactive CD treatment strategy, recently coined as “treat-to-target”, where patients are treated to achieve not only an initial response-to-therapy, but also longer clinical remissions and improved mucosal healing. However, this approach requires the regular assessment of disease activity using objective markers enabling treatments to be tailored to the individual patient. Therefore, there is an unmet need for new tools to improve diagnostic accuracy, to quantify disease burden, and to monitor treatment efficacy. Our application in response to PAR-19-056, “Robust quantitative MR imaging markers of response to therapy in Crohn's disease” is aimed at developing and evaluating noninvasive, contrast and radiation-free quantitative imaging markers for assessing disease activity and for monitoring response to therapy. Currently available non-contrast MR imaging (MRI) sequences such as diffusion-weighted MRI (DW-MRI) and the calculated apparent diffusion coefficient (ADC) maps are attractive but limited in providing robust and reproducible markers. Different imaging protocols or different scanners result in different ADC values. Our primary goal is to develop robust and reliable quantitative DW-MR imaging markers for the non-contrast and non-invasive assessment of CD. The proposed novel MRI acquisition and model fitting techniques will provide measurements of slow and fast diffusion as well as fraction of fast diffusion, as highly accurate, quantitative biomarkers for cell proliferation, density and size, and tissue perfusion—all indices that characterize the extent of disease activity (i.e., inflammation) in the tissue micro-structure of the bowel. To achieve this goal, we will first develop and implement an advanced distortion and motion corrected (DiMoCo) DW-MRI technique and a spatially constrained probabilistic intravoxel incoherent motion (SPIM) model to compute robust and reproducible markers. Next, we will reduce the imaging time with estimated x4 acceleration with an accelerated image acquisition and a new, advanced deep learning-based parameter estimation technique. The proposed imaging techniques and software tools will provide robust quantitative markers, thereby enabling the accurate assessment of CD activity and response to therapy. They will also reduce the need for invasive endoscopy procedures as well as the total number of other tests typically ordered for monitoring disease, effectively reducing both the disease burden and the overall cost of healthcare. Another important goal is to develop and broadly disseminate open source software that will enable the standardized evaluation of other diseases presently evaluated with DW-MRI that would benefit from the advanced diagnostic and assessment capabilities of the proposed DiMoCo SPIM-DW-MRI technique.

Project Narrative: Crohn's disease (CD) is a chronic inflammatory disease of the gastrointestinal tract that impacts over 800,000 patients in the US alone, where characteristic periods of remission and relapse necessitate frequent, costly hospitalizations and symptoms such as bloody diarrhea, abdominal pain, general malaise, and fever significantly compromise quality of life. There is a need for new tools aimed at the reliable assessment of disease activity by using objective markers to improve diagnostic accuracy, quantify disease burden, and monitor treatment efficacy, enabling subsequent treatments tailored to the individual patient's disease profile. The primary objective of this project is three-fold: First, to develop and validate a novel, noninvasive, contrast- and radiation-free quantitative MR imaging and image analysis technique, i.e., DiMoCo-SPIM that will generate robust and reproducible imaging markers from DW-MRI that will enable clinicians to more accurately assess bowel inflammation and response to therapy; second, to develop a machine learning technique that reduces the imaging time (x4 acceleration); and third, to make image analysis software developed in this proposal available in multiple clinical domains that presently evaluate various diseases with MRI and would benefit from the improved diagnostic and assessment capabilities of this computationally-driven technique.