Proposal

The objective of the proposed study is the development of an in-silico, virtual human stomach model called “eStomach” which simulates all the key biomechanics/biophysics in gastric digestion: electromechanics for gastric motility, fluid dynamics with flow-structure interaction, and flow mediated biochemical reactions for digestive process. To incorporate patient variability in morphology and key parameters, we will employ a data-assimilation approach by using patient-specific MRI imaging data. eStomach will not only allow us to investigate the details of gastric biomechanics but also enable “virtual” gastric surgeries and prediction of expected outcomes. This could revolutionize bariatric surgery by providing more predictable and potentially better outcomes for patients. 

Previous in-silico models of stomach biomechanics including those developed by our group20-23 are missing the ability to directly take in imaging data from patients and have also not been verified nor validated against in-vivo data. This was mainly because the in-vivo measurement of flow variables in the human stomach is extremely difficult. In the research effort proposed for this seed grant, we will improve our existing stomach biofluid dynamics model significantly by directly incorporating the morphology from the patient imaging data and performing verification and validation against the in-vivo flow data. For this purpose, we will develop a physics informed neural network (PINN) model to extract the flow data from the time-dependent contrast enhanced MRI data. This eStomach model couples four main physical domains: electrophysiology, mechanics, fluid dynamics, and biochemistry and will also incorporate patient-specific data. The model, therefore, will provide an unprecedented capability for modeling the function of the stomach as well as virtual surgery planning.