Visualizing Gastric Motility: Utilizing FLASH 2 Real-Time MRI and Artificial Intelligence for Quantification of Antral Peristalsis.

Motivation: Evaluation of gastric motility patterns and disorders is mostly dependent on external interventions (gastric manometry) or methods involving radioactive radiation (scintigraphy). Goal(s): The goal was to establish a non-invasive and fast MRI protocol to overcome these limitations. Approach: Using FLASH 2 real-time MR imaging, three sagittal slices with high frame rates of the antrum were acquired simultaneously, segmented using a trained artificial intelligence, a human-in-the-loop protocol and evaluated regarding frequency, velocity, and amplitude. Results: Integration into standard DICOM software OsiriX, allowed an easy learnable workflow and fast acquisition of the peristaltic status of the antrum, offering various insights in physiology and pathophysiology. Impact: FLASH 2 real-time MRI has been shown to be suitable for visualizing gastric motility of the stomach in fed state. Together with specifically trained artificial intelligence this could speed up the process to evaluate the peristaltic status of the stomach.