School of Engineering Department of Mechanical and Aerospace Engineering 185 Apply Physics Informed Neural Network (PINN) on Hemodynamics of Coronary Artery Supervisor: HU Wenqi / MAE Student: LUI Hiu Pang / AE Course: UROP 1100, Spring Coronary artery disease (CAD) is a leading global health challenge driven by atherosclerosis. In this study, a Physics-Informed Neural Networks (PINNs) is applied to predict flow fields in a simplified coronary artery model with plaque blockage. Compared to a conventional multilayer perceptron, PINN integrates governing physics laws into its training framework, so it is less dependent on an external dataset. Flow acceleration across the narrowed region, and a significant pressure drop across the bulge are observed in the resultant flow field prediction. These predictions are well aligned with the governing physics laws, indicating the potential of adopting PINNs for rapid early disease screening. Magnetic Soft Materials with Directional Stiffness Supervisor: HU Wenqi / MAE Student: ZHANG Yunming / ISD Course: UROP 2100, Fall UROP 3100, Spring This research presents a new type of soft robotic structure that uses magnets for safe and effective movement in complex environments, such as inside the human body. Traditional magnetic soft robots can lose control if the magnetic field is slightly misaligned. The new design combines soft and rigid materials, allowing it to bend in desired directions while resisting unwanted twists. This makes it more reliable for medical applications, such as guiding tools through blood vessels or performing minimally invasive surgeries. The study showcases two prototypes: a worm-like robot that navigates difficult paths and a capsule robot that mimics natural intestinal movements for drug delivery, enhancing safety and effectiveness in healthcare.
RkJQdWJsaXNoZXIy NDk5Njg=