Radiation is critical to modern medicine and frequently used for procedures such as diagnostic imaging, nuclear medicine, and radiotherapy. It is used extensively in procedures involving X-rays, CT scans and fluoroscopy by radiologists, interventionalists, nuclear medicine doctors and more. Long-term radiation exposure can pose adverse health risks to medical workers including increasing risk of cancer, reproductive health concerns, and genetic damage. Radiation protection is necessary for medical workers who are routinely exposed to considerable radiation levels, but current widely used lead aprons introduce significant orthopedic strain onto the body. While alternative protection devices have been produced, they have not faced widespread adoption as they are incompatible with other healthcare worker needs, including cost-effectiveness, durability levels, and easy integration into the clinical workflow. We propose a novel lightweight material capable of providing adequate radiation protection, reducing orthopedic strain, and addressing the clinical needs of healthcare professionals.

This project involves the design, simulation, and layout of a CMOS image sensor using Cadence tools. Students will build and validate a 3-transistor active pixel sensor (3T APS), column and row C2MOS shift registers, and a column read-out circuit. The final design integrates a 4×4 pixel array with scanning logic and fits within a 2.5mm × 2.5mm layout. The imager will be tested through transient simulations showing pixel readouts. Deliverables include DRC/LVS-verified schematics, layouts, presentations, and a four-page IEEE-style final report.

This project investigates the feasibility of running and fine-tuning lightweight Large Language Models (LLMs) on resource-constrained edge devices, specifically the NVIDIA Jetson platform. Our objective is to systematically evaluate multiple small-scale LLMs in terms of their inference accuracy, GPU memory usage, power consumption, and overall computational efficiency. By leveraging built-in monitoring tools and external measurement methods, we aim to provide a comprehensive comparison of model performance and resource demands. Following the evaluation phase, we fine-tune a selected model on a downstream task to assess how well Jetson handles on-device training or adaptation. This study highlights the trade-offs between model performance and hardware limitations, offering insights for real-world deployment of LLMs in low-power environments. The results will help guide model selection and optimization strategies for developers targeting edge AI applications with minimal computational overhead.

When an aneurysm ruptures, blood begins to build up in the space between the brain and the skull. After surgeons operate to stop the bleeding, patients face 2-3 weeks of recovery in the ICU. During this period, patients risk developing vasospasm—the spontaneous narrowing of cerebral blood vessels caused by inflammation and tissue damage—which can cut off blood flow to the brain (Mascia, 2009). The resulting damage can lead to long-term disability or even death (Shah, 2023). VasculaSure is a drug delivery system that continuously releases a vasodilating agent directly into the ventricles of the brain over the course of three weeks. Once released into the cerebrospinal fluid, the vasodilator circulates around the brain and exerts therapeutic effects on blood vessels in spasm, restoring healthy blood flow to brain tissue.