Dr. Thu Nguyen, Dr. Harrison Lewis, Dr. Matthias Gobbert, Department of Mathematics And Statistics
Dr. Shiming Yang, Dr. Peter Hu, University Of Maryland Baltimore School Of Medicine.
Leveraging recent advances in deep learning methodologies holds great potential to construct accurate and interpretable models tailored specifically for trauma-related applications. With the pressing prevalence of traumatic brain injury (TBI) and hemorrhagic shock across both military and civilian healthcare sectors, there is an urgent demand for more refined predictive models. Within the burgeoning domain of big data analytics in healthcare, efforts are underway to optimize resource allocation and enhance patient outcomes through thorough data scrutiny. By proactively identifying and addressing secondary injuries and hemorrhagic events, this project seeks to elevate the standard of patient care and clinical decision-making. Our research strategy revolves around harnessing deep learning techniques for medical data representation and processing, creating transformer-based predictive models for sequential data, and pioneering transformer-based approaches for hemorrhage image segmentation. These coordinated endeavors are poised to overcome existing limitations in predictive modeling and medical image analysis, thereby driving advancements in trauma care.