The project assesses patient cardiovascular risk and matches patients to the treatments most likely to be effective. The project addresses this problem through sophisticated computational methods that identify new markers of disease, improve the ability to measure both new and existing markers, and construct personalized models that can provide highly accurate assessments of individual risk. The core focus of the research addresses the poor performance of existing tools for cardiovascular decision support through advanced methods at the intersection of machine learning, data mining, signal processing, and applied algorithms; with the research guided by knowledge of cardiac pathophysiology.
This project impacts patient care for a disease that causes roughly one death every 38 seconds in the United States and imposes a burden of over half a trillion dollars in the U. S. each year. More generally, many of the ideas explored here (e.g., personalization of risk models) extends to a wide variety of other disorders in a straightforward manner and leads to wide improvements in outcomes while controlling costs. The research also strengthens interdisciplinary research in EECs and medicine throughout the computer science research community.
