Dr. Peyster serves as the Assistant Professor of Advanced Heart Failure and Transplant Medicine at the University of Pennsylvania. His research focuses on applying sophisticated computational methods to multidimensional diagnostic data in order to enhance the understanding of cardiovascular diseases. With clinical training in advanced heart failure/transplant medicine and scientific training in translational research, his work focuses on designing rigorous experiments that facilitate the translation of innovative technologies into patient care. He has led multiple first-in-field investigations leveraging ‘machine learning’ computer-vision technology to diagnose and predict diseases of both the native and transplanted heart. As a result of these efforts, he has received multiple national research awards, been awarded several US patents, presented at numerous national conferences, and serves as key personnel on multiple NIH-funded projects. Current areas of active research focus on multi-modal data integration, combining digital pathology analysis with other data types to develop novel diagnostic and prognostic platforms. These include utilizing ‘histo-immunologic’ data to predict transplanted heart rejection, ‘histo-clinical’ data to predict long-term transplanted heart failure, and ‘histo-proteomic’ data to improve donor organ harvest and allocation decisions. Ultimately, the goal of his research is to identify and integrate promising analytical methods to improve disease understanding and enhance precision medical practices.
In this talk, we will review the computational methods that are changing how researchers look at data, and which are poised to change how clinicians treat their patients. We will demystify tech ‘buzz words’ like artificial intelligence, machine learning, big data, and computer vision, and discuss why these tools are well-suited for addressing longstanding issues in transplant medicine. Recent translational applications using AI to perform automated analysis of pathology samples will be presented, highlighting the potential of this approach to improve the diagnosis and prediction of allograft rejection and vasculopathy. Lastly, future directions for AI-enabled precision medicine in transplantation will be introduced, followed by audience questions. and answers.