PHIND Seminar: Opportunistic Disease Prediction using Already-Acquired Medical Imaging and Deep Learning - Akshay Chaudhari, PhD
Li Ka Shing Building - Room LK 120
291 Campus Drive
Palo Alto, CA 94305
Over 88 million computed tomography (CT) scans are performed annually in the US, with abdominal CT accounting for ~20 million. While these scans answer specific clinical questions, a majority of the information in the rich 3D scans unrelated to the referral question is not evaluated. In this work, we will demonstrate how extracting additional biological insights beyond those required to answer the original clinical question can be used for predicting the onset of future disorders. We will demonstrate how we can analyze CT images using deep learning tools to opportunistically predict future cardiometabolic disorders with high accuracy. We will depict how we can combine medical imaging data with data from electronic medical records to improve the accuracy of such models. Overall, opportunistic imaging has the potential to be a paradigm-changing new tool to improve health outcomes through early detection and intervention without requiring additional diagnostic testing since it uses CT imaging that has already been acquired.
Dr. Chaudhari is an Assistant Professor in the Department of Radiology and (by courtesy) in the Department of Biomedical Data Science. He leads the Machine Intelligence in Medical Imaging research group at Stanford and has a primary research interest that lies at the intersection of artificial intelligence and medical imaging. Dr. Chaudhari is interested in the application of artificial intelligence techniques to all aspects of medical imaging, including automated schedule and reading prioritization, image reconstruction, quantitative analysis, and prediction of patient outcomes. His interests range from developing novel data-efficient machine learning algorithms to clinical deployment and validation of patient outcomes, both for medical imaging acquisition and subsequent analysis. He also conducts research in combining imaging with clinical, natural language, and time series data.