BME PhD Preliminary Exam Announcement for Conner C. Earl (C. Goergen, advisor) Everyone is invited to the public presentation beginning at 3:00pm. Title: Simplified, Improved, and Automated 4D Cardiac Image Analysis Date: 9/19/22 Time: 3pm-5pm (3pm-4pm open portion, 4pm-5pm closed portion) Location: MJIS 2001 or Zoom: https://purdue-edu.zoom.us/j/91304131947 Advisory committee: Craig J. Goergen, Chair; Guang Lin; Vitaliy L. Rayz; Larry W. Markham; Jonathan H. Soslow Abstract: Cardiovascular disease is the leading cause of death worldwide. Cardiac imaging is an increasingly utilized tool to help with early identification and treatment of disease, however as imaging techniques advance in complexity and application, there is a greater need for accessibility and feasibility for clinical and research use. For example, 3D and 4D imaging methods arguably provide a greater morphological and quantitative assessment of cardiac fitness, however these methods also introduce complexity to analysis and interpretation of results. Deep learning for biomedical image segmentation and classification is a promising method that could potentially be used to automate analysis and reduce complexity but requires an informed approach to do so. This proposal builds on previously developed strengths in our lab in using 4D ultrasound to evaluate mouse models of cardiac disease and 4D cardiac magnetic resonance imaging in the characterization of Duchenne muscular dystrophy-associated cardiomyopathy. Using these frameworks, we plan to simplify, improve, and automate 4D cardiac image analysis in both experimental and clinical cardiac image datasets. We hypothesize that these improvements will decrease analysis time of 4D image datasets by more than 90% and that automated image analysis will achieve >95% accuracy for both image segmentation and classification. Achieving these goals will address a growing need in cardiac imaging to provide accessible techniques and technologies to researchers and healthcare providers with the goal of improving cardiac outcomes for patients worldwide.