AI in Medicine - From Imaging Signatures & Physics-Based Simulation of Drug Response to Knowledge Distillation in Medical Imaging

Abstract:

Artificial Intelligence is rapidly transforming medicine, from discovering new therapies to predicting treatment outcomes. This talk presents an overview of recent advances in machine learning–driven drug response prediction, focusing on immunotherapy benefit estimation from CT scans and the use of deep feature learning to capture tumor biology beyond tissue-based biomarkers. I will discuss our two recent works, one on DeepCT and another on physics-based simulationto predict benefits from immunotherapy and targeted therapy. I will also present how semi-supervised knowledge distillation can leverage unlabeled data for robust medical image segmentation. The discussion will conclude with reflections on Responsible AI practices—highlighting model reliability, interpretability, and fairness as prerequisites for clinical deployment.

About Rizwan:

Dr. Rizwan Qureshi is a Senior Member of IEEE and a Research Scientist specializing in Responsible AI, Vision–Language Models, and Computational Health. His research focuses on reliability, interpretability, and safety of foundation models. He has published in leading venues such as IEEE Transactions on Medical Imaging, CVPR, and ACM Computing Surveys, with over 3,700 citations and an h-index of 24. Dr. Qureshi has worked with Dr. Mubarak Shah at the Center for Research in Computer Vision (CRCV), University of Central Florida, and collaborates globally on AI safety, health informatics, and model evaluation. His current work bridges cognitive science and machine learning to build AI systems that are not only powerful—but responsible.