Title: Beyond the Chatbot: Transforming Cancer Diagnosis and Treatment Through Generative AI and Precision Transformers
Speaker: Dongxiao Zhu, PhD
Professor, Department of Computer Science
Director, Institute for AI and Data Science
Director, Trustworthy AI Lab
Wayne State University
Time: Thursday, April 30, 12:00 PM – 1:00 PM (U.S. Eastern Time)
Registration link: (free and open to the research community) https://events.teams.microsoft.com/event/73b9111b-b231-44e6-901a-8a1f6848a66d@e51cdec9-811d-471d-bbe6-dd3d8d54c28b
Abstract: Clinicians are increasingly challenged by the scale and heterogeneity of clinical data-from sparsely labeled imaging to fragmented patient records. This talk presents a unified framework that integrates Generative AI and Vision Transformers to bridge the gap from data understanding to high-precision clinical action. We begin by examining Large Language Models (LLMs) as clinical assistants for distilling complex patient histories, supporting automated consultations, and enabling personalized, interpretable risk assessment. To address critical concerns around HIPAA compliance and black-box hallucinations, we introduce white-box AI approaches that provide secure, localized, and interpretable predictions through attention-based feature attribution.
The second part focuses on precision imaging, where transformer-based architectures and foundation models address the label scarcity bottleneck in cancer segmentation. We highlight three key innovations: (1) FocalUNETR, which captures local–global interactions for boundary-aware segmentation in low-contrast CT; (2) AutoProSAM, which enables automated 3D multi-organ segmentation without manual prompting; (3) MulModSeg, which improves unpaired multi-modal (CT/MR) segmentation via modality-conditioned text embeddings. Finally, we present FluenceFormer, a physics-informed framework for automated radiotherapy planning that predicts clinically deliverable fluence maps. This end-to-end "assessment-to-action" pipeline illustrates how AI can evolve from a diagnostic support tool to a clinically actionable system for precise and reliable cancer treatment.
Speaker Bio: Dr. Dongxiao Zhu is a Professor in the Department of Computer Science, Director of the Institute for AI and Data Science and Director of the Trustworthy AI Lab at Wayne State University. He received his Ph.D. in Bioinformatics from the University of Michigan, Ann Arbor. Dr. Zhu's research focuses on trustworthy artificial intelligence for high-stakes domains, with a growing emphasis on biomedical and cancer research applications. His work develops robust, interpretable, and fair machine learning methods that enable reliable decision-making from complex, multimodal data such as electronic health records, medical imaging, and population-level data. In oncology, his research is particularly relevant to risk prediction, treatment planning, and reducing disparities in cancer outcomes, where trust, reliability, and transparency are critical for clinical adoption. He has published over 100 peer-reviewed papers and served on senior program committees of leading AI and biomedical informatics venues, including NeurIPS, ICML, ICLR, AAAI, AMIA, and MICCAI, as well as journals such as Bioinformatics, Medical Physics, and Journal of Medical Internet Research. He maintains a strong and sustained external funding portfolio, totaling approximately $12 million, with $8 million as principal investigator. His interdisciplinary efforts bridge computer science, medicine, and public health, aligning with the broader goal of leveraging AI to improve cancer care and outcomes.
Host and Contact: Yang Shi, Karmanos Cancer Institute (yangsh(AT)karmanos.org)
------------------------------
Yang Shi
Wayne State University
------------------------------