On November 21, 2025, our institute held a thematic lecture titled Trustworthiness of AI in Medicine at the Suzhou Campus of Nanjing University. The lecture was delivered by Professor Huazhu Fu, Chief Scientist at the Institute of High Performance Computing (IHPC), A*STAR, Singapore. It was hosted by our institute's Appointed Associate Professor Li Wenbin and Appointed Assistant Professor Fang Yuqi, with participation from numerous doctoral and master's students.
Professor Fu highlighted his latest research advances in Clinically Usable and Trustworthy Medical Artificial Intelligence. He emphasized that while AI holds transformative potential in medical imaging diagnosis and clinical decision support, it still faces two core challenges in real-world clinical settings: insufficient trustworthiness and limited usability. During his presentation, he systematically explained the critical roles of uncertainty quantification and out-of-distribution detection in enhancing AI safety. He also showcased the latest results from his team's platform for fundus disease diagnosis, GlobeReady, which can be applied across different domains without requiring retraining. This platform maintains stable performance across various devices, populations, and clinical environments, providing a significant example for the practical implementation of medical AI.
During the interactive session, attendees actively engaged with questions on topics such as trustworthy AI, cross-domain robustness, clinical interpretability, and the model deployment pipeline. Professor Fu addressed each question in detail, fostering a lively and enthusiastic discussion.
The lecture concluded successfully.
