Huazhong University of Science and Technology (HUST) and the Hong Kong University of Science and Technology (HKUST) held their joint forum and symposium on trustworthy AI in medical–engineering integration on Nov 20.

Eight scholars from HKUST, the Hong Kong Polytechnic University, the University of Hong Kong, Peking University, Southwestern University of Finance and Economics, and HUST delivered in-depth academic presentations to more than 100 participants on-site, with over sixty joining online.

Shi Xuanhua, Vice Dean of HUST’s School of Computer Science and Technology, opened the event by outlining the school’s recent research achievements in AI and intelligent computing. He highlighted the value of co-hosting this forum with HKUST, emphasizing its role in fostering academic exchange, sparking interdisciplinary innovation.

Centered on the theme of “Trustworthy AI and Medical–Engineering Integration”, the forum explored new pathways for computational health and translational applications by bringing together AI, computing science, and medicine.

Four speakers led the first session. Yang Qiang, Chair Professor at HKUST, discussed the challenges of large models and the response of federated learning, examining how federated and transfer learning can enhance the practical deployment of large models.

Wang Leye, Tenured Associate Professor at Peking University’s School of Computer Science, presented his work on federated collaborative diagnosis using unaligned multimodal data across hospitals.

Associate Professor Zhang Qingpeng from the University of Hong Kong introduced mathematical and machine-learning models that analyze tumor immune microenvironment dynamics to support individualized cancer treatment.

Yang Xin, Vice Dean of the School of Computing and Artificial Intelligence at Southwestern University of Finance and Economics, offered a systematic overview of federated continual learning for large models.

The second session featured presentations by four additional scholars. He Kun, Distinguished Professor under HUST's distinguished scholar program, outlined advances in transfer-based adversarial attacks, jailbreak attacks, and hallucination mitigation.

Xiong Hui, Associate Vice President and Chair Professor of AI at HKUST (Guangzhou), examined recent breakthroughs and applications shaping the next generation of embodied VLA systems and emphasized the irreplaceable role of human wisdom in guiding emerging intelligent technologies.

Professor Wu Dongrui from HUST discussed advances in precise, secure, and privacy-preserving EEG signal decoding, highlighting the importance of adversarial safety and privacy protection for large-scale brain-computer interface deployment.

Chen Lei, Chair Professor at HKUST, presented new system-level strategies for optimized and scalable large-model inference through proactive key-value cache management.

Yi Hui, Chair of the committee at HUST's School of Computer Science and Technology, concluded the event by affirming the forum's contribution to advancing the understanding of trustworthy AI in healthcare.