Prof. Donghyun Kim(SPIE Fellow)

Yonsei University

Donghyun Kim received B.S. and M.S. from Seoul National University in Electronics Engineering. He graduated from the Massachusetts Institute of Technology with Ph.D. in electrical engineering in the area of novel multi-dimensional display technologies and smart optical filters. He worked on next generation fiber-optic access communication systems at Corning Inc. as a senior research scientist and then investigated cellular biophotonic sensors for cell-based assays at Cornell University as a postdoctoral fellow. He has been leading Biophotonics Engineering Laboratory of Yonsei University, Seoul, Korea. The main theme of his research at Yonsei has been focused on sensing and imaging applications in biomedical engineering based on near-field manipulation by plasmonic techniques and metastructures. He has given 100+ invited lectures on related topics and written more than 200+ peer-reviewed journal and conference publications on nano and biophotonics, many of which were the results of domestic and international collaboration. He also holds 30+ international patents. Dr. Kim served as Undergraduate Chair of the School of Electrical Engineering of Yonsei University and chaired international conferences including CLEO Pacific Rim 2024. He is currently Vice President of the Optical Society of Korea. He is a Fellow of the SPIE and a Senior Member of the IEEE.

Prof. Haiteng Jiang

Zhejiang University

Jiang Haiteng is a National High-Level Young Talent, Chief Young Scientist of the China Brain Project, and Director of the Clinical Base Office of the National Key Laboratory of Brain-Machine Intelligence. He is a PI at MOE Frontier Science Center for Brain Science and Brain-Machine Integration, the Affiliated Mental Health Center of Zhejiang University School of Medicine and Liangzhu Laboratory. His team has been dedicated to developing novel brain information analysis methods, exploring cognitive and brain disease mechanisms, and facilitating clinical translation. He currently leads three national-level projects and his representative research findings have been published in high-impact journals such as Advanced Science, Annals of Neurology, and Neuropsychopharmacology as a corresponding or first author. He has received several awards, including the LiZao Neuroscience Award from Zhejiang University School of Medicine, the Emerging Countries Travel Scholar Award at the 21st International Conference on Biomagnetism, and the 2018 Neuroscience Society Hot Topic Award.

Speech Title: "Translational Psychiatry with Real-World Clinical EEG"

Abstract: Real-world data (RWD) research involves the analysis of practical, heterogeneous data collected during routine clinical practice. Unlike controlled trial data, RWD is characterized by diversity, non-structural formats, complexity, and imperfections, posing challenges but also offering valuable insights into real clinical scenarios. Electroencephalography (EEG), a non-invasive technique for monitoring brain electrical activity, has become a critical tool in diagnosing and researching neurological and psychiatric disorders such as epilepsy, depression, and sleep disorders. This report highlights our research group’s progress in leveraging deep learning (DL) and transfer learning (TL) techniques to automate sleep staging and enhance the precision of psychiatric disorder diagnoses using real-world EEG and polysomnography (PSG) data. Sleep staging, traditionally a labor-intensive manual process, benefits from DL’s ability to extract complex patterns from raw EEG signals, improving accuracy and efficiency. Meanwhile, TL enables the adaptation of pre-trained models to diverse patient populations, addressing data scarcity and variability in clinical settings. By integrating real-world EEG and PSG data, our work aims to bridge the gap between research and clinical application, ultimately advancing personalized medicine and improving diagnostic outcomes for sleep and psychiatric disorders.

Prof. Dakun Lai

University of Electronic Science and Technology of China (UESTC)

Professor Lai is currently the director of the Biomedical Imaging and Electrophysiology Lab at the University of Electronic Science and Technology of China (UESTC), and the Vice Chairman of Chinese Heart Rhythm Society (CHRS). He received his Ph.D. in Medical Electronics from Fudan University in 2008. Then he completed a three-year Postdoctoral Associate in Biomedical Engineering at the University of Minnesota, USA. From 2012, he has been on the faculty of the School of Electronic Science and Technology, UESTC, China, where he was appointed as a Professor of Electrical Science and Technology. Dr. Lai is members of IEEE and the Engineering in Medicine and Biology Society, and the member of American Heart Associate. He is also an editor and reviewer of several international journals. He has published 80 peer-reviewed papers in Circulation, Physics in Medicine and Biology, IEEE TITB, IEEE Sensors J etc. and holds 30 Chinese Patents. His research interests and main contributions include medical electronics, bioelectromagnetism and cardiac/nural electrophysiology. He has pioneered the development of various wearable ECG devices with flexible dry electrodes and made significant contributions to deep learning based bioelectrical signal analysis, and detection and prediction of severe cardiac arrhythmias and neuro-disorder.

Invited Speaker

 

Prof. Gang Wang
Xi'an Jiaotong University

Gang Wang is a Professor at the School of Life Science and Technology, Xi'an Jiaotong University. His current research focuses on intelligent monitoring of dynamic neurophysiological information and neural rehabilitation engineering. He has successively led four projects funded by the National Natural Science Foundation of China. He has published 50 SCI-indexed papers in renowned journals such as NEUROIMAGE, IEEE JBHI, IEEE TNSRE, IEEE TBME, and IJNS. Additionally, he has applied for and obtained 21 national invention patents and holds 9 software copyrights. He serves as a Committee Member of the Medical Neural Engineering Branch of the Chinese Society of Biomedical Engineering and an Associate Editor for Medical & Biological Engineering & Computing. He has received the First Prize of Shaanxi Higher Education Institutions' Scientific and Technological Award.

Speech Title: "Assessing Propofol Anesthesia Susceptibility Based on Source-Space EEG Connectivity in Low Alpha Band"

Abstract:  Background: The susceptibility of different individuals to the same dosage of the same anesthetic drug is influenced by many factors. In addition to basic conditions such as age and gender, susceptibility to anesthesia is also related to brain activity during the resting state. However, the specific physiological mechanisms involved are still poorly understood. Methods: Twenty healthy volunteers who participated in propofol-induced sedation were divided into two groups according to their susceptibility to anesthesia and the electroencephalogram were recorded in baseline and moderate sedation states. Functional connectivity in the baseline was measured by the debiased weighted phase lag index between different brain regions to find band-specific differences in source space and sensor space respectively. Classifiers for anesthesia susceptibility were trained according to connectivity and based on bi-encoder autoencoder and convolutional neural network. Results:  In the baseline state, the specific frequency band was mainly in the low alpha band, and showed that the subjects more sensitive to propofol had more weaker brain activities. Source-space functional connectivity in the specific band during the resting state could successfully assess the individual's susceptibility to propofol with an accuracy of 82.52%. Conclusions: The source-space functional connectivity in the specific frequency band during the resting state serves as a reliable biomarker that can effectively assess the susceptibility to propofol during anesthesia. This study offers novel insights to help anesthesiologists enable precision anesthesia.

 

Assoc. Prof. Jianjun Meng
Shanghai Jiao Tong University

Jianjun Meng is an associate professor at Shanghai Jiao Tong University (SJTU). He studied Mechanical Engineering and got a Bachelor's and Ph.D. degrees at SJTU. He conducted postdoctoral research at both the University of Minnesota and Carnegie Mellon University in the USA from 2014 to 2019. His research interests focus on noninvasive BCIs, neural prosthetics, and neural engineering. He has received several awards, including the First Prize for China’s Natural Science of the Ministry of Education, and has been selected in the Shanghai Pujiang Talent Program Class A. He is the PI for the general program of the National Natural Science Foundation of China and co-PI for the National Key Research and Development Program of China. He is the author or co-author of over 50 scientific SCI-indexed journal publications, including Science Robotics, National Science Review, and IEEE Trans. On BME, etc. He co-authored an academic book chapter in “Neural Engineering”. He is a senior member of IEEE (since 2022), an associate editor for IEEE Reviews in Biomedical Engineering (since 2024), IEEE Journal of Biomedical and Health Informatics (since 2024), and Frontiers in Human Neuroscience (since 2022).

 

Speech Title: "Motor Imagery-Based Brain-Computer Interface: Its Decoding and Control"

Abstract:  Conventional motor imagery paradigms and decoding methods often face challenges due to the relatively broad definition of motor imagery tasks, leading to difficulties in performing task for some subjects and poor signal decoding accuracy. Our research group have proposed a rhythmic motor imagery-based task that can elicit steady-state movement-related rhythms. During upper limb motor imagery tasks induced by rhythmic movement execution and observation, this approach generates more stable coherence features and achieves high classification accuracy in multi-class upper limb motor imagery tasks. Furthermore, in lower limb motor execution detection, the analysis of MEG signals and time-frequency features enables accurate discrimination between left- and right-foot movements, providing an effective brain-computer interface technology for lower limb rehabilitation.

 

Dr. Fan Lin
Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology

Lin Fan has been working at Tongji Hospital, Tongji Medical College, Huazhong University of Science and Technology, since his graduation in 2011. He is a member of the Chinese Non-invasive Electrophysiology Professional Committee and the Electrocardiography Branch of the Wuhan Medical Doctor Association. Lin specializes in the diagnosis of various arrhythmias, pacemaker programming and analysis, as well as esophageal and intracardiac electrophysiological examinations. He was awarded the 3rd "China Non-invasive Cardiac Electrophysiology" Promotion and Application Award. Currently, his research focuses on artificial intelligence in cardiovascular physiological signals. He leads one National Natural Science Foundation of China (NSFC) General Program and one Tongji Hospital Youth AI Project. His key findings have been published in renowned journals, including Med (2 papers), European Heart Journal-Digital Health, IEEE Journal of Biomedical and Health Informatics (JBHI), and the Chinese Journal of Cardiac Arrhythmias. Additionally, he holds multiple authorized invention patents.

Speech Title: "AI in Atrial Fibrillation Research: Clinician Perspectives on Challenges and Opportunities"

Abstract: This clinician-focused analysis explores AI's integration in atrial fibrillation (AF) research, revealing significant opportunities alongside challenges. AI demonstrates high specificity in detecting paroxysmal AF episodes (≥6 minutes) and distinguishing AF from multiple other arrhythmias using heart rhythm data. Critically, clinician-AI collaboration achieves 100% diagnostic accuracy while reducing physician workload by 76.7%, optimizing clinical efficiency. Furthermore, the HBBI-AI model predicts AF risk from sinus rhythm intervals alone, outperforming traditional models (e.g., C2HEST), with explainability analysis linking risk to autonomic imbalance—suggesting pathways for personalized prevention. While these advances signal significant opportunities in real-time monitoring and precision risk assessment, clinicians emphasize persistent challenges, including model interpretability, integration into existing workflows, and validation across diverse populations. This clinician-centered perspective underscores AI’s dual role as both a transformative tool and an evolving solution requiring collaborative refinement to realize its full potential in AF research and care.

Assoc. Prof. Liu Yan
Sichuan University

Liu Yan, Associate Professor. She has been engaged in medical image analysis, pattern recognition, and machine learning related work. Her specific research directions includes image segmentation, image registration, deep learning, etc. At present, She have published multiple high-quality papers in authoritative academic journals and conferences, and have been invited multiple times to give presentations at mainstream conferences. She has been the PI of multiple national and provincial-level projects, including the National Natural Science Foundation of China and key research and development projects in Sichuan Province.

Speech Title: "Medical Image Segmentation Method Driven by Geometric Shape Features"

Abstract: This report focuses on geometric shape feature-guided medical image segmentation. It explores how to leverage anatomical characteristics to enhance segmentation accuracy. Relevant methods integrating shape constraints into segmentation models are elaborated, with experiments verifying their effectiveness in improving boundary precision and reducing false segmentation in medical images.

Dr. Rajeev K. Singla

West China Hospital, Sichuan University, China

 Dr. Rajeev Kumar (also known as Rajeev K. Singla) is an assistant researcher at the Institutes for Systems Genetics, West China Hospital, Sichuan University. He has extensive experience in the fields of natural products, metabolic disorders, neurological diseases, infectious diseases, chemoinformatics, and public health. He is the founder and editor-in-chief of Indo Global Journal of Pharmaceutical Sciences, associate editor of Journal of Translational Medicine,  Journal of Alzheimer’s Disease, Frontiers in Pharmacology, and Frontiers in Nutrition, a member of the Topical Advisory Panel of Cancers (MDPI), a member of the Advisory Board of Heliyon (Cell Press), and a member of the Editorial Board of BMC Complementary Medicine and Therapies and CNS & Neurological Disorders - Drug Targets. He is among the top 2% of scientists according to Stanford University, Scopus, and Elsevier reports (based on publications in 2022, 2023, and 2024). He has published more than 140 papers. According to Scopus, he has published papers with 670 co-authors from 40 countries so far. According to Web of Science, he has 447 peer review records and 298 verified editing records. To date, he has published 4 books and 19 chapters. He serves as an expert reviewer for the Sichuan University Innovation Training Program Academy and participates in guiding the College Students Innovation and Entrepreneurship Competition. In addition, he was awarded the 2024 Sichuan Province Foreign High-level Talent Certificate this year and won the second prize for the 2022-2023 Suzhou Natural Science Excellent Academic Paper Award (SZLW2024084). In 2024, he successfully applied for a Sichuan Provincial Science and Technology Department project to find alternative therapies extracted from coconut shells to fight Pseudomonas aeruginosa (Project No.: 2025YFHZ0213).

Assoc. Prof. Zhiwei Wang

Huazhong University of Science and Technology

 Dr. Zhiwei Wang is an Associate Professor and Ph.D. advisor at Huazhong University of Science and Technology (HUST). He is a recipient of the Wuhan Youth Talent Program. His research focuses on artificial intelligence and medical image processing, with particular interests in minimally invasive surgical robotics and endoscopic diagnosis and treatment systems. He has published over 45 papers in top journals and conferences such as MedIA, TMI, JBHI, ICCV, MICCAI, AAAI, IJCAI, and ACMMM, and holds more than 20 invention patents. He has led multiple national and provincial research projects. His AR navigation system for glioma surgery was selected as a major AI application scenario in Hubei Province, significantly improving intraoperative visualization and precision. His deep learning-based CTA diagnostic system for coronary and pulmonary arteries received the Hubei Provincial Science and Technology Progress Award (Second Prize) and has been deployed in Wuhan Central Hospital, enhancing early screening efficiency and accuracy for cardiovascular diseases.

Speech Title: "Mask is Not All You Need: Box-supervised Polyp Segmentation in Images and Videos"

 Abstract: Pixel-level masks have long been considered essential for accurate polyp segmentation. However, these dense annotations are costly, time-consuming, and require expert knowledge, limiting the scalability of medical AI systems. In this talk, I will present a new line of research that challenges this reliance and shows that box-level annotations, even when coarse or sparse, can effectively support high-quality segmentation. We begin with IBoxCLA, which decouples learning of location, size, and shape, using proxy maps and contrastive latent anchors to achieve accurate segmentation from bounding boxes instead of pixel masks. Building on this, MonoBox addresses loosely drawn boxes by introducing a monotonicity constraint that guides learning in uncertain boundaries, removing the need for precise box-to-boundary alignment and enabling robust optimization under noisy supervision. Extending to videos, PSDNet reduces supervision to a single annotated frame per video. It combines tracking-based and semantic consistency teachers, with backward verification to ensure reliable pseudo labels across frames. Together, these methods demonstrate that accurate segmentation is achievable with minimal supervision. They substantially reduce annotation costs, improve generalization, and offer scalable solutions for image and video-based clinical applications.

Assoc. Prof. Gaoyan Zhang

Tianjin University, China