◇ 2025年9月29日(月)加齢研セミナーのご案内
| 日時： | 2025年9月29日（月）13時～14時 |
| 場所： | 加齢医学研究所 スマート・エイジング研究棟6階 セミナー室 |
| 演題： | 医療における人工知能：眼疾患への応用とそれ以上の可能性 |
| 講師： | Yalin, Zheng先生 |
| 所属： | リヴァプール大学 健康生命科学部 |
| 要旨： | This talk presents cutting-edge AI applications in healthcare, showcasing innovative deep learning approaches for medical image analysis and diagnosis. The speaker demonstrates several breakthrough contributions including a graph neural network framework for retinal disease segmentation that achieves state-of-the-art performance in glaucoma detection, a multiple instance learning system for cancer classification in gigapixel histology images, and a longitudinal deep learning model for predicting age-related macular degeneration that incorporates temporal patient data over multiple years. The research extends to diabetic neuropathy diagnosis using confocal microscopy, ECG analysis from printed records, and novel approaches that bridge traditional mathematical models with modern neural networks. However, the presentation critically addresses the significant challenges facing AI in healthcare, citing systematic reviews that reveal most AI diagnostic research lacks clinical utility due to methodological flaws, biases, and poor generalizability across different populations. The speaker emphasizes the urgent need for responsible AI development that goes beyond academic exercises to include comprehensive lifecycle thinking, proper validation, bias mitigation, explainability for clinical acceptance, and adherence to regulatory standards, advocating for collaborative efforts between AI researchers, clinicians, and regulatory bodies to ensure safe and effective clinical implementation while maintaining ethical considerations and patient safety. |
| 担当： | 瀧 靖之（臨床加齢医学研究分野・内線8559） |
| 連絡先： | オガワ 淑水（臨床加齢医学研究分野・内線8492） |