香港中文大学研发的视网膜图像分析技术，作为解决未确诊的糖尿病的方法。未确诊的糖尿病是一个全球性的健康问题，大约24%–75%病例未被诊断，因此导致大量糖尿病并发症和巨额医疗费用。目前许多糖尿病诊断方法依赖于代谢指数，并且存在相当大的变异性。相比之下，视网膜图像分析以数字方法表述了整体血糖状态，提供一个稳定的标记，而且快速、方便、及非侵入性，超过90%的敏感性和特异性。
Yangpu Entrepreneurship Star Competition “Entrepreneurship Star” Award, 22 March 2024
The Chinese University of Hong Kong developed Automatic Retinal Image Analysis (ARIA) technology to solve undiagnosed diabetes problems. Undiagnosed diabetes is a global health problem, with approximately 24%–75% of cases going undiagnosed, resulting in numerous diabetes complications and huge medical costs. Many current diabetes diagnostic methods rely on metabolic indices, and considerable variability exists. In contrast, retinal image analysis digitally represents overall glycemic status, providing a stable marker that is fast, convenient, and non-invasive, with over 90% sensitivity and specificity.
References:
Zee B, Lee J, Lai M, Chee P, Rafferty J, Thomas R, Owens D. “Digital solution for detection of undiagnosed diabetes using machine learning-based retinal image analysis.” BMJ Open Diabetes Res Care. 2022 Dec; 10(6): e002914 https://drc.bmj.com/content/10/6/e002914
Press Release 新闻稿
杨浦首发2024“人才秀带”三大新政：撬动教育资源、推动产业融合和保障安居环境