I(eye)-SCREEN: A real-world AI-based infrastructure for screening and prediction of progression in age-related macular degeneration (AMD) providing accessible shared care

I(eye)-Screen的目的是开发一种基于人工智能（AI）的诊断决策支持系统，用于在视力丧失之前的早期阶段筛查和监测年龄相关性黄斑变性（AMD）。晚期老年性黄斑变性是50岁以下法定失明的主要原因，有1.1亿人处于危险之中。该多学科联盟汇集了临床视网膜专家网络、从事人工智能开发前沿工作的计算机科学家、以社区为基础的验光师/验光师基础设施以及在数字平台性能方面经验丰富的中小企业，通过基于云的基础设施开发创新且值得信赖的人工智能工具，用于广泛、实时的AMD筛查和监控。为了实现在早期AMD中找到“大海捞针”的宏伟目标，光学相干断层扫描（OCT），一种高分辨率，轻松的成像方式，在广泛的体积扫描中提供了视网膜的详细特征。将开发突破性的医学成像人工智能方法，从稀疏的纵向OCT数据中实现数据高效和稳健的学习，以系统地分析密集的数据量，并识别疾病活动的（亚）临床标志物。整个欧洲的临床站点将收集纵向队列服务，使用眼科诊所提供的高端OCT设备校准和微调算法。然后，创新的人工智能技术将被创造出来，将检测和监控工具转移到隔壁验光师办公室使用的低成本设备上。该项目的时机与最近监管部门批准的首个治疗方法完全吻合，该疗法可以阻止AMD主要萎缩性类型的进展。由此产生的基于人工智能的“共享护理”策略为视力维持护理提供了不受限制的可及性，具有最大的健康公平性，并为筛查反映视网膜生物标志物的全身性、心血管和神经退行性疾病提供了榜样。

The aim of I(eye)-Screen is to develop an artificial intelligence (AI)-based diagnostic decision support system for screening and monitoring of age-related macular degeneration (AMD) at an early stage before vision loss occurs. Late AMD is the leading cause of legal blindness >50 years with 110 mio individuals at risk. The multidisciplinary consortium brings together a network of clinical retina experts, computer scientists working at the cutting edge of AI development, an infrastructure of community-based opticians/optometrists and an SME experienced in digital platform performance to develop innovative and trustworthy AI tools for broad, real-time AMD screening and monitoring via a cloud-based infrastructure with unlimited access. To achieve the ambitious goal of finding “the needle in the haystack” in early AMD, Optical Coherence Tomography (OCT), a high-resolution, effortless imaging modality is used providing a detailed characterization of the retina in extensive volumetric scans. Breakthrough AI approaches for medical imaging will be developed to enable data-efficient and robust learning from sparse longitudinal OCT data to systematically analyse dense data volumes and identify (sub)clinical markers of disease activity. Clinical sites throughout Europe will collect a longitudinal cohort serving for calibrating and fine-tuning algorithms using the high-end OCT device available at eye clinics. Innovative AI technology will then be created to transferthe detection and monitoring tools to low-cost devices used in next door opticians’/optometrists’ offices. The timing of the project perfectly fits the recent regulatory approval of the first therapy to halt progression of the major atrophic type of AMD. The resulting AIbased “shared care” strategy offers unrestricted accessibility to vision-maintaining care with greatest health equity and provides a role model for screening for systemic, cardiovascular and neurodegenerative disease reflecting retinal biomarkers.