发表机构
Emory University; National University of Singapore; Singapore Eye Research Institute, Singapore National Eye Centre; University of Ibadan; Duke University School of Medicine; Duke-NUS Graduate Medical School-Singapore; Georgia Institute of Technology; Emory Empathetic AI for Health Institute, Emory University; Abyss Processing Pte Ltd（埃默里大学; 新加坡国立大学; 新加坡国立眼科中心新加坡眼科研究所; 伊巴丹大学; 杜克大学医学院; 新加坡杜克-新加坡国立大学研究生医学院; 佐治亚理工学院; 埃默里大学埃默里共情健康人工智能研究所; Abyss Processing私人有限公司）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
本研究开发并评估了一种可解释AI框架，结合低成本手持相机进行青光眼筛查，在西非人群中性能接近台式相机，有望提升资源有限地区的社区筛查能力。
AI 中文摘要
目的：开发并评估一种可解释的人工智能（AI）框架，用于在西非人群中使用低成本便携式手持视网膜眼底照片进行青光眼筛查，并将其性能与临床台式眼底成像进行比较。方法：我们使用了尼日利亚一项基于社区的研究数据，该研究包含681名参与者（1362只眼），包括414只青光眼、478只疑似青光眼和470只非青光眼眼。眼底照片分别使用低成本手持便携式Volk Viva视网膜相机和Canon CR-2-AF台式相机获取。我们针对每种设备分别微调了组件模型，以执行血管分割、杯盘边界分割和特征提取，用于检测视神经头特征。最终分类模型结合这些组件，将扫描结果分类为青光眼、疑似青光眼或非青光眼。使用特征权重分析和梯度加权类激活映射进行解释。结果：模型在Volk Viva和Canon CR-2-AF图像上均表现良好：血管分割：0.98 Dice系数（DC）（Volk）和0.94 DC（Canon）；杯盘分割：0.95 DC（Volk）和0.96 DC（Canon）；视神经头特征检测：受试者工作特征曲线下面积（AUC）为0.83±0.03（Volk）和0.87±0.04（Canon）；分类模型：AUC为0.85±0.01（Volk）和0.93±0.01（Canon）。每张图像的报告提供模型决策置信度分数和决策依据可视化，以支持临床解释。结论：Volk Viva在组件模型中的结果与Canon CR-2-AF相当合理，在分类方面也相差不远。这表明，可解释AI结合低成本便携式成像可能增强社区层面的青光眼筛查，尤其是在专家获取和资源有限的场景中。
英文摘要
Purpose: To develop and evaluate an interpretable artificial intelligence (AI) framework for glaucoma screening from low-cost portable, handheld retinal fundus photographs in a West African population and to compare its performance with clinical tabletop fundus imaging. Methods: We used data from a community-based study of 681 participants (1,362 eyes) in Nigeria, comprising 414 glaucoma, 478 glaucoma suspect, and 470 non-glaucoma eyes. Fundus photographs were acquired using the low-cost handheld, portable Volk Viva retinal camera and the Canon CR-2-AF tabletop camera. We fine-tuned component models separately to each device to perform vessel segmentation, cup and disc boundary segmentation, and feature extraction to detect optic nerve head features. A final classification model combined these components to classify scans as glaucoma, glaucoma suspect or non-glaucoma. Feature-weight analysis and Gradient-weighted Class Activation Mapping were used for interpretation. Results: The models performed well on both Volk Viva and Canon CR-2-AF images: Vessel segmentation: 0.98 Dice Coefficient (DC) (Volk) and 0.94 DC (Canon); Cup and disc segmentation: 0.95 DC (Volk) and 0.96 DC (Canon); Optic nerve head feature detection: area under the receiver operating characteristic curve (AUCs) of 0.83$\pm$0.03 (Volk) and 0.87$\pm$0.04 (Canon); Classification model: AUCs of 0.85$\pm$0.01 (Volk) and 0.93$\pm$0.01 (Canon). Reports for each image, present model decision confidence scores and decision-rationale visualizations to support clinical interpretation. Conclusions: Volk Viva results were reasonably comparable to Canon CR-2-AF in the component models and not far behind in classification. This shows that interpretable AI combined with low-cost, portable imaging may enhance community-level glaucoma screening, especially in settings with limited specialist access and resources.
Comments31 pages, 2 Tables, 5 Figures, 1 Supplementary Material
