论文标题
多模式，多任务，多注意（M3）的深度学习检测：迈向与年龄相关的黄斑变性的自动化和可访问的分类
Multi-modal, multi-task, multi-attention (M3) deep learning detection of reticular pseudodrusen: towards automated and accessible classification of age-related macular degeneration
论文作者
论文摘要
客观网状假发素（RPD）是与年龄相关的黄斑变性（AMD）的关键特征，是由标准颜色摄影（CFP）的人类专家检测到的很差，通常需要先进的成像方式，例如基底底基自动荧光（FAF）。目的是在RPD检测中开发和评估新型的“ M3”深度学习框架的性能。材料和方法开发了一个深度学习框架M3，以便单独使用CFP，单独使用FAF或两者兼有> 8000 CFP-FAF图像对，以准确检测RPD的存在（与年龄相关的眼病研究2）。 M3框架包括多模式（从单个图像模式或多个图像模式检测），多任务（同时训练不同的任务以提高概括性）和多注意（改善结合功能表示）操作。将RPD检测的性能与最先进的深度学习模型和13位眼科医生进行了比较。还评估了检测其他两个AMD特征（地理萎缩和色素异常）的性能。 RPD检测的结果，M3在接收器工作特性（AUROC）下实现了区域0.832、0.931和0.933，单独使用CFP，单独使用FAF和两者。 CFP上的M3性能大大优于人类视网膜专家（中位F1得分为0.644对0.350）。外部验证（在鹿特丹研究，荷兰）仅在CFP上表现出很高的精度（AUROC 0.965）。 M3框架还准确检测到地理萎缩和色素异常（分别为0.909和0.912），证明了其普遍性。结论本研究表明了对新型深度学习框架的成功发展，强大的评估以及外部验证，该框架可以访问，准确和自动化的AMD诊断和预后。
Objective Reticular pseudodrusen (RPD), a key feature of age-related macular degeneration (AMD), are poorly detected by human experts on standard color fundus photography (CFP) and typically require advanced imaging modalities such as fundus autofluorescence (FAF). The objective was to develop and evaluate the performance of a novel 'M3' deep learning framework on RPD detection. Materials and Methods A deep learning framework M3 was developed to detect RPD presence accurately using CFP alone, FAF alone, or both, employing >8000 CFP-FAF image pairs obtained prospectively (Age-Related Eye Disease Study 2). The M3 framework includes multi-modal (detection from single or multiple image modalities), multi-task (training different tasks simultaneously to improve generalizability), and multi-attention (improving ensembled feature representation) operation. Performance on RPD detection was compared with state-of-the-art deep learning models and 13 ophthalmologists; performance on detection of two other AMD features (geographic atrophy and pigmentary abnormalities) was also evaluated. Results For RPD detection, M3 achieved area under receiver operating characteristic (AUROC) 0.832, 0.931, and 0.933 for CFP alone, FAF alone, and both, respectively. M3 performance on CFP was very substantially superior to human retinal specialists (median F1-score 0.644 versus 0.350). External validation (on Rotterdam Study, Netherlands) demonstrated high accuracy on CFP alone (AUROC 0.965). The M3 framework also accurately detected geographic atrophy and pigmentary abnormalities (AUROC 0.909 and 0.912, respectively), demonstrating its generalizability. Conclusion This study demonstrates the successful development, robust evaluation, and external validation of a novel deep learning framework that enables accessible, accurate, and automated AMD diagnosis and prognosis.
