arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2011.05142cs.CVcs.AIcs.LGeess.IV

Multi-modal, multi-task, multi-attention (M3) deep learning detection of reticular pseudodrusen: towards automated and accessible classification of age-related macular degeneration

National Center for Biotechnology Information, National Library of Medicine, National Institutes of Health (NIH)（美国国立卫生研究院国家医学图书馆国家生物技术信息中心）；National Eye Institute, National Institutes of Health（美国国立卫生研究院国家眼科研究所）；University of Wisconsin（威斯康星大学）；Erasmus Medical Center（伊拉斯谟医学中心）；Uniformed Services University of the Health Sciences（美国卫生 uniformed 大学）；University of Texas at Arlington（德克萨斯大学阿灵顿分校）；Clinical Center, National Institutes of Health（美国国立卫生研究院临床中心）
Qingyu Chen, Tiarnan D. L. Keenan, Alexis Allot, Yifan Peng, Elvira Agrón, Amitha Domalpally, Caroline C. W. Klaver, Daniel T. Luttikhuizen, Marcus H. Colyer, C… 展开作者

Qingyu Chen, Tiarnan D. L. Keenan, Alexis Allot, Yifan Peng, Elvira Agrón, Amitha Domalpally, Caroline C. W. Klaver, Daniel T. Luttikhuizen, Marcus H. Colyer, Catherine A. Cukras, Henry E. Wiley, M. Teresa Magone, Chantal Cousineau-Krieger, Wai T. Wong, Yingying Zhu, Emily Y. Chew, Zhiyong Lu

备注：
5 figures and 4 tables, To appear in Journal of the American Medical Informatics Association
阅读 PDF
arXiv 页面 ↗HTML 原文 ↗

英文摘要：

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.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

↑