A foundation model for generalizable disease detection from retinal images.
A foundation model for generalizable disease detection from retinal images.
复制标题
DOI：
10.1038/s41586-023-06555-x
复制
发表时间：
2023-10
期刊：
影响因子：
64.8
通讯作者：
Keane, Pearse A.
中科院分区：
文献类型：
--
作者：
Zhou, Yukun;Chia, Mark A.;Wagner, Siegfried K.;Ayhan, Murat S.;Williamson, Dominic J.;Struyven, Robbert R.;Liu, Timing;Xu, Moucheng;Lozano, Mateo G.;Woodward-Court, Peter;Kihara, Yuka;Altmann, Andre;Lee, Aaron Y.;Topol, Eric J.;Denniston, Alastair K.;Alexander, Daniel C.;Keane, Pearse A.
医学人工智能(AI)在识别视网膜图像中的健康状况迹象以及加快眼部疾病和系统性疾病的诊断方面提供了巨大的潜力。然而，人工智能模型的开发需要大量的注释，而且模型通常是特定于任务的，对不同临床应用的泛化能力有限。在这里，我们提出了RETFound，这是一个视网膜图像的基础模型，它从未标记的视网膜图像中学习概括的表示，并为几个应用中的标记高效模型适应提供了基础。具体来说，RETFound通过自我监督学习的方式在160万 未标记的视网膜图像上进行训练，然后适应具有明确标记的疾病检测任务。我们表明，自适应RETFound在诊断和预测危及视力的眼病以及预测复杂系统性疾病(如心力衰竭和心肌梗死)的事件方面始终优于几个比较模型，使用更少的标记数据。RETFound提供了一个通用的解决方案，可以提高模型性能，减轻专家的注释工作量，使视网膜成像能够广泛应用于临床。RETFound是一个视网膜图像的基础模型，它从未标记的图像中学习泛化表示，通过自监督学习对160万 的未标记图像进行训练，然后适应于具有明确标记的疾病检测任务。
Medical artificial intelligence (AI) offers great potential for recognizing signs of health conditions in retinal images and expediting the diagnosis of eye diseases and systemic disorders. However, the development of AI models requires substantial annotation and models are usually task-specific with limited generalizability to different clinical applications. Here, we present RETFound, a foundation model for retinal images that learns generalizable representations from unlabelled retinal images and provides a basis for label-efficient model adaptation in several applications. Specifically, RETFound is trained on 1.6 million unlabelled retinal images by means of self-supervised learning and then adapted to disease detection tasks with explicit labels. We show that adapted RETFound consistently outperforms several comparison models in the diagnosis and prognosis of sight-threatening eye diseases, as well as incident prediction of complex systemic disorders such as heart failure and myocardial infarction with fewer labelled data. RETFound provides a generalizable solution to improve model performance and alleviate the annotation workload of experts to enable broad clinical AI applications from retinal imaging. RETFound, a foundation model for retinal images that learns generalizable representations from unlabelled images, is trained on 1.6 million unlabelled images by self-supervised learning and then adapted to disease detection tasks with explicit labels.
登录
查看更多内容
使用深度神经网络自动检测视网膜照片中的 39 种眼底疾病和状况
影响因子：
16.6
作者：
Cen LP;Ji J;Lin JW;Ju ST;Lin HJ;Li TP;Wang Y;Yang JF;Liu YF;Tan S;Tan L;Li D;Wang Y;Zheng D;Xiong Y;Wu H;Jiang J;Wu Z;Huang D;Shi T;Chen B;Yang J;Zhang X;Luo L;Huang C;Zhang G;Huang Y;Ng TK;Chen H;Chen W;Pang CP;Zhang M
通讯作者：
Zhang M
影响因子：
64.8
作者：
Bycroft C;Freeman C;Petkova D;Band G;Elliott LT;Sharp K;Motyer A;Vukcevic D;Delaneau O;O'Connell J;Cortes A;Welsh S;Young A;Effingham M;McVean G;Leslie S;Allen N;Donnelly P;Marchini J
通讯作者：
Marchini J
影响因子：
10.9
作者：
Ayhan, Murat Seckin;Kuhlewein, Laura;Berens, Philipp
通讯作者：
Berens, Philipp
影响因子：
20.1
作者：
Guenthner, Roman;Hanssen, Henner;Schmadere, Christoph
通讯作者：
Schmadere, Christoph
影响因子：
3.7
作者：
Ahn JM;Kim S;Ahn KS;Cho SH;Lee KB;Kim US
通讯作者：
Kim US