Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Division of Endocrinology and Metabolism, Department of Internal Medicine, St. Vincent's Hospital, College of Medicine, The Catholic University of Korea, Seoul, Republic of Korea.
Department of Biostatistics, Epidemiology and Informatics, Perelman School of Medicine, University of Pennsylvania, Philadelphia, PA, USA; Department of Computer Engineering, Ajou University, Suwon, Republic of Korea; Institute for Biomedical Informatics, University of Pennsylvania, Philadelphia, PA, USA.
Nutr Metab Cardiovasc Dis. 2022 May;32(5):1218-1226. doi: 10.1016/j.numecd.2022.01.010. Epub 2022 Jan 13.
We aimed to develop and evaluate a non-invasive deep learning algorithm for screening type 2 diabetes in UK Biobank participants using retinal images.
The deep learning model for prediction of type 2 diabetes was trained on retinal images from 50,077 UK Biobank participants and tested on 12,185 participants. We evaluated its performance in terms of predicting traditional risk factors (TRFs) and genetic risk for diabetes. Next, we compared the performance of three models in predicting type 2 diabetes using 1) an image-only deep learning algorithm, 2) TRFs, 3) the combination of the algorithm and TRFs. Assessing net reclassification improvement (NRI) allowed quantification of the improvement afforded by adding the algorithm to the TRF model. When predicting TRFs with the deep learning algorithm, the areas under the curve (AUCs) obtained with the validation set for age, sex, and HbA1c status were 0.931 (0.928-0.934), 0.933 (0.929-0.936), and 0.734 (0.715-0.752), respectively. When predicting type 2 diabetes, the AUC of the composite logistic model using non-invasive TRFs was 0.810 (0.790-0.830), and that for the deep learning model using only fundus images was 0.731 (0.707-0.756). Upon addition of TRFs to the deep learning algorithm, discriminative performance was improved to 0.844 (0.826-0.861). The addition of the algorithm to the TRFs model improved risk stratification with an overall NRI of 50.8%.
Our results demonstrate that this deep learning algorithm can be a useful tool for stratifying individuals at high risk of type 2 diabetes in the general population.
我们旨在开发和评估一种使用视网膜图像对英国生物库参与者进行 2 型糖尿病筛查的非侵入性深度学习算法。
用于预测 2 型糖尿病的深度学习模型在 50077 名英国生物库参与者的视网膜图像上进行了训练，并在 12185 名参与者上进行了测试。我们评估了其预测传统风险因素（TRFs）和糖尿病遗传风险的性能。接下来，我们比较了使用 1）仅图像的深度学习算法，2）TRFs，3）算法和 TRFs 的组合，三种模型预测 2 型糖尿病的性能。评估净重新分类改善（NRI）可以量化将算法添加到 TRF 模型中所带来的改善。当使用深度学习算法预测 TRFs 时，验证集的曲线下面积（AUCs）获得年龄，性别和 HbA1c 状态分别为 0.931（0.928-0.934），0.933（0.929-0.936）和 0.734（0.715-0.752）。当预测 2 型糖尿病时，使用非侵入性 TRFs 的组合逻辑模型的 AUC 为 0.810（0.790-0.830），而仅使用眼底图像的深度学习模型的 AUC 为 0.731（0.707-0.756）。在将 TRFs 添加到深度学习算法后，判别性能提高到 0.844（0.826-0.861）。将算法添加到 TRFs 模型可改善风险分层，总体 NRI 为 50.8％。
我们的结果表明，该深度学习算法可以成为一般人群中 2 型糖尿病高危个体分层的有用工具。
