Predicting sex from retinal fundus photographs using automated deep learning.
Predicting sex from retinal fundus photographs using automated deep learning.
复制标题
DOI：
10.1038/s41598-021-89743-x
复制
发表时间：
2021-05-13
影响因子：
4.6
通讯作者：
Keane PA
中科院分区：
文献类型：
--
作者：
Korot E;Pontikos N;Liu X;Wagner SK;Faes L;Huemer J;Balaskas K;Denniston AK;Khawaja A;Keane PA
深度学习可能会改变医疗保健，但模型开发在很大程度上依赖于先进技术专业知识的可用性。在这里，我们介绍了由临床医生开发的无编码的深度学习模型，该模型根据视网膜眼底照片预测所报告的性别。一个模型接受了来自英国生物库数据集的84743张视网膜眼底照片的训练。对来自第三眼科转诊中心的252张眼底照片进行了外部验证。对于内部验证，无代码深度学习(CFDL)模型的接收器操作特征曲线(AUROC)下的面积为0.93。敏感度、特异度、阳性预测值和准确度分别为88.8%、83.6%、87.3%和86.5%，外部验证分别为83.9%、72.2%、78.2%和78.6%。临床医生目前还没有意识到男性和女性之间存在明显的视网膜特征差异，这突显了模型可解释性对这项任务的重要性。当黄斑中心凹病理出现在外部验证数据集中时，模型的表现明显较差，Acc：69.4%，而在健康眼睛中为85.4%，这表明中心凹是模型性能的显著区域，OR(95%CI)：0.36(0.19，0.70)p = 0.0022。自动机器学习(AutoML)可以使临床医生驱动的新见解和疾病生物标记物的自动发现成为可能。
Deep learning may transform health care, but model development has largely been dependent on availability of advanced technical expertise. Herein we present the development of a deep learning model by clinicians without coding, which predicts reported sex from retinal fundus photographs. A model was trained on 84,743 retinal fundus photos from the UK Biobank dataset. External validation was performed on 252 fundus photos from a tertiary ophthalmic referral center. For internal validation, the area under the receiver operating characteristic curve (AUROC) of the code free deep learning (CFDL) model was 0.93. Sensitivity, specificity, positive predictive value (PPV) and accuracy (ACC) were 88.8%, 83.6%, 87.3% and 86.5%, and for external validation were 83.9%, 72.2%, 78.2% and 78.6% respectively. Clinicians are currently unaware of distinct retinal feature variations between males and females, highlighting the importance of model explainability for this task. The model performed significantly worse when foveal pathology was present in the external validation dataset, ACC: 69.4%, compared to 85.4% in healthy eyes, suggesting the fovea is a salient region for model performance OR (95% CI): 0.36 (0.19, 0.70) p = 0.0022. Automated machine learning (AutoML) may enable clinician-driven automated discovery of novel insights and disease biomarkers.
登录
查看更多内容
影响因子：
3.7
作者：
Coppola G;Di Renzo A;Ziccardi L;Martelli F;Fadda A;Manni G;Barboni P;Pierelli F;Sadun AA;Parisi V
通讯作者：
Parisi V
影响因子：
4
作者：
Collins GS;de Groot JA;Dutton S;Omar O;Shanyinde M;Tajar A;Voysey M;Wharton R;Yu LM;Moons KG;Altman DG
通讯作者：
Altman DG
影响因子：
4.2
作者：
Cheung, Carol Y.;Thomas, George N.;Wong, Tien Yin
通讯作者：
Wong, Tien Yin
影响因子：
13.7
作者：
Owen CG;Rudnicka AR;Welikala RA;Fraz MM;Barman SA;Luben R;Hayat SA;Khaw KT;Strachan DP;Whincup PH;Foster PJ
通讯作者：
Foster PJ
影响因子：
3.7
作者：
Adhi M;Aziz S;Muhammad K;Adhi MI
通讯作者：
Adhi MI