■增大的杯盘比(CDR)是青光眼视神经病变的标志。CDR的手动评估可能比自动化方法更不准确且更耗时。这里,我们试图开发和验证一种基于深度学习的算法,以自动确定眼底图像中的CDR.
■使用基于人群的观察研究的眼底数据估计CDR的算法开发。
■来自英国生物银行(UKBB)的总共181.768张眼底图像,Drishti_GS,和EyePACS。
■FastAI和PyTorch库用于在来自UKBB的眼底图像上训练基于卷积神经网络的模型。构建模型以确定图像可分级性(分类分析)以及估计CDR(回归分析)。然后使用来自EyePACS和Drishti_GS的多种族数据集验证最佳性能模型用于青光眼筛查。
■接收器工作特性曲线和确定系数下的面积。
■我们的分级模型vgg19_batch归一化(bn)在16.045个图像的验证集上实现了97.13%的精度,精度为99.26%,接收器工作特性曲线下面积为96.56%。使用回归分析,我们的最佳性能模型(在vgg19_bn架构上训练)获得了0.8514的确定系数(95%置信区间[CI]:0.8459-0.8568),而在用于确定CDR的12.183个验证图像集上,均方误差为0.0050(95%CI:0.0048-0.0051),平均绝对误差为0.0551(95%CI:0.0543-0.0559).使用20个类别的0.2的容差将回归点转换为分类指标;分类指标实现了99.20%的准确度。EyePACS数据集(98.172健康,3270青光眼)然后用于外部验证青光眼分类模型,准确地说,灵敏度,特异性为82.49%,72.02%,82.83%,分别。
■我们的模型在确定图像可分级性和估计CDR方面是精确的。尽管我们的人工智能衍生的CDR估计实现了高精度,青光眼筛查的CDR阈值将根据其他临床参数而变化.
■专有或商业披露可在本文末尾的脚注和披露中找到。
UNASSIGNED: An enlarged cup-to-disc ratio (CDR) is a hallmark of glaucomatous optic neuropathy. Manual assessment of the CDR may be less accurate and more time-consuming than automated methods. Here, we sought to develop and validate a deep learning-based algorithm to automatically determine the CDR from fundus images.
UNASSIGNED: Algorithm development for estimating CDR using fundus data from a population-based observational study.
UNASSIGNED: A total of 181 768 fundus images from the United Kingdom Biobank (UKBB), Drishti_GS, and EyePACS.
UNASSIGNED: FastAI and PyTorch libraries were used to train a convolutional neural network-based model on fundus images from the UKBB. Models were constructed to determine image gradability (classification analysis) as well as to estimate CDR (regression analysis). The best-performing model was then validated for use in glaucoma screening using a multiethnic dataset from EyePACS and Drishti_GS.
UNASSIGNED: The area under the receiver operating characteristic curve and coefficient of determination.
UNASSIGNED: Our gradability model vgg19_batch normalization (bn) achieved an accuracy of 97.13% on a validation set of 16 045 images, with 99.26% precision and area under the receiver operating characteristic curve of 96.56%. Using regression analysis, our best-performing model (trained on the vgg19_bn architecture) attained a coefficient of determination of 0.8514 (95% confidence interval [CI]: 0.8459-0.8568), while the mean squared error was 0.0050 (95% CI: 0.0048-0.0051) and mean absolute error was 0.0551 (95% CI: 0.0543-0.0559) on a validation set of 12 183 images for determining CDR. The regression point was converted into classification metrics using a tolerance of 0.2 for 20 classes; the classification metrics achieved an accuracy of 99.20%. The EyePACS dataset (98 172 healthy, 3270 glaucoma) was then used to externally validate the model for glaucoma classification, with an accuracy, sensitivity, and specificity of 82.49%, 72.02%, and 82.83%, respectively.
UNASSIGNED: Our models were precise in determining image gradability and estimating CDR. Although our artificial intelligence-derived CDR estimates achieve high accuracy, the CDR threshold for glaucoma screening will vary depending on other clinical parameters.
UNASSIGNED: Proprietary or commercial disclosure may be found in the Footnotes and Disclosures at the end of this article.