Deep Learning Algorithm Detects Presence of Disorganization of Retinal Inner Layers (DRIL)-An Early Imaging Biomarker in Diabetic Retinopathy.
Deep Learning Algorithm Detects Presence of Disorganization of Retinal Inner Layers (DRIL)-An Early Imaging Biomarker in Diabetic Retinopathy.
复制标题
深度学习算法检测出视网膜内层（DRIL）的混乱的存在 - 糖尿病性视网膜病中的早期成像生物标志物。
DOI：
10.1167/tvst.12.7.6
复制
发表时间：
2023-07-03
影响因子：
3
通讯作者：
Yuan, Alex
中科院分区：
文献类型：
--
作者：
Singh, Rupesh;Singuri, Srinidhi;Batoki, Julia;Lin, Kimberly;Luo, Shiming;Hatipoglu, Dilara;Anand-Apte, Bela;Yuan, Alex
关键词：
开发和训练一种基于深度学习的算法，用于检测光学相干断层扫描(OCT)上的视网膜内层紊乱(DRIL)，以筛查一组糖尿病视网膜病变(DR)患者。在这项横断面研究中，受试者年龄在18岁以上，在2009年1月至2019年9月期间接受了ICD-9/10诊断为伴有和不伴有视网膜病变的2型糖尿病，并进行了Cirrus HD-OCT成像。在应用纳入和排除标准后，最终共有664名患者(1201只眼的5992例B超)被纳入分析。CirrusHD-OCT的五行水平栅格扫描是从共享的电子健康记录中获得的。两名训练有素的评分员对扫描是否存在Dril进行了评估。第三位医生评分员对任何分歧进行了仲裁。在所分析的5,992次B超中，有1,397次(∼30%)显示存在DRIL。分级扫描被用于标记用于卷积神经网络(CNN)开发和训练的训练数据。在单CPU系统上，表现最好的有线电视新闻网训练花费了∼35分钟。标签数据分为90：10，用于内部培训/验证和外部测试目的。通过这种训练，我们的深度学习网络能够预测新的OCT扫描中是否存在DRIL，准确率为88.3%，特异度为90.0%，敏感度为82.9%，Matthews相关系数为0.7。本研究表明，基于深度学习的OCT分类算法可以用于DRL的快速自动识别。这一开发的工具可以在研究和临床决策环境中帮助筛查Dril。深度学习算法可以在OCT扫描中检测到视网膜内层的紊乱。
To develop and train a deep learning-based algorithm for detecting disorganization of retinal inner layers (DRIL) on optical coherence tomography (OCT) to screen a cohort of patients with diabetic retinopathy (DR). In this cross-sectional study, subjects over age 18, with ICD-9/10 diagnoses of type 2 diabetes with and without retinopathy and Cirrus HD-OCT imaging performed between January 2009 to September 2019 were included in this study. After inclusion and exclusion criteria were applied, a final total of 664 patients (5992 B-scans from 1201 eyes) were included for analysis. Five-line horizontal raster scans from Cirrus HD-OCT were obtained from the shared electronic health record. Two trained graders evaluated scans for presence of DRIL. A third physician grader arbitrated any disagreements. Of 5992 B-scans analyzed, 1397 scans (∼30%) demonstrated presence of DRIL. Graded scans were used to label training data for the convolution neural network (CNN) development and training. On a single CPU system, the best performing CNN training took ∼35 mins. Labeled data were divided 90:10 for internal training/validation and external testing purpose. With this training, our deep learning network was able to predict the presence of DRIL in new OCT scans with a high accuracy of 88.3%, specificity of 90.0%, sensitivity of 82.9%, and Matthews correlation coefficient of 0.7. The present study demonstrates that a deep learning-based OCT classification algorithm can be used for rapid automated identification of DRIL. This developed tool can assist in screening for DRIL in both research and clinical decision-making settings. A deep learning algorithm can detect disorganization of retinal inner layers in OCT scans.
登录
查看更多内容
影响因子：
8.1
作者：
Radwan, Salma H.;Soliman, Ahmed Z.;Koozekanani, Dara D.
通讯作者：
Koozekanani, Dara D.
影响因子：
8.1
作者：
Babiuch, Amy S.;Han, Michael;Singh, Rishi P.
通讯作者：
Singh, Rishi P.
影响因子：
4.6
作者：
Venerito, Vincenzo;Angelini, Orazio;Iannone, Florenzo
通讯作者：
Iannone, Florenzo
影响因子：
4.4
作者：
Chicco, Davide;Jurman, Giuseppe
通讯作者：
Jurman, Giuseppe
影响因子：
3
作者：
Nadri, Gauhar;Saxena, Sandeep;Kruzliak, Peter
通讯作者：
Kruzliak, Peter