Predicting Axial Length From Choroidal Thickness on Optical Coherence Tomography Images With Machine Learning Based Algorithms.
Predicting Axial Length From Choroidal Thickness on Optical Coherence Tomography Images With Machine Learning Based Algorithms.
复制标题
DOI：
10.3389/fmed.2022.850284
复制
发表时间：
2022
影响因子：
3.9
通讯作者：
中科院分区：
文献类型：
--
作者：
我们制定并测试了集合学习模型，从以中央凹为中心的二维单光学相干断层扫描（OCT）图像中显示的脉络膜厚度（CT）中分类轴向长度（AXL）。回顾性横断面研究。我们分析了188名患者355只眼睛的710张OCT图像。每只眼有2张OCT图像。从每张图像的3个点估计CT值。我们使用了五种机器学习基础算法来构建分类器。本研究训练并验证了基于二分类（AXL <或> 26 mm）和多分类（AXL < 22 mm, 22 ~ 26 mm, > 26 mm）的AXL眼分类模型。在使用Pearson相关系数、lasso模式搜索算法和方差膨胀因子进行分析后，没有多余或重复的特征。其中鼻侧CT与AXL的相关性最高，其次为中部。在二值分类中，准确率、召回率、阳性预测值（PPV）、阴性预测值（NPV）、F1评分和ROC曲线下面积（AUC）值分别为94.37、100、90.91、100、86.67和95.61%，达到了较高的准确率。在多类分类中，准确率、加权召回率、加权PPV、加权NPV、加权F1评分和宏观AUC分别为88.73、88.73、91.21、85.83、87.42和93.42%，分类器的准确率也很高。如OCT图像所示，我们的二分类器和多分类器可以很好地从CT中分类AXL。我们证明了所提出的分类器的有效性，并为医生提供了一个辅助工具。
We formulated and tested ensemble learning models to classify axial length (AXL) from choroidal thickness (CT) as indicated on fovea-centered, 2D single optical coherence tomography (OCT) images. Retrospective cross-sectional study. We analyzed 710 OCT images from 355 eyes of 188 patients. Each eye had 2 OCT images. The CT was estimated from 3 points of each image. We used five machine-learning base algorithms to construct the classifiers. This study trained and validated the models to classify the AXLs eyes based on binary (AXL < or > 26 mm) and multiclass (AXL < 22 mm, between 22 and 26 mm, and > 26 mm) classifications. No features were redundant or duplicated after an analysis using Pearson’s correlation coefficient, LASSO-Pattern search algorithm, and variance inflation factors. Among the positions, CT at the nasal side had the highest correlation with AXL followed by the central area. In binary classification, our classifiers obtained high accuracy, as indicated by accuracy, recall, positive predictive value (PPV), negative predictive value (NPV), F1 score, and area under ROC curve (AUC) values of 94.37, 100, 90.91, 100, 86.67, and 95.61%, respectively. In multiclass classification, our classifiers were also highly accurate, as indicated by accuracy, weighted recall, weighted PPV, weighted NPV, weighted F1 score, and macro AUC of 88.73, 88.73, 91.21, 85.83, 87.42, and 93.42%, respectively. Our binary and multiclass classifiers classify AXL well from CT, as indicated on OCT images. We demonstrated the effectiveness of the proposed classifiers and provided an assistance tool for physicians.
登录
查看更多内容
影响因子：
4.4
作者：
Ikuno, Yasushi;Kawaguchi, Kana;Yasuno, Yoshiaki
通讯作者：
Yasuno, Yoshiaki
影响因子：
1.9
作者：
de Asis Bartol-Puyal, Francisco;Isanta, Carlos;Pablo, Luis
通讯作者：
Pablo, Luis
DOI：
10.1137/0907087
发表时间：
1986-10-01
期刊：
SIAM JOURNAL ON SCIENTIFIC AND STATISTICAL COMPUTING
影响因子：
--
作者：
SANTOSA, F;SYMES, WW
通讯作者：
SYMES, WW
影响因子：
15.1
作者：
Dong L;Yang Q;Zhang RH;Wei WB
通讯作者：
Wei WB
影响因子：
3.7
作者：
Omodaka K;An G;Tsuda S;Shiga Y;Takada N;Kikawa T;Takahashi H;Yokota H;Akiba M;Nakazawa T
通讯作者：
Nakazawa T