Automatic Localization of the Scleral Spur Using Deep Learning and Ultrasound Biomicroscopy.
Automatic Localization of the Scleral Spur Using Deep Learning and Ultrasound Biomicroscopy.
复制标题
DOI：
10.1167/tvst.10.9.28
复制
发表时间：
2021-08-02
影响因子：
3
通讯作者：
Lin S
中科院分区：
文献类型：
--
作者：
Wang W;Wang L;Wang T;Wang X;Zhou S;Yang J;Lin S
本研究的目的是开发一个卷积神经网络（CNN）的超声生物显微镜（UBM）图像的开角眼的巩膜棘的自动定位。获取UBM图像，一位青光眼专家提供所有图像中巩膜突位置的参考坐标。开发了基于EfficientNetB3架构的CNN模型来检测每个图像中的巩膜毛刺。使用预测误差和欧氏距离来评估CNN模型的定位性能。使用由专家提供并由CNN模型预测的巩膜刺位置测量和分析小梁虹膜角500（TIA500）和开角距离500（AOD500）。CNN使用2328张图像的训练数据集开发，并使用258张图像的独立数据集进行测试。CNN模型的平均绝对预测误差为X坐标48.06 ± 45.40 µm，Y坐标30.84 ± 27.03 µm。X坐标的平均绝对观察者内变异性为47.80 ± 44.45 µm，Y坐标的平均绝对观察者内变异性为29.50 ± 25.77 µm。CNN的平均欧几里得距离为60.41 ± 49.02 µm，观察者内平均欧几里得距离为59.78 ± 47.12 µm。对于所有测试图像，TIA500的平均绝对误差为1.26 ± 1.38度，AOD500的平均绝对误差为0.039 ± 0.051 mm。CNN可以检测开角眼的UBM图像上的巩膜棘，其性能与青光眼专家相似。用于自动化巩膜突定位的深度学习算法将有助于定量评估房角的开放和房角闭合的风险。
The purpose of this study was to develop a convolutional neural network (CNN) for automated localization of the scleral spur in ultrasound biomicroscopy (UBM) images of open-angle eyes. UBM images were acquired, and one glaucoma specialist provided reference coordinates of scleral spur locations in all images. A CNN model based on the EfficientNetB3 architecture was developed to detect the scleral spur in each image. The prediction errors and Euclidean distance were used to evaluate localization performance of the CNN model. Trabecular-iris angle 500 (TIA500) and angle-opening distance 500 (AOD500) were measured and analyzed using the scleral spur locations provided by the specialist and predicted by the CNN model. The CNN was developed using a training dataset of 2328 images and tested using an independent dataset of 258 images. The mean absolute prediction errors of CNN model were 48.06 ± 45.40 µm for X-coordinates and 30.84 ± 27.03 µm for Y-coordinates. The mean absolute intraobserver variability was 47.80 ± 44.45 µm for X-coordinates and 29.50 ± 25.77 µm for Y-coordinates. The mean Euclidean distance of the CNN was 60.41 ± 49.02 µm and the intraobserver mean Euclidean distance was 59.78 ± 47.12 µm. The mean absolute error in TIA500 was 1.26 ± 1.38 degrees for all test images and in AOD500 was 0.039 ± 0.051 mm. A CNN can detect the scleral spur on UBM images of open-angle eyes with performance similar to that of a glaucoma specialist. Deep learning algorithms for automating scleral spur localization would facilitate the quantitative assessment of the opening of the angle and the risk in angle closure.
登录
查看更多内容
DOI：
10.1001/jama.2014.3192
发表时间：
2014-05-14
期刊：
JAMA
影响因子：
--
作者：
Weinreb RN;Aung T;Medeiros FA
通讯作者：
Medeiros FA
影响因子：
10.9
作者：
Fu, Huazhu;Li, Fei;Zhou, Rouxi
通讯作者：
Zhou, Rouxi
影响因子：
4.2
作者：
Fu, Huazhu;Baskaran, Mani;Aung, Tin
通讯作者：
Aung, Tin
影响因子：
--
作者：
Leung, Christopher Kai-shun;Yung, Wing-ho;Lam, Dennis Shun-chiu
通讯作者：
Lam, Dennis Shun-chiu
影响因子：
3.1
作者：
Maslin JS;Barkana Y;Dorairaj SK
通讯作者：
Dorairaj SK