Kuwayama Soichiro, Ayatsuka Yuji, Yanagisono Daisuke, Uta Takaki, Usui Hideaki, Kato Aki, Takase Noriaki, Ogura Yuichiro, Yasukawa Tsutomu
Department of Ophthalmology & Visual Science, Nagoya City University Graduate School of Medical Sciences, Nagoya, Japan.
Technology Laboratory, Cresco Ltd., Tokyo, Japan.
J Ophthalmol. 2019 Apr 9;2019:6319581. doi: 10.1155/2019/6319581. eCollection 2019.
Although optical coherence tomography (OCT) is essential for ophthalmologists, reading of findings requires expertise. The purpose of this study is to test deep learning with image augmentation for automated detection of chorioretinal diseases.
A retina specialist diagnosed 1,200 OCT images. The diagnoses involved normal eyes (=570) and those with wet age-related macular degeneration (AMD) (=136), diabetic retinopathy (DR) (=104), epiretinal membranes (ERMs) (=90), and another 19 diseases. Among them, 1,100 images were used for deep learning training, augmented to 59,400 by horizontal flipping, rotation, and translation. The remaining 100 images were used to evaluate the trained convolutional neural network (CNN) model.
Automated disease detection showed that the first candidate disease corresponded to the doctor's decision in 83 (83%) images and the second candidate disease in seven (7%) images. The precision and recall of the CNN model were 0.85 and 0.97 for normal eyes, 1.00 and 0.77 for wet AMD, 0.78 and 1.00 for DR, and 0.75 and 0.75 for ERMs, respectively. Some of rare diseases such as Vogt-Koyanagi-Harada disease were correctly detected by image augmentation in the CNN training.
Automated detection of macular diseases from OCT images might be feasible using the CNN model. Image augmentation might be effective to compensate for a small image number for training.
尽管光学相干断层扫描（OCT）对眼科医生至关重要，但解读检查结果需要专业知识。本研究的目的是测试使用图像增强技术的深度学习，以自动检测脉络膜视网膜疾病。
一位视网膜专家对1200张OCT图像进行了诊断。诊断包括正常眼睛（=570例）以及患有湿性年龄相关性黄斑变性（AMD）（=136例）、糖尿病性视网膜病变（DR）（=104例）、视网膜前膜（ERM）（=90例）和另外19种疾病的眼睛。其中，1100张图像用于深度学习训练，通过水平翻转、旋转和平移将其扩充至59400张。其余100张图像用于评估训练后的卷积神经网络（CNN）模型。
自动疾病检测显示，在83张（83%）图像中，第一个候选疾病与医生的诊断结果相符，在7张（7%）图像中，第二个候选疾病与医生的诊断结果相符。CNN模型对正常眼睛的精确率和召回率分别为0.85和0.97，对湿性AMD为1.00和0.77，对DR为0.78和1.00，对ERM为0.75和0.75。在CNN训练中，通过图像增强技术正确检测出了一些罕见疾病，如原田病。
使用CNN模型从OCT图像中自动检测黄斑疾病可能是可行的。图像增强技术可能有助于弥补训练图像数量较少的问题。
