Liu Baoyi, Zhang Bin, Hu Yijun, Cao Dan, Yang Dawei, Wu Qiaowei, Hu Yu, Yang Jingwen, Peng Qingsheng, Huang Manqing, Zhong Pingting, Dong Xinran, Feng Songfu, Li Tao, Lin Haotian, Cai Hongmin, Yang Xiaohong, Yu Honghua
Guangdong Eye Institute, Department of Ophthalmology, Guangdong Provincial People's Hospital, Guangdong Academy of Medical Sciences, The Second School of Clinical Medicine, Southern Medical University, Guangzhou, China.
School of Computer Science and Engineering, South China University of Technology, Guangzhou, China.
Ann Transl Med. 2021 Jan;9(1):43. doi: 10.21037/atm-20-1431.
This study aimed to predict the treatment outcomes in patients with diabetic macular edema (DME) after 3 monthly anti-vascular endothelial growth factor (VEGF) injections using machine learning (ML) based on pretreatment optical coherence tomography (OCT) images and clinical variables.
An ensemble ML system consisting of four deep learning (DL) models and five classical machine learning (CML) models was developed to predict the posttreatment central foveal thickness (CFT) and the best-corrected visual acuity (BCVA). A total of 363 OCT images and 7,587 clinical data records from 363 eyes were included in the training set (304 eyes) and external validation set (59 eyes). The DL models were trained using the OCT images, and the CML models were trained using the OCT images features and clinical variables. The predictive posttreatment CFT and BCVA values were compared with true outcomes obtained from the medical records.
For CFT prediction, the mean absolute error (MAE), root mean square error (RMSE), and R of the best-performing model in the training set was 66.59, 93.73, and 0.71, respectively, with an area under receiver operating characteristic curve (AUC) of 0.90 for distinguishing the eyes with good anatomical response. The MAE, RMSE, and R was 68.08, 97.63, and 0.74, respectively, with an AUC of 0.94 in the external validation set. For BCVA prediction, the MAE, RMSE, and R of the best-performing model in the training set was 0.19, 0.29, and 0.60, respectively, with an AUC of 0.80 for distinguishing eyes with a good functional response. The external validation achieved a MAE, RMSE, and R of 0.13, 0.20, and 0.68, respectively, with an AUC of 0.81.
Our ensemble ML system accurately predicted posttreatment CFT and BCVA after anti-VEGF injections in DME patients, and can be used to prospectively assess the efficacy of anti-VEGF therapy in DME patients.
本研究旨在基于治疗前光学相干断层扫描（OCT）图像和临床变量，使用机器学习（ML）预测糖尿病性黄斑水肿（DME）患者在接受3个月每月一次抗血管内皮生长因子（VEGF）注射后的治疗结果。
开发了一个由四个深度学习（DL）模型和五个经典机器学习（CML）模型组成的集成ML系统，以预测治疗后的中心凹厚度（CFT）和最佳矫正视力（BCVA）。训练集（304只眼）和外部验证集（59只眼）共纳入了363只眼的363张OCT图像和7587条临床数据记录。DL模型使用OCT图像进行训练，CML模型使用OCT图像特征和临床变量进行训练。将预测的治疗后CFT和BCVA值与从医疗记录中获得的真实结果进行比较。
对于CFT预测，训练集中表现最佳的模型的平均绝对误差（MAE）、均方根误差（RMSE）和R分别为66.59、93.73和0.71，区分具有良好解剖学反应的眼睛的受试者工作特征曲线下面积（AUC）为0.90。在外部验证集中，MAE、RMSE和R分别为68.08、97.63和0.74，AUC为0.94。对于BCVA预测，训练集中表现最佳的模型的MAE、RMSE和R分别为0.19、0.29和0.60，区分具有良好功能反应眼睛的AUC为0.80。外部验证的MAE、RMSE和R分别为0.13、0.20和0.68，AUC为0.81。
我们的集成ML系统准确预测了DME患者抗VEGF注射后的治疗后CFT和BCVA，可用于前瞻性评估DME患者抗VEGF治疗的疗效。
