Table 1.

Application of artificial intelligence to retinal diseases.

Disease	AI Devices	AI Tool
DR	IDx-DR, EyeArt, and AEYE-DS	multiple biomarker detectors, utilizing CNN; feature-based C approach to detect microaneurysms from color fundus photos
AMD	U-net DL segmenter, DL framework, ResNet, iPredict, DenseNet	identify and distinguish early AMD OCT biomarkers; address treatment choice
RVO	SVM, ResNet-50, Inception-v3, DenseNet-121, and SE-ReNeXt-50	identify nonperfused areas and detect foveal avascular zone using CNN and DNN
ROP	SVM, i-ROP, DeepROP	analyze traditional features and identify optimal combinations for plus disease diagnosis, using CNN

DR = diabetic retinopathy; CNN = convolutional neural networks; ML = machine learning; AMD = age-related macular degeneration; DL = deep learning; RVO = retinal vein occlusion; SVM = support vector machine; DNN = deep convolutional neural network; ROP = retinopathy of prematurity.