Detection of Retinal Disease from Optical Coherence Tomography Images Using CNN Models

Optical Coherence Tomography (OCT) is a scanning method to diagnose tissues in a non-invasive way. Even the micro-architecture of the retina can be seen clearly in the OCT image, which makes it manageable to screen for any unusual patterns. This paper aims to classify retinal state into four classes: Diabetic Macular Edema (DME), DRUSEN, Neovascularization (CNV), and NORMAL by using five different CNN architectures, such as VGG16, VGG19, ResNet50, MobileNet, and our proposed LayerV model. The LayerV model, which technically improved the algorithm's parameter and has an accuracy of 97.45%, is an updated version of VGG16 that is suited for multimodal images taken using various illness configurations under various conditions.