An Efficient Model for Detection and Classification of Internal Eye Diseases using Deep Learning

Abstract

Natural eye is influenced by the distinctive eye illnesses some of them are great cause of vision loss. Many Artificial Intelligence (AI) approaches have been proposed for the identification of such diseases. The proposed method intends to plan an AI based automated network for eye illness identification and grouping to help the ophthalmologists all the more viably distinguishing and ordering of internal eye diseases like Choroid Neovascularisation (CNV), Diabetic Macular Edema (DME) and Drusen by utilizing the Optical Coherence Tomography (OCT) pictures portraying various tissues. The procedure utilized for planning this framework includes diverse deep learning convolutional neural organization (CNN) models. The proposed methodology is called efficient because it is performed on a large scale data-set which has four classes and improves the performance to a great level. The best picture subtitling model is chosen after execution investigation by looking at different picture inscribing frameworks for helping ophthalmologists to identify and order eye illnesses. The proposed methodology achieves the performance to a great level, 83.66% of accuracy for the test images when the data-set is divide in the format of 70-30 ratio.