Paper—Identifying Retinal Diseases on OCT Image Based on Deep Learning Identifying Retinal Diseases on OCT Image Based on Deep Learning https://doi.org/10.3991/ijoe.v18i15.33639 Abdelhafd Errabih 1 , Mohyeddine Boussarhane 2 , Benayad Nsiri 2() , Abdelalim Sadiq 1 , My Hachem El Yousf Alaoui 2 , Rachid Oulad Haj Thami 3 , Brahim Benaji 2 1 Laboratory of Information Modelling and Communication Systems, IbnToufail University, Kenitra, Morocco 2 Health Technologies Engineering Department – Research Group in Biomedical Engineering and Pharmaceutical Sciences – Higher School of Arts and Crafts (ENSAM), Mohammed V University, Rabat, Morocco 3 ADMIR Lab, National School of Computer Science and Systems Analysis (ENSIAS), ENSAM – Mohammed V University, Rabat, Morocco benayad.nsiri@ensam.um5.ac.ma Abstract—Computer-aided diagnosis has the potential to replace or at least support medical personnel in their everyday responsibilities, such as diagnosis, therapy, and surgery. In the area of ophthalmology, artifcial intelligence approaches have been incorporated in the diagnosis of the most frequent ocu- lar disorders, namely choroidal neovascularization (CNV), diabetic macular edema (DMO), and DRUSEN; these illnesses pose a signifcant risk of vision loss. Optical coherence tomography (OCT) is an imaging technology used to diagnose the aforementioned eye disorders. It enables ophthalmologists to see the back of the eye and take various slices of the retina. The present research seeks to automate the diagnosis of retinopathy, which includes CNV, DME, and DRUSEN. The approach employed is a deep learning-based, and transfer learning technique, applying to a public dataset of OCT pictures and two per- tained neural network models VGG16 and InceptionV3, which are trained on the big database “ImageNET.” That allows them to be able to extract the main features of millions of images. Furthermore, fne-tuning approaches are applied to outperform the feature extraction method, by modifying the hyperparameters. The fndings showed that the VGG16 model performed better in classifcation than the InceptionV3 architecture, with a 0.93 accuracy. Keywords—artifcial intelligence, deep learning, convolutional neural network, transfer learning, optical coherence tomography, DRUSEN, choroidal neovascularization, diabetic macular edema iJOE ‒ Vol. 18, No. 15, 2022 141