Automatic Detection of Diabetic Retinopathy from Retinal Fundus Images Using MobileNet Model

Diabetes can lead to a number of eye disorders, such as glaucoma, cataracts, and diabetic retinopathy, which can be defined as damage to the retinal blood vessels of the eye. If these damaged vessels remain untreated, this may result in decreased flow of blood, irritation, and eventually irreversible vision loss. We can stop vision loss and prevent the development of diabetic retinopathy with timely treatment. The Computer-Aided Detection technique assists ophthalmologists in the detection of ophthalmologic diseases by automatically analyzing retinal fundus images. Deep learning algorithms have performed exceptionally well in various computer vision applications, giving them a distinct advantage over traditional approaches. Therefore, continuous discussion and evaluation of the connected techniques are necessary. In this paper, the pretrained convolutional neural network, MobileNet is used to detect DR in retinal fundus images. The Asia Pacific Tele-Ophthalmology Society 2019 Blindness Detection (APTOS-2019 BD) dataset, which was collected by India’s Arvind Hospital, was utilized in this study. To assess the efficacy of the model, the APTOS-2019 BD dataset is split into two sections, 80% of which are used for training the proposed model and 20% for testing it to assess the model’s efficacy. The effectiveness of the suggested technique is evaluated based on a number of factors, including accuracy, precision, sensitivity, and specificity.