IDENTIFICATION OF GLAUCOMA FROM FUNDUS IMAGES AND USING DEEP LEARNING TECHNIQUES
Sandeep Singh, Nidhi Srivastav, Achin Srivastav
1,2Department of Computer Science and Engineering, Swami Keshvanand Institute of Technology, Management & Gramothan, Jaipur, India
3Department of Mechanical Engineering, Swami Keshvanand Institute of Technology, Management & Gramothan, Jaipur, India
Abstract: The Retinopathy, often known as diabetic retinopathy (DR), is a common consequence of diabetes that can have an impact on a person’s vision. If it is not identified at an early stage, it can result in blindness. Unfortunately, diabetic retinopathy is not a process that can be reversed, and treatment can only keep vision intact. Detection and treatment of diabetic retinopathy (DR) at an early stage can dramatically reduce the vision loss risk. The manual diagnostic procedure of the DR retinal fundus photographs by ophthalmologists is time consuming, difficult, and expensive. Additionally, it is possible to make an incorrect diagnosis. In recent years, deep learning has emerged as one of the most popular methods, and it has demonstrated improved performance in a variety of domains, particularly in the field of medical picture analysis and classification. Proposed Model uses DenseNet network for the purpose to classify and diagnosis of the diabetic retinopathy. These networks are widely utilized deep learning method in the field of medical image analysis, and they are extremely efficient. Performance of proposed model is evaluated by using Accuracy in training, learning rate, training loss, validation loss, validation accuracy, precision, recall, and sensitivity.
Keywords: Machine Learning, Deep Learning, Glaucoma, optical coherence tomography, CNN, DR, Dense Net121