PROPOSED ARCHITECTURE FOR DIABETIC RETINOPATHY DISEASE OF BLINDNESS USING DEEP LEARNING AND IMAGE PROCESSING

Abstract

Diabetic Retinopathy (DR) is one of the major causes of blindness in the western world. Increasing life expectancy, indulgent lifestyles and other contributing factors mean the number of people with diabetes is projected to continue rising. Regular screening of diabetic patients for DR has been shown to be a cost-effective and important aspect of their care. The accuracy and timing of this care is of significant importance to both the cost and effectiveness of treatment. If detected early enough, effective treatment of DR is available, making this a vital process. The diagnosis of diabetic retinopathy (DR) through colour fundus images requires experienced clinicians to identify the presence and significance of many small features which, along with a complex grading system, makes this a difficult and time consuming task. As it is a leading disease in the west, there needs to be an easy solution and way to identify the disease. We, through this app are trying to do the same and help in some real world problem of the society. We propose a CNN approach to diagnosing DR from digital fundus images and accurately classifying its severity. We develop a network with CNN architecture and data augmentations which can identify the intricate features involved in the classification task such as micro-aneurysms, exudates and hemorrhages on the retina and consequently provide a diagnosis automatically and without user input. We train this network using a high-end graphics processor unit (GPU) on the publicly available Kaggle dataset and demonstrate impressive results, particularly for a high-level classification task. On the data set of 80,000 images used our proposed CNN achieves a sensitivity of 95% and an accuracy of 75% on 5,000 validation images. What should be no surprise in an image recognition task, most of the top contestants used deep convolution neural networks (CNNs), and so did we. Our solution consisted of multiple steps: