Diabetic Retinopathy Image Classification Using Support Vector Machine
Manoj Kumar Behera, Sujata Chakravarty
Abstract
Diabetes is a leading cause of various serious diseases like kidney failure, cardiovascular problems and many more, blindness is one of them. As it progress it may affect the retina of the eye and leads to Diabetic retinopathy (DR). In this disease the retinal blood vessels start swelling and eventually retina gets damaged. This generally makes the patient total blind. The effective way of handle this disease is to detect and cure it at its earlier state. This needs regular screenings of the eye. There are several automatic screening techniques for the determination of the disease still there is a scope for the improvement in accuracy of prediction. One of the way to improve the accuracy is by considering important artifacts present in the retina; those play a vital role in disease prediction. In case of DR one important artifact is Exudates. Therefore, in this research two well-known predefined feature extraction techniques scale-invariant feature transform (SIFT) and speeded up robust features (SURF) have been used simultaneously on each retinal images to capture the Exudates regions. These Exudates of each image stored in a feature matrix and used by the support vector machine(SVM) classifier for prediction of DR. For a 100 set of test images the average sensitivity of the model is 94%.