NEURAL NETWORKS ALGORITHM BASED DIABETIC RETINOPATHY DIAGNOSIS SYSTEM

Abstract

 Due to diabetes the blood vessels are damaged in retina of human eyes. The damage of blood vessels is the important reason for blindness.  Diabetic retinopathy means the blood cells are damaged in retina. The major symptom of this disease is the vision is unclear and complexity to differentiating colours. The final state of this disease is the affected peoples loss their entire vision.  The diabetes peoples checked their vision periodically. The communication technology plays a major role in various fields. In medical field also the computer techniques and algorithms are used to predict and diagnosis the diseases.    Diabetic retinopathy is predicted in earlier stage, to prevent the affected person from visual impairment or blindness. This paper describes about various machine learning algorithms are used to predict the Diabetic retinopathy disease in earlier stage.  In this proposed work k-NN, ANN and SVM classifiers are used to predict the diseases. SVM algorithm is used extract the features from unstructured data.