WITHDRAWN: Diabetic retinopathy detection using convolutional nueral networks algorithm

Abstract

Abstract Change in lifestyles, increase in life expectancy and several other factors contribute to the increase in number of Diabetic patients. Diabetic Retinopathy (DR) is one among the major causes of blindness in Diabetic patients. Diagnosing the Diabetic Retinopathy (DR) using color fundus images requires expertise in identifying the presence of many small features with a difficult grading system and time-consuming process for clinicians. Regularly monitoring the diabetic patients for DR is cost effective. We here proposed a CNN (Convolutional Neural Networks) based approach for diagnosing DR from fundus images and its severity. CNN, a deep learning branch, has prominent role and successful record in applications involving image analysis and interpretation. CNN network architecture can identify the features of retina involved in classifying. This network is trained with the image datasets available in Kaggle and impressive results were obtained. The timing and accuracy of this network will have significant importance to cost and effectiveness of treatment.