MLPDR: High Performance ML Algorithms for the Prediction of Diabetes Retinopathy

In today's world, millions of imaging analyses are performed every week, and medical imaging plays a significant role in that. It is a process of creating visual representation based on the functioning of human organs or tissues to identify abnormalities or study diseases. Diabetes has been regarded as one of the most prevalent diseases by many researchers. Blindness or other retinal alterations based on the eye's vision, such as difficulty reading or seeing distant things, may occur, which has an impact on diabetes in humans. A new computer aided diagnosis system based on image processing of retinal images can be analyzed in order to detect diabetic patients in advance. In this study, the eye retina based on image processing techniques is analyzed and machine learning techniques have been used to extract features of eye and several computational intelligence Machine Learning techniques like Support Vector Machines (SVM) and Principal Component Analysis (PCA). These Computational Intelligence techniques have been used to detect the diabetic in advance. Several performance metrics have been considered such as Root Mean Square Error (RMSE), Peak Signal Noise Ratio (PSNR), Entropy (E), and Accuracy (A) in obtaining the better accuracy results for detection of diabetic retinopathy in early stages. The experimental results show that, SVM outperforms with an accuracy of 95%.