Detection of Diabetic Retinopathy based on Classification Algorithms Vaishnavi J, Ravi Subban, Anousouya M and Punitha Stephen, Department of Computer Science, Pondicherry University, India Abstract—Diabetic retinopathy is the retinal abnormality for the diabetic patient due to imbalanced blood glucose level. To help the ophthalmologist in diagnosing the diabetic retinopathy using retinal images taken from various publicly available databases, efficient classifiers are used to analyze the images and to identify the symptoms of diabetic retinopathy such as microanueryms, hemorrhages, exudates, cotton wool spots etc. The manual screening method of diabetic retinopathy is time consuming and it is difficult for the ophthalmologist in producing accurate results. Thus an automatic computerized system is enabled to detect the diabetic retinopathy with less time consumption and obtain effective results. In this paper, a brief survey is presented on the various classification methods used in the detection of diabetic retinopathy. The various classifiers are efficient in classifying diabetic retinopathy as proliferative and non- proliferative diabetic retinopathy. The machine learning classifiers such as Naïve Bayes, Gaussian, random forest, support vector machine and neural network achieves the greater accuracy of more than 90%. Using these algorithms severity grading of the abnormalities is also improved. Keywords-Bayesian, Gaussian model, Random forest, Hybrid fuzzy model,Support vector machine; neural network Introduction I. INTRODUCTION Diabetic mellitus is the imbalanced level of glucose in the blood. A patient with long term of diabetes leads to diabetic retinopathy. It can be proliferative and non-proliferative. The different stages of diabetic retinopathy is mild non- proliferative, moderate non-proliferative, severe non- proliferative and proliferative. In the beginning stage of non- proliferative diabetic retinopathy the blood vessels becomes thin and leads to blood leakage which is called as microanueryms, the extension of microanuerysms leads to hemorrhages. In the stage of proliferative diabetic retinopathy, some new abnormal blood vessels begin to grow like a loop and network like structure. In the early diagnosing system it was difficult to analyze the retinal images, thus the proposed automatic assessment is capable of diagnosing the retinal images to reduce the complications of the ophthalmologists. The retinal images are analyzed to remove background noises, enhancing the contrast, illumination adjustments by using filters. The region to be diagnosed is done using the pattern matching, thresholding, region growing methodologies. The extension of preprocessing stage is segmentation which segments out the Figure 1.Abnormal Retinal Image abnormalities and normal regions using several segmentation methodologies like Support Vector Machine, Neural network, clustering etc. Finally the segmented images are classified using classifiers such as naïve Bayes, Gaussian method, and Random Forest, SVM and Neural Network Classification is to identify the set of categories to which the observation proceeds. Classification can be done statistically as well as by machine learning. The statistical classification can be performed using logistic regression. In machine learning algorithm, the observations are called as instances and the variables are called features. The organization of the paper is as follows: In section 2.the survey of various classifications techniques has been discussed. The detailed description of the table is discussed in section 3.Finally the proposed work is concluded in section 4 II. SURVEY USING CLASSIFICATION ALGORITHMS A.NAÏVE BAYES CLASSIFIER Istvan Lazar et al. [1] proposed retinal microanueryms detection through local rotating cross section profile analysis. This method proposes candidate extraction, feature extraction, classification and score determination. Peak detection on each pixel is based on the size, height, and shape of the peak. Features of the cross section profiles are classified using Naive Bayes classification. Retinopathy Online Challenge (ROC) is proposed which scores high comparing the other methods. The classifier naïve Bayes is used to exclude the candidate and to estimate the feature probability distribution. The extracted features are size, height and shape. Cemal Köse et al. [2] proposed a simple method for segmentation and measurement Vol. 14 CIC 2016 Special Issue International Journal of Computer Science and Information Security (IJCSIS) https://sites.google.com/site/ijcsis/ ISSN 1947-5500 International Conference on Advances in Computational Intelligence and Communication (CIC 2016) Pondicherry Engineering College, Puducherry, India October 19 & 20 - 2016 75