Comparative Study and Detection of Diabetic Retinopathy in Retinal Images Using Computational Approach

Retinal image segmentation and classification is a challenge task in diagnosing and treating for Diabetic Retinopathy (DR) over the past decade. Usually, retinal image is used to assess the diabetic diseases, as it offers complementary information for acquiring the retinal image sequences. This long outstanding problem to classify the DR significantly requires more time for a physician. Therefore, developed an automated computational approach for physicians with less time and speed up the diagnosing procedure. The proposed work based on machine learning techniques for achieving blood vessel classification using the optic disc segmented features of retinal image. The segments are generated through the image processing mechanism, which ensure the effectiveness of optimal segment selection that yields to detect the optic disc and blood vessel more accurately. In, this proposed work detailed comparative study for image processing and machine learning techniques in DR are analyzed. Finally, the effectiveness of the proposed work is carried out by using this various machine learning algorithm and attained the better performance value. The proposed work achieves the best results values for blood vessel classification in DR and computed the performance metrics in terms of accuracy, sensitivity and specificity respectively.