Diabetic State Recognition and Detection Based on Intelligent Methods

One of the main problems in medical science is the absence of recognition, detection and prediction of differentillness for working precisely and practical. Diabetes is one of the illnesses which caught today’s people due to disregardenvironmental issues. Diabetes has several states which one of them is retinopathy that losing eyesight is one the mostimportant risk of it. Microaneurysm are red spots which are the primary symptoms occurred in retina. The mostimportant problem is early detection of retinopathy for protecting eyesight and consequences of this disease for sometime delay. This research proposed a new approach for diabetic retinopathy recognition and detection by using hybridmethods. This approach has four levels. In first level, pre-processing done for probabilistic noise removal andstandardization of input dataset. Then Spiking Neural Network (SNN) done for image segmentation based on edgedetection. In the following step, dimension reduction, feature selection and extraction done by percolation theory whichfeatures are blood vessels edges and intensity of edges. At the end, combination method of SNN and percolation theoryhas done for detection the area of retinopathy. Results show that proposed method has the better accuracy in comparisonto others.

کلیدواژه ها:

Retinopathy ، Image Segmentation ، Edge Detection ، Spiking Neural Network (SNN) ، Percolation Theory