Retinopathy Detection Using Probabilistic Neural Network

We propose a diabetic retinopathy (DR) analysis algorithm based on probabilistic neural network (PNN). This algorithm is used to recognize the pattern problem. By this algorithm, we can help in diagnosis of a diabetic patient regarding their damage to the back of retina (eye) occurred in tissue of blood vessels using probabilistic neural network. PNN is also known as feed forward neural network. This algorithm has been tested on a small image database and compared with the performance of a human eye. Confusion matrix and kappa coefficient are used to find the accuracy rate of the diabetic eye.