Neural Network as an Ophthalmologic Disease Classifier

In this paper, we explore the neural network as a disease classifier. In our investigation, the sets of parameters describing glaucomatous and healthy eyes are taken. These sets represent the structure of the optical nerve disc which resides in a patient’s eye fundus image. As a separate case, the excavation can be seen in the image as well. These two sets describe the elliptical shape of both structures and compound the initial data for analysis. Thus, the distinction of classes represented by the data sets becomes possible. In this article, a multi-layer neural network is explored. Selection of the optimal number of hidden neurons is taken into consideration. We also explore here the principal component analysis for feature reduction. The classification results are discussed as well.