Convolution neural network and deep-belief network (DBN) based automatic detection and diagnosis of Glaucoma

Diagnosis of Glaucoma eye disease is a challenging task for CADx (computer-aided diagnostics) systems. An automatic CADx framework is developed for diagnosing glaucoma eye disease by handcrafted feature-based segmentation in retinal images. In this manuscript, automatic glaucoma eye disease detection based on deep learning (DL), with deep-belief network (DBN) is proposed. In addition, a contextualizing DL structure is proposed for obtaining various levels of portraying fundus images to separate among glaucoma and non-glaucoma modes, where the network uses output of other CNNs as information of context to support performance. The3 existing machine learning models are (1) SVM (support vector machine) (2) RF (random forest)(3) k-NN (k-nearest neighbor), which is executed and assessed on tests. The efficiency of the Glaucoma-Deep model is analyzed by the statistical measures like sensitivity, specificity, accuracy, precision. Finally, an official conclusion executed through the softmax straight classifier is to divide glaucoma and non-glaucoma retinal fundus images.