A Novel System For Glaucoma Diagnosis Using Artificial Neural Network Classification

摘要

Morphological shape of the optic nerve's disc and excavation presents an important feature in the identification of eyes' diseases such as glaucoma. Leading to the optic nerve head (ONH) destruction, this sickness is considered as the second leading cause of blindness worldwide and mainly in least developed countries. Since early detection is crucial to cure this disease, this paper describes a new decision-making system based on Artificial Neural Network (ANN) classifier. The suggested method has the advantage of taking into consideration both instrumental parameters (Cup-to-Disc Ratio, ISNT rule and eyes' asymmetry) and factor risks (age, gender, genetic history and origin). Experiments are performed on a real dataset of ophthalmologic images of normal and glaucomatous cases. The experimental results show high accuracy compared with some existing systems.