Convolutional Neural Networks for Automated Glaucoma Detection: Performance and Limitations

Glaucoma is a set of eye disorders that, if left untreated, can cause optic nerve damage, resulting in vision loss and blindness. While glaucoma is often linked with high eye stress, it can also develop with normal or low pressure. The most common variety, primary open-angle glaucoma, is known as the silent thief of sight because it causes slow vision loss with no symptoms. Ethnicity, age, diabetes, bloodline, and hypertension are all dangerous factors. Regular eye exams are crucial for early detection. To aid in glaucoma detection, a model utilizing eye fundus images is proposed. Fundus images provide valuable information about the optic nerve’s health and abnormalities. The model employs a Convolutional Neural Network (CNN) to classify fundus images and detect glaucoma. By automating the process, the proposed system aims to improve accuracy. This CNN-based model has the potential to enhance glaucoma detection, enabling prompt interventions and better patient outcomes.