Published Paper Details:

Glaucoma is a chronic and progressive eye disease characterized by optic nerve damage and gradual loss of vision. Early detection and timely intervention are critical to preventing irreversible vision impairment. In recent years, advancements in machine learning, particularly Convolutional Neural Networks (CNNs), have shown promise in automated glaucoma detection through the analysis of retinal images. This study proposes a robust glaucoma detection framework that integrates image enhancement techniques like Retinex, state-of-the-art CNN architectures, and statistical methodologies such as Design of Experiments (DOE). The Retinex enhancement process is employed to improve the quality and contrast of fundus images, enhancing the subsequent CNN-based feature extraction. CNNs, designed to identify distinctive patterns and features indicative of glaucoma, are optimized using DOE to achieve enhanced detection performance. The proposed framework exhibits a high degree of accuracy, sensitivity, and specificity, making it a potential candidate for effective glaucoma screening and diagnosis. The integration of image enhancement, deep learning, and statistical optimization in this framework underscores the significant potential for advancements in glaucoma detection, aiding ophthalmologists in early diagnosis and appropriate patient management.