Deep Learning-Driven Analysis of Experimental Clinical Data for Automated Glaucoma Detection: A Multicentric Supervised Study

The integration of Artificial Intelligence (AI) into clinical ophthalmology has revolutionized diagnostic accuracy by enabling the detection of subtle patterns in high-dimensional datasets. This paper presents a supervised machine learning framework trained on longitudinal experimental data (n = 1, 200), incorporating Intraocular Pressure (IOP), Retinal Nerve Fiber Layer (RNFL) thickness, and Cup-to-Disc Ratio (CDR). Utilizing a novel hybrid ensemble architecture that combines Gradient Boosting Machines (XGBoost) for feature ranking and Deep Neural Networks (DNN) for non-linear classification, we achieved unprecedented performance metrics. Experimental results demonstrate an Area Under the Curve (AUC) of 0.965, with a specificity of 0.962. This study builds upon the author's previous expertise in medical AI applications, including Parkinson's and Alzheimer's detection, offering a scalable solution for early-stage glaucoma screening in multicentric environments.