This medical AI solution employs deep learning-based image segmentation to assist ophthalmologists in early glaucoma detection. The system utilizes a custom U-Net architecture trained on thousands of annotated retinal fundus images to accurately segment the optic disc and optic cup regions—critical markers for glaucoma diagnosis. The model calculates the cup-to-disc ratio (CDR) automatically, flagging high-risk cases for further clinical evaluation. Built with TensorFlow, the pipeline includes preprocessing stages for image normalization, augmentation, and quality assessment to ensure reliable predictions across varying image conditions. The system achieves high sensitivity and specificity rates, significantly reducing screening time and enabling mass screening programs in underserved areas. Integration APIs allow seamless deployment within existing hospital information systems and telemedicine platforms.