Early detection of optic-nerve damage remains a major unmet need in community eye-care settings, where access to high-resolution retinal imaging is limited and interpretation of fundus photographs varies widely between clinicians. HawkVi is a cloud-based intelligent analytical platform that delivers automated, quantitative evaluation of the optic nerve directly from standard fundus images. At its core, the system integrates an nnU-Net–driven segmentation pipeline that precisely delineates the optic disc and optic cup, enabling robust extraction of clinically meaningful biomarkers such as the vertical optic disc–to–cup ratio, the structure of the neuro-retinal rim, patterns related to the retinal nerve fibre layer, and subtle indicators of early temporal pallor. The adaptive, self-configuring architecture of nnU-Net ensures high segmentation fidelity across heterogeneous datasets and multiple imaging systems (Topcon, Canon, NIDEK, Remidio).

The segmentation output is coupled with a deep-learning classification module trained to detect early signs of optic-nerve injury, supporting screening for glaucoma, optic neuritis, and neuro-degenerative disorders. Validation across several clinical centres demonstrated stable and consistent performance across demographic groups, imaging environments, and camera types. With processing times of less than two minutes per image, HawkVi enables real-time, cloud-based decision support suitable for remote eye-care workflows, community screening, and high-volume clinical settings.

This work highlights the transformative potential of intelligent analytical tools, combining standardized nnU-Net segmentation with cloud-deployed inference to provide accessible, scalable, and objective optic-nerve evaluation for routine clinical practice and population-level eye-health programs.