Diagnostic AI Systems
Three deep learning models for automated retinal disease detection, designed for integration into clinical and community screening workflows.
Detects and grades AMD severity from fundus photographs. Identifies drusen, pigmentary changes, geographic atrophy, and choroidal neovascularisation.
Five-level ETDRS-aligned severity grading. Detects microaneurysms, haemorrhages, exudates, venous beading, and neovascularisation at disc and elsewhere.
Automated cup-to-disc ratio measurement from disc-centred fundus images. Provides continuous CDR estimate with clinical threshold interpretation.