Diabetic retinopathy is a leading cause of blindness worldwide and one of the fastest-growing diabetes-related complications. Early detection is critical but often limited by access to medical specialists, particularly in low-income regions.
The challenge:
Elemental Concept was engaged to create a fast, reliable, and scalable screening solution that could accurately detect retinal damage and support clinicians in early diagnosis.
Our approach:
Using 37,266 retinal images, we trained a deep learning algorithm (80% training, 20% validation) to distinguish between healthy and diabetic retinopathy-affected eyes. The AI-based grading system was integrated into screening software that applies image-processing and recognition techniques to assess retinal damage with high precision and consistency.
Results:
The system achieved 94.3% accuracy in detecting diabetic retinopathy and 96.7% accuracy in grading its severity. This innovation enables rapid, cost-effective screening, improves diagnostic efficiency, and reduces human error — offering a transformative solution for global eye health.
