Diabetic retinopathy is an eye condition that can cause vision loss and blindness in people who have diabetes and is one of the fastest growing global serious health conditions.
- Elemental Concept developed a diabetic eye screening system which using an artificial intelligence-based grading algorithm to assess the retinal damage with high reliability and accuracy.
- Using 37,266 retinal images, of which 80% were used to train a Deep Learning algorithm and 20% used to validate its performance. We created a screening system that successfully identifies diabetes retinopathy patients from healthy patients with incredible accuracy:
- Accuracy to distinguish Normal/Diabetic Retinopathy: 94.3% | Non-Proliferative/Proliferative: 96.7%
Potential Impact
Rapid screening of diabetes retinopathy is valuable, especially in low-income countries, challenged by limited access to medical specialists. Our solution saw us apply AI to image processing and facial recognition applications to develop a screening software that could assist medical experts, improve their efficiency and minimise human error across the health sector. The project showcased our ability to design and develop AI applications to analyse and extract meaningful insights from images.