Background: Medical image segmentation is a well-studied subject within the field of image processing. The goal of this research is to create an AI retinal screening grading system that is both accurate and fast. We introduce a new segmentation network which achieves state-of-the-art results on semantic segmentation of colour fundus photographs. By applying the network to identify anatomical markers of diabetic retinopathy (DR) and diabetic macular edema (DME) including: micro-aneurysms, haemorrhages, etc., we collect sufficient information to classify patients into R0 (no DR) and R1 or above (DR), as well as M0 (no DME) and M1 (DME).
Methods: The AI grading system was trained on public and private screening data to evaluate the presence of DR and DME. The system’s core algorithm is a novel deep learning segmentation network (W-net) that locates and segments relevant anatomical features in a retinal image. Both eyes of the patients are graded individually, based on the detected features and classified according to the standard feature-based grading protocol used in the NHS Diabetic Eye Screening Programme.
Results: The algorithm performance was evaluated with a series of patient retinal images from routine diabetic eye screenings and achieved state-of-the-art results. It correctly predicted 98% of retinopathy events (95% confidence interval [CI], 97.1-98.8) and 68.9% of maculopathy events (95% CI, 58.1-79.7). Non-disease events prediction rate was 68.6% for retinopathy and 81.3% for maculopathy.
Conclusion: This novel deep learning segmentation model trained on a colour fundus photograph data set and tested on patient data from annual diabetic retinopathy screenings can detect and classify with high accuracy the DR and DME status of a person with diabetes. The system can be easily reconfigured according to any grading protocol, without starting a long AI training procedure. The incorporation of the AI grading system can increase the graders’ productivity and improve the final outcome of the screening process.
O. Katz: Employee; Self; NECAM, Research Support; Self; Northgate Public Services. D. Presil: Employee; Self; NECAM, Research Support; Self; Northgate Public Services. L. Cohen: Employee; Self; NECAM, Research Support; Self; Northgate Public Services. R. Nachmani: Employee; Self; NECAM, Research Support; Self; Northgate Public Services. N. Kirshner: Employee; Self; NEC, Research Support; Self; Northgate Public Service. D. R. Owens: Advisory Panel; Self; 360 Consulting. T. H. Lev: Employee; Self; NECAM (NEC Corporation of America), Research Support; Self; Northgate Public Services. A. Hadad: Consultant; Self; Necam.
