Abstract
Purpose :
Predicting diabetic retinopathy (DR) progression could enable individualised screening with prompt referral for high-risk individuals for sight-saving treatment, whilst reducing screening burden for low-risk individuals. We validated trained deep learning systems (DLS) that predict 1, 2 and 3 year emergent referable DR and maculopathy using risk factor characteristics (tabular DLS), colour fundal photographs (image DLS) or both (multimodal DLS).
Methods :
DLS were trained using data from 162,339 development set eyes from the south-east London diabetic eye screening programme (DESP, UK, Figure 1A), of which 110,837 eyes had eligible longitudinal data, and the remaining 51,502 eyes were used for DLS pretraining. Internal and external (Birmingham DESP, UK, Figure 1B) validation datasets included 27,996, and 6,928 eyes respectively. DLS outcomes were predicting emergent (1) referable DR (R2+), (2) referable maculopathy (M1) or (3) either (R2+ | M1) within 1, 2 or 3 years.
Results :
Internal validation multimodal DLS emergent referable DR, maculopathy or either area-under-the receiver operating characteristic (AUROC) were 0.95 (95% CI: 0.92-0.98), 0.84 (0.82-0.86), 0.85 (0.83-0.87) for 1 year, 0.92 (0.87-0.96), 0.84 (0.82-0.87), 0.85 (0.82-0.87) for 2 years, and 0.85 (0.80-0.90), 0.79 (0.76-0.82), 0.79 (0.76-0.82) for 3 years. External validation multimodal DLS emergent referable DR, maculopathy or either AUROC were 0.93 (0.88-0.97), 0.85 (0.80-0.89), 0.85 (0.76-0.85) for 1 year, 0.93 (0.89-0.97), 0.79 (0.74-0.84), 0.80 (0.76-0.85) for 2 years, and 0.91 (0.84-0.98), 0.79 (0.74-0.83), 0.79 (0.74-0.84) for 3 years. Multimodal and image DLS performance was significantly better than tabular DLS for all emergent referable disease outcomes and prediction intervals. Multimodal DLS was significantly better than image DLS prediction of emergent referable maculopathy within 2 and 3 years on internal validation.
Conclusions :
DLS accurately predict 1, 2 and 3 year emergent referable DR and referable maculopathy using fundal photographs, with additional risk factor characteristics conferring further improvements in prognostic performance. Proposed DLS are a step towards individualised risk-based screening, whereby AI-assistance allows high-risk individuals to be closely monitored while reducing screening burden for low-risk individuals.
This abstract was presented at the 2024 ARVO Annual Meeting, held in Seattle, WA, May 5-9, 2024.