AI has shown high accuracy in detecting diabetic retinopathy (DR) in children and young adults with diabetes, highlighting the need for more research in this area.
A study conducted on 1,274 participants aged 3-26 found that 19.4% had any DR, with 2.35% classified as referable DR by AI.
The performance metrics indicated that AI had strong sensitivity and specificity rates, especially in younger children with less diabetes duration, suggesting it could be a valuable resource in low-resource settings for screening purposes.