This article is more than 5 years old. Information may no longer be current.
Digital imaging can aid in screening for DR
Automated grading of digital retinal images can improve diabetic retinopathy screening, according to a recent study in Scotland.
John A. Olson, MD, and colleagues at the University of Aberdeen compared several mydriatic screening methods for evaluating diabetic retinopathy status in 586 patients with diabetes.
Slit-lamp screening by optometrists achieved the lowest sensitivities (73%) but had no technical failures. Manual grading of red-free digital images showed an average sensitivity of 93% and an average specificity of 87%. Color slides showed an average sensitivity of 96% and an average specificity of 89%. Technical failure rates were lower with digital imaging — 4.4% compared to 11.9% with color slides.
Automated grading showed an average sensitivity of 83% and an average specificity of 71%. It also detected diabetic macular edema with an average sensitivity of 76% and an average specificity of 85%.
The study is published in Diabetic Medicine.