Purpose: To investigate the use of automated image analysis for the detection of diabetic retinopathy (DR) in fundus photographs captured with and without pha rmacological pupil dilation using a digital non-mydriatic camera. Methods: A to tal of 83 patients (165 eyes) with type 1 or type 2 diabetes, representing the f ull spectrum of DR, were photographed with and without pharmacological pupil dil ation using a digital non-my-driatic camera. Two sets of five overlapping, non -stereoscopic, 45-degree field images of each eye were obtained. All images we re graded in a masked fashion by two readers according to ETDRS standards and di sagreements were settled by an independent adjudicator. Automated detection of r ed lesions as well as image quality control was made: detection of a single red lesion or insufficient image quality was categorized as possible DR. Results: At patient level, the automated red lesion detection and image quality control com bined demonstrated a sensitivity of 89.9%and specificity of 85.7%in detecting DR when used on images captured without pupil dilation, and a sensitivity of 97. 0%and specificity of 75.0%when used on images captured with pupil dilation. Fo r moderate non-proliferative or more severe DR the sensitivity was 100%for ima ges captured both with and without pupil dilation. Conclusion: Our results demon strate that the described automated image analysis system, which detects the pre sence or absence of DR, can be used as a first-step screening tool in DR screen ing with considerable effectiveness.
展开▼
