An automated retinal imaging method for the early diagnosis of diabetic retinopathy.

BACKGROUND: Diabetic retinopathy is a microvascular complication of long-term diabetes and is the major cause for eyesight loss due to changes in blood vessels of the retina. Major vision loss due to diabetic retinopathy is highly preventable with regular screening and timely intervention at the earlier stages. Retinal blood vessel segmentation methods help to identify the successive stages of such sight threatening diseases like diabetes.

OBJECTIVE: To develop and test a novel retinal imaging method which segments the blood vessels automatically from retinal images, which helps the ophthalmologists in the diagnosis and follow-up of diabetic retinopathy.

METHODS: This method segments each image pixel as vessel or nonvessel, which in turn, used for automatic recognition of the vasculature in retinal images. Retinal blood vessels were identified by means of a multilayer perceptron neural network, for which the inputs were derived from the Gabor and moment invariants-based features. Back propagation algorithm, which provides an efficient technique to change the weights in a feed forward network, is utilized in our method.

RESULTS: Quantitative results of sensitivity, specificity and predictive values were obtained in our method and the measured accuracy of our segmentation algorithm was 95.3%, which is better than that presented by state-of-the-art approaches.

CONCLUSIONS: The evaluation procedure used and the demonstrated effectiveness of our automated retinal imaging method proves itself as the most powerful tool to diagnose diabetic retinopathy in the earlier stages.