Analysis of maculopathy in color fundus images

Diabetes has become a life threatening disease. The effect of diabetes on human retina is called Diabetic Retinopathy (DR). If DR involves macula, it is known as maculopathy. Under this condition, the visibility of the patient gets affected adversely. In this paper, we have proposed a method for grading the stages of maculopathy. In the first stage, the abnormalities are detected. Second stage consists of detection of fovea. In the third stage, different macular masks are generated and the amount of abnormality present over them is calculated. Depending on the amount of abnormality, the severity level is decided. The method is applied on DRIVE, DIARETDB1 and HRF databases which have both normal and pathological images. The proposed method shows an overall sensitivity of 97.5% and specificity of 98.7% in grading maculopathy.