Automatic Detection of Microaneurysm using Local Rotating Cross Sectional Profile Analysis

Diabetes is a metabolic disease which damage blood vessel. Damage in retinal blood vessel of the human eye due to effect of diabetes is termed as Diabetic Retinopathy (DR). This abnormality in the blood vessel will cause a leakage of blood from the vessel and forms a red dot. This is named as Microaneurysm. Microaneurysm in the retina causes a swelling of retinal tissue which in turn causes clouding in vision. Microaneurysm is the first symptom of Diabetic Retinopathy. Detection of microaneurysm is mainly used for the early diagnosis and treatment of DR in order to prevent the vision loss. The proposed method is used to detect microaneurysms in colour fundus images using local rotating cross-section profile analysis. Based on the profile parameters such as peak value, height and width, microaneurysms are identified. The method is successfully applied to the publically available MESSIDOR database fundus images and the results related with the detection of microaneurysms are obtained with high accuracy.