Exudate Segmentation in Retinal Images of Diabetic Retinopathy Using Saliency Method Based on Region

Diabetic retinopathy is a complication of diabetes that attacks the eye organs. One of the early signs of the disease is the appearance of an exudate wound that occurs because there is lipid in the abnormal blood vessels which can cause blindness if it appears around the macula. Early detection of exudates is expected to reduce the risk of blindness to diabetic retinopathy patients. One of the problems in the detection process of exudates is that the size of the object is quite small compared to the overall image. In this research, we propose the segmentation of exudates with saliency method based on region obtained using intensity thresholding. There are three main stages in this research, namely optic disc removal, location detection of exudates, and exudates segmentation. Optical disc removal is done using midpoint circle algorithm. At the detection stage of the exudates location, the image will be subdivided into smaller sub-images called patches then classified into exudate patch and exudate-free patch based on the intensity thresholding obtained from each image. For a sub-image classified as exudate patch, then it is segmented using saliency method. The evaluation method is done on the diaretDB1 dataset by calculating the value of accuracy, sensitivity, and specificity of the proposed method respectively 99.33%, 81.65% and 99.42%. The proposed method is expected to be able to detect exudates well.