Abstract
Abstract: :
Purpose: Diabetic retinopathy is a leading cause of vision loss in developed countries. Screening programs can identify the disease at an earlier and more treatable stage. Retinal digital imaging is becoming available as a means of screening for diabetic retinopathy. Using images of the ocular fundus, we are developing algorithms for the detection and classifications of various lesions associated with diabetic retinopathy. Moreover digital images have the potential to be processed by automatic analysis systems. Methods: A program for detecting and quantifying diabetic retinopathy is proposed. The program performs an image enhancement of the digital fundus photographs. This enhancement is carried out by subtracting background illumination of the photographs and applying of Frei and Chen operator (an edge-finding operator) to the result. In this way, the system automatically detects and discriminates between hard exudates, cotton wools spots and haemorrhages and provides the number of the different lesions and their location in the ocular fundus. Results: A total of 59 digital fundus photographs were examined of whom 7 (12%) present hard exudates, 4 (7%) cotton wools spots, 2 (3%) haemorrhages and 46 (78%) the three lesions all together. In all of them, the visual discrimination and automatic detection of hard exudates, cotton wools spots and haemorrhages were successful. However, it can also be seen that a number of the false negative cases arise when the different lesions were very close together. Conclusions: The long term goal of the project is to automate the screening for diabetic retinopathy with retinal images. We have described the preliminary development of a tool to provide automatic analysis of digital fundus photographs taken as part of routine monitoring of diabetic retinopathy. We have proved its efficiency for detecting and discriminating between various lesions associated with diabetic retinopathy: hard exudates, cotton wools spot and haemorrhages.
Keywords: diabetic retinopathy • image processing • imaging/image analysis: clinical