Adaptive Super-Candidate Based Approach for Detection and Classification of Drusen on Retinal Fundus Images

Abstract

Abstract. Identification and characterization of drusen is essential for the severity assessment of age-related macular degeneration (AMD). Presented here is a novel super-candidate based approach, combined with robust preprocessing and adaptive thresholding for detection of drusen, resulting in accurate segmentation with the mean lesion-level overlap of 0.75, even in cases with non-uniform illumination, poor contrast and confounding anatomical structures. We also present a feature based lesionlevel discrimination analysis between hard and soft drusen. Our method gives sensitivity of 80% for high specificity above 90% and high sensitivity of 95% for specificity of 70% on representative pathological databases (STARE and ARIA) for both detection and discrimination.

Automatic differentiation of color fundus images containing drusen or exudates using a contextual spatial pyramid approach

“…In literature, other methods which have focused on the identification of either drusen or exudates have been described [10][11][12][13][15][16][17]. In these works, the focus was put on discriminating control cases from abnormal cases, i.e.…”

“…Automatic software solutions have been proposed to allow for more costeffective mass-screening, reducing the amount of specialized personnel required and making mass-screening feasible. Most of these automatic software solutions analyze CF images for presence of lesions which are associated with DR or AMD [9][10][11][12][13][14][15][16][17]. Lesions associated with DR include microaneurysms, hemorrhages, exudates and cotton wool spots, whereas for AMD, these include drusen.…”

Automated method for the detection and segmentation of drusen in colour fundus image for the diagnosis of age‐related macular degeneration

“…Sundaresan et al [26] proposed a super candidate based approach for the detection and discrimination of soft and hard drusen. Retinal image was first pre-processed with sigmoid function and histogram equalisation.…”

Automatic Screening and Grading of Age-Related Macular Degeneration from Texture Analysis of Fundus Images

“…Other methods first detect drusen regions and a classification based on drusen features, using, for example, linear discriminant analysis, k -nearest neighbors, gentle boost, random forest, or support vector machine classifiers, is then performed for AMD screening or assessing the risk of progression to the advanced stage [ 24 – 26 ]. The results show good performance, comparable to trained human observers.…”