Automated identification of diabetic retinal exudates in digital colour images

Abstract

Aim: To identify retinal exudates automatically from colour retinal images. Methods: The colour retinal images were segmented using fuzzy C-means clustering following some key preprocessing steps. To classify the segmented regions into exudates and non-exudates, an artificial neural network classifier was investigated.Results: The proposed system can achieve a diagnostic accuracy with 95.0% sensitivity and 88.9% specificity for the identification of images containing any evidence of retinopathy, where the trade off between sensitivity and specificity was appropriately balanced for this particular problem. Furthermore, it demonstrates 93.0% sensitivity and 94.1% specificity in terms of exudate based classification. Conclusions: This study indicates that automated evaluation of digital retinal images could be used to screen for exudative diabetic retinopathy. Intraretinal fatty (hard) exudates are a visible sign of diabetic retinopathy and a marker for the presence of coexistent retinal oedema. If present in the macular area, they are a major cause of treatable visual loss in the nonproliferative forms of diabetic retinopathy. It would be useful to have an automated method of detecting exudates in digital retinal images produced from diabetic retinopathy screening programmes.Sinthanayothin 1 identified exudates in grey level images based on a recursive region growing technique. The sensitivity and specificity reported were 88.5% and 99.7%; however, these measurements were based on 10610 windows where each window was considered as an exudate or a non-exudate region. The reported sensitivity and specificity only represent an approximate accuracy of exudate recognition, because any particular 10610 window may be only partially affected by exudates. Gardner et al 2 used a neural network (NN) to identify the exudates in grey level images. The authors reported a sensitivity of 93.1%. Again this was the result of classifying whole 20620 regions rather than a pixel level classification. One novelty of our proposed method here is that we locate exudates at pixel resolution rather than estimate for regions. We evaluate the performance of our system applying both lesion based and image based criteria in colour retinal images. MATERIALS AND METHODSWe used 142 colour retinal images obtained from a Canon CR6-45 non-mydriatic retinal camera with a 45˚field of view as our initial image dataset. This consisted of 75 images for training and testing our NN classifier in the exudate based classification stage. The remaining 67 colour images were employed to investigate the diagnostic accuracy of our system for identification of images containing any evidence of retinopathy. The image resolution was 7606570 at 24 bit RGB.Preprocessing Typically, there is wide variation in the colour of the fundus from different patients, related to race and iris colour. The first step is therefore to normalise the retinal images across the set. We selected a particular retinal image as a reference and then used histogram specification 3 to modify the ...

Automated Detection of Hard Exudates in Fundus Images Using Improved OTSU Thresholding and SVM

“…We found that EXs appear most contrasted in the green channel, and optic disc appears most continuous and most contrasted against the background in the red channel [19]. So we coarsely segmented EXs in the green channel as shown in Figure 2(a), and segmented optic disc in the red channel which was EXs Optic disc shown in Figure 2(b).…”

To Find Location of Optic Disc in Digital Fundus Images

“…The four main vessels originating from the OD were geometrically modeled using two parabolas, and the OD position was located as their common vertex. Inspired by previous works, Youssif et al [6]. The OD center location was identified using the variance of intensity produced by the blood vessels within the OD.…”

Automated Segmentation of Optical Nerves by Neural Network based Region Growing

“…In a clustering technique along with the region growing, each pixel is associated with one of the finite number of threshold is grown to form disjoint regions. The contextual clustering method proposed by [4,9] is a supervised algorithm. It uses a 3 X 3 overlapping windows of pixels to form a segmented image.…”