Publicly available database of both fundus fluorescein fngiogram photographs and corresponding color fundus images of 30 healthy persons and 30 patients with diabetic retinopathy.
As manual analysis and diagnosis of large amount of images are time consuming, automatic detection and grading of diabetic retinopathy are desired. In this study, we use fundus fluorescein angiography and color fundus images simultaneously, extract 6 features employing curvelet transform, and feed them to support vector machine in order to determine diabetic retinopathy severity stages. These features are area of blood vessels, area, regularity of foveal avascular zone (FAZ), and the number of micro-aneurisms therein, total number of micro-aneurysms, and area of exudates. In order to extract exudates and vessels, we respectively modify curvelet coefficients of color fundus images and angiograms. The end points of extracted vessels in predefined region of interest based on optic disk are connected together to segment FAZ and can be further improved by adding several morphological operators. To extract micro-aneurysms from angiogram, first extracted vessels are subtracted from original image, and after removing detected background by morphological operators and enhancing bright small pixels, micro-aneurisms are detected. Our simulations show that the proposed system has sensitivity and specificity of 100% for grading diabetic retinopathy into 3 groups: 1) no diabetic retinopathy, 2) mild/moderate nonproliferative diabetic retinopathy, 3) severe nonproliferative/proliferative diabetic retinopathy.
Please reference the following paper if you would like to use any part of this dataset or method:
*** SH. Hajeb, H. Rabbani, MR. Akhlaghi, “Diabetic Retinopathy Grading by Digital Curvelet Transform", Computational and Mathematical Methods in Medicine, vol. 2012, Article ID 761901, 11 pages, 2012.1607-1614, July 2012.
Fundus Images with Exudates
This page contains 35 720*576 colour retinal images with signs of the diabetic retinopathy (microaneurysms and exudates). The data used in a study which presents a curvelet-based algorithm for detection of optic disk (OD) and exudates on low contrast images. This algorithm which is composed of three main stages does not require user initialization and is robust to the changes in the appearance of retinal fundus images. At first, bright candidate lesions in the image are extracted by employing DCUT and modification of curvelet coefficients of enhanced retinal image. For this purpose, the authors apply a new bright lesions enhancement on green plane of retinal image to obtain adequate illumination normalisation in the regions near the OD, and to increase brightness of lesions in dark areas such as fovea. Following this step, a new OD detection and boundary extraction method based on DCUT and level set method is introduced. Finally, bright lesions map (BLM) image is generated and to distinguish between exudates and OD (i.e. a false detection for the final exudates detection), the extracted candidate pixels in BLM that are not in OD regions (detected in previous step) are considered as actual bright lesions.
Please reference the following paper if you would like to use any part of this dataset or method:
***M. Esmaeili, H. Rabbani, A. M. Dehnavi, A. Dehghani, “Automatic Detection of Exudates and Optic Disk in Retinal Images Using Curvelet Transform", IET Image Processing, vol. 6, no. 7, pp. 1005-1013, Oct. 2012.