Automatic Grading of Severity of Diabetic Macular Edema Using Color Fundus Images

Diabetic macular edema (DME) is the main reason for vision loss in diabetic patients and results in blurring in the vision and leads to vision loss if left untreated. In this work, we have proposed an automatic unsupervised method to classify severity of diabetic macular edema in color fundus images. Particle Swarm Optimization (PSO) algorithm is made use of for effective segmentation of exudates. Optic disc and fovea are detected employing mathematical morphology. Region of macula is marked by Early Treatment Diabetic Retinopathy Studies (ETDRS) grading scale. Severity of disease such as normal, stage 1 or stage 2 of diabetic macular edema is detected by the location of exudates. The proposed method is evaluated using 100 images of public ally available MESSIDOR database and performance figures of 82.5% for sensitivity, 100% for specificity and 93% for accuracy are obtained.

[1]  B. Zinman,et al.  Diabetic retinopathy and diabetic macular edema: pathophysiology, screening, and novel therapies. , 2003, Diabetes care.

[2]  Bunyarit Uyyanonvara,et al.  Automatic detection of diabetic retinopathy exudates from non-dilated retinal images using mathematical morphology methods , 2008, Comput. Medical Imaging Graph..

[3]  Rafael C. González,et al.  Local Determination of a Moving Contrast Edge , 1985, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  James Kennedy,et al.  Particle swarm optimization , 2002, Proceedings of ICNN'95 - International Conference on Neural Networks.

[5]  F L Ferris,et al.  Photocoagulation for diabetic retinopathy. Early Treatment Diabetic Retinopathy Study Research Group. , 1991, JAMA.

[6]  I. Introduction,et al.  An Overview of PSO- Based Approaches in Image Segmentation , 2012 .

[7]  Jon Atli Benediktsson,et al.  An efficient method for segmentation of images based on fractional calculus and natural selection , 2012, Expert Syst. Appl..

[8]  Kenneth W. Tobin,et al.  Automatic retina exudates segmentation without a manually labelled training set , 2011, 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[9]  Ahmed S. Fahmy,et al.  Segmentation of Diabetic Macular Edema in fluorescein angiograms , 2011, 2011 IEEE International Symposium on Biomedical Imaging: From Nano to Macro.

[10]  C. Sinthanayothin,et al.  images retinal blood vessels from digital colour fundus Automated localisation of the optic disc , fovea , and , 1999 .

[11]  N. Bressler,et al.  Diabetic Macular Edema: Current and Emerging Therapies , 2012, Middle East African journal of ophthalmology.

[12]  Michael H. Goldbaum,et al.  Locating the optic nerve in a retinal image using the fuzzy convergence of the blood vessels , 2003, IEEE Transactions on Medical Imaging.

[13]  R. Hornero,et al.  Feature Extraction and Selection for the Automatic Detection of Hard Exudates in Retinal Images , 2007, 2007 29th Annual International Conference of the IEEE Engineering in Medicine and Biology Society.

[14]  C. Walker.  Ophthalmology , 1859, Bristol medico-chirurgical journal.

[15]  N. Otsu.  A threshold selection method from gray level histograms , 1979 .

[16]  Zhou Xian-cheng.  Image Segmentation Based on Modified Particle Swarm Optimization and Fuzzy C-Means Clustering , 2009, 2009 Second International Conference on Intelligent Computation Technology and Automation.

[17]  Alireza Osareh,et al.  A Computational-Intelligence-Based Approach for Detection of Exudates in Diabetic Retinopathy Images , 2009, IEEE Transactions on Information Technology in Biomedicine.

[18]  Sungbin Lim,et al.  Automatic classification of diabetic macular edema in digital fundus images , 2011, 2011 IEEE Colloquium on Humanities, Science and Engineering.

[19]  Ataollah Ebrahimzadeh,et al.  Brain tissue segmentation using an unsupervised clustering technique based on PSO algorithm , 2010, 2010 17th Iranian Conference of Biomedical Engineering (ICBME).