Grading Of Diabetic Macular Edema Research Articles

Application of higher-order spectra for automated grading of diabetic maculopathy.

Diabetic macular edema (DME) is one of the most common causes of visual loss among diabetes mellitus patients. Early detection and successive treatment may improve the visual acuity. DME is mainly graded into non-clinically significant macular edema (NCSME) and clinically significant macular edema according to the location of hard exudates in the macula region. DME can be identified by manual examination of fundus images. It is laborious and resource intensive. Hence, in this work, automated grading of DME is proposed using higher-order spectra (HOS) of Radon transform projections of the fundus images. We have used third-order cumulants and bispectrum magnitude, in this work, as features, and compared their performance. They can capture subtle changes in the fundus image. Spectral regression discriminant analysis (SRDA) reduces feature dimension, and minimum redundancy maximum relevance method is used to rank the significant SRDA components. Ranked features are fed to various supervised classifiers, viz. Naive Bayes, AdaBoost and support vector machine, to discriminate No DME, NCSME and clinically significant macular edema classes. The performance of our system is evaluated using the publicly available MESSIDOR dataset (300 images) and also verified with a local dataset (300 images). Our results show that HOS cumulants and bispectrum magnitude obtained an average accuracy of 95.56 and 94.39% for MESSIDOR dataset and 95.93 and 93.33% for local dataset, respectively.

Automated detection of exudates and macula for grading of diabetic macular edema

Medical systems based on state of the art image processing and pattern recognition techniques are very common now a day. These systems are of prime interest to provide basic health care facilities to patients and support to doctors. Diabetic macular edema is one of the retinal abnormalities in which diabetic patient suffers from severe vision loss due to affected macula. It affects the central vision of the person and causes total blindness in severe cases. In this article, we propose an intelligent system for detection and grading of macular edema to assist the ophthalmologists in early and automated detection of the disease. The proposed system consists of a novel method for accurate detection of macula using a detailed feature set and Gaussian mixtures model based classifier. We also present a new hybrid classifier as an ensemble of Gaussian mixture model and support vector machine for improved exudate detection even in the presence of other bright lesions which eventually leads to reliable classification of input retinal image in different stages of macular edema. The statistical analysis and comparative evaluation of proposed system with existing methods are performed on publicly available standard retinal image databases. The proposed system has achieved average value of 97.3%, 95.9% and 96.8% for sensitivity, specificity and accuracy respectively on both databases.

Severity Grading of DME from Retina Images: A Combination of PSO and FCM with Bayes Classifier

Diabetic macular edema (DME) is the main cause of visual impairments in patients with diabetic retinopathy and leads to vision loss if left untreated. In this paper, an automatic approach for severity grading of DME is introduced. The approach involves preprocessing, combination of Particle Swarm Optimization (PSO) algorithm and Fuzzy C-Means Clustering for exudates segmentation, optic disc elimination, fovea and macular region localization, and classification. The Bayes classifier separates the lesions to exudates and nonexudates. The severity of the disease is graded into categories such as normal, grade 1 and grade 2 based on the location of exudates. Region of macula is marked by Early Treatment Diabetic Retinopathy Studies (ETDRS) grading scale. The proposed method is evaluated using 200 images of publically available MESSIDOR database and performance figures of 91% for sensitivity, 98% for specificity and 94.5% for accuracy are obtained.

An epidemiological surrey on diabetic macular edema in Huizhou city

Objective To investigate the prevalence of diabetic macular edema (DME) in Huizhou city and to explore the relationship of DME with the degrees of diabetic retinopathy (DR) and visual loss and the course of diabetes mellitus (DM). Methods Population samples in Huizhou were randomly selected to survey with questionnaires for DM screening. The patients then received vision tests, +90D front- position eyeground examination under a slit-lamp microscope, fundus photography, and fluorescence fundus an-giography (FFA) for DR and DME. Results Of 645 with DM, 262 developed DR and 92 occurred DME. The incidence of DME was 0.91% in Huizhou population, 1.43% in patients with DM, and 35.1% in the diabetic patients with DR. In DME 38.20% was mild, 31.46% was moderate, and 30.33% was severe. DME worsened with the increasing degree of DR (r= 0.945, P= 0.035). Visual loss was significantly correlated with the degree of DME (r= 0.972, P= 0.005). The grading of DME was markedly associated with the disease course of DM (r= 0.954, P= 0.008). Conclusions DME leads to severe vision impairment in diabetic patients. The patients with DME should be screened and treated earlier to reduce the incidence of DR-induced blindness. The degree of DME is markedly related with the stages of DR, visual loss, and course of diabetes. Key words: Diabetic macular edema; Diabetic retinopathy; Diabetes mellitus

An Extension of the Early Treatment Diabetic Retinopathy Study (ETDRS) System for Grading of Diabetic Macular Edema in the Astemizole Retinopathy Trial

Purpose: To compare the effects of astemizole, an antihistamine, versus placebo on the 1-year course of diabetic macular edema (DME) and to illustrate use of a modified ETDRS system for grading areas of retinal thickening and hard exudates that may be useful in clinical trials of treatments for this disorder. Methods: Between June 1994 and September 1997, at 2 clinics, 63 patients who had, in at least one eye (the study eye), DME that had not previously been treated with macular photocoagulation, and for which photocoagulation was not currently recommended by the investigator, were enrolled and randomly assigned to astemizole or placebo. Fifty-four of the 63 patients (86%, 26 in Clinic 1 and 28 in Clinic 2) completed 1 year of followup and had adequate 7-field stereoscopic film-based color fundus photographs of the study eye at the baseline and 1-year visits. DME was > 0.33 disc diameters (DD) from the center of the macula in 48% of study eyes and involved the center in 13%. Photographs were graded using the ETDRS protocol modified to allow estimates of areas of retinal thickening (RT) and hard exudate (HE) to be made on continuous scales in disc area (DA) units. Principal outcome measures were mean change in the square root of RT area (the average diameter of the area in DD), mean change in area of HE, and change in the degree to which RT involved or threatened the center of the macula. Results: At baseline, RT area in the 54 study eyes ranged from 0.09 to 4.0 DA (median 1.1). At the 1-year visit the square root of RT area (RTdd) had decreased by ≥ 0.3 DD in 10 eyes, increased by ≥ 0.3 DD in 19 and was about the same in 25. Mean change at 1 year was +0.09 DD (SD 0.57) for astemizole versus +0.19 DD (SD 0.48) for placebo, for a difference of −0.10 DD (95% CI −0.38, +0.19; p = 0.51). Adjustments for baseline and time-dependent risk factors did not change this result appreciably, although there was a trend towards a difference in favor of astemizole in the subgroup of patients with more severe retinopathy. Other morphologic outcomes paralleled change in RTdd. Change in RTdd did vary by clinic: −0.03 DD in Clinic 2, versus + 0.32 DD in Clinic 1, for a difference of −0.35 DD (95% CI −0.62, −0.07; p = 0.014). Clinic 1 is a tertiary retinal referral center in Pennsylvania and Clinic 2 a retinal clinic closely affiliated with a large diabetes clinic in Copenhagen. The unexpected clinic difference in outcome provided an opportunity for further analyses using the modified ETDRS system. In comparison to Clinic 1, Clinic 2 patients were more often male, were younger at diagnosis of diabetes, and had less severe retinopathy and better visual acuity, but these differences did not appear to explain the trend for lesser increase in RTdd. Conclusion: No effect of astemizole was found, but the confidence interval for the principal outcome, mean change in RTdd, included both a modest beneficial effect and a small harmful effect. This outcome measure did demonstrate a small difference in outcome by clinic, which could not be explained by baseline characteristics but may reflect differences in access to and/or continuity of care or other unmeasured differences associated with different referral patterns. Although optical coherence tomography may supplant photography as a measure of central RT, photographic assessments of change in RT and HE areas analyzed with the methods described herein may be useful outcomes in trials assessing treatment of early stages of DME. Application of these methods to other data sets is needed to confirm this conclusion.