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

• Published in: Medical & biological engineering & computing 2015-12, Vol.53 (12), p.1319-1331
• Main Authors: Mookiah, Muthu Rama Krishnan, Acharya, U. Rajendra, Chandran, Vinod, Martis, Roshan Joy, Tan, Jen Hong, Koh, Joel E. W., Chua, Chua Kuang, Tong, Louis, Laude, Augustinus
• Format: Article
• Language: English
• Subjects: Adult; Analysis; Automation; Biomedical and Life Sciences; Biomedical engineering; Biomedical Engineering and Bioengineering; Biomedicine; Computer Applications; Computer engineering; Datasets; Diabetes; Diabetic retinopathy; Diabetic Retinopathy - classification; Diabetic Retinopathy - diagnosis; Diagnostic Techniques, Ophthalmological; Discriminant analysis; Edema; Grading; Human Physiology; Humans; Image Interpretation, Computer-Assisted - methods; Imaging; Information systems; Macular Edema - classification; Macular Edema - diagnosis; Medical diagnosis; Medical services; Middle Aged; Morphology; Original Article; Radiology; Redundancy; Regression analysis; ROC Curve; Spectra; Studies; Support vector machines; Visual acuity; Young Adult
• ISSN: 0140-0118; 1741-0444
• DOI: 10.1007/s11517-015-1278-7

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.