Title: Virus infection on brain white matter: statistical analysis of DT MRI scans

Authors: R. Kalpana, S. Muttan, N. Kumarasamy

Addresses: Department of Electronics and Communication Engineering, CEG Campus, Anna University, Chennai 600025, India. ' Department of Electronics and Communication Engineering, CEG Campus, Anna University, Chennai 600025, India. ' YRG CARE Medical Centre-VHS, Taramani, Chennai 600113, India

Abstract: Brain White Matter (BWM) is formed on nerve fibres (due to myelin coating) and it undergoes continuous changes with aging throughout the life of humans. This age-related morphology, however, also gets further modified when a subject is infected by a viral disease such as HIV. In the present study, textural variations in BWM specific to a set of control images vs. images of a HIV + subject are assayed via an algorithm based on the Grey Level Co-occurrence Matrix (GLCM); and the associated Haralick features that are extracted from the GLCM. The clinical images of the brain used in the study are gathered from the noninvasive technique of Diffusion Tensor Magnetic Resonance Imaging (DT MRI). Results observed in the images of normal (nonpathogenic) subjects show progressive changes in the image parameters with aging; whereas, a wider variation is observed in the relevant parameters of the images pertinent to pathogenic subjects who exhibit cognitive (or motor impairments) due to HIV infection. The findings of the proposed method are reasonably correlated with clinical diagnosis in two HIV+ cases.

Keywords: brain white matter; DT MRI; diffusion tensor MRI; magnetic resonance imaging; texture; GLCM; grey level co-occurrence matrix; HIV infection; Haralick features; MRI scans; age-related morphology; viral diseases; clinical images; bioinformatics; cognitive impairment; motor impairment.

DOI: 10.1504/IJBRA.2011.041738

International Journal of Bioinformatics Research and Applications, 2011 Vol.7 No.3, pp.273 - 286

Published online: 24 Jan 2015 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article