Diffusion tensor imaging of early relapsing-remitting multiple sclerosis with histogram analysis using automated segmentation and brain volume correction

Abstract

Diffusion tensor magnetic resonance imaging (DTI) reveals measurable abnormalities in normal-appearing brain tissue (NABT) in established multiple sclerosis (MS). However, it is unclear how early this occurs. Recent studies have employed whole brain histogram analysis to improve sensitivity, but concern exists regarding reliability of tissue/cerebrospinal fluid segmentation and possible intersubject brain volume differences, which can introduce partial volume error: To address this, 28 early relapsing-remitting MS subjects [median disease duration 1.6 years; median Expanded Disability Status Scale (EDSS) score 1.5] and 20 controls were compared with whole brain histogram analysis using an automated segmentation algorithm to improve reproducibility. Brain parenchymal volumes (BPV) were estimated for each subject in the analysis. The mean, peak height and peak location were calculated for DTI parameters [fractional anisotropy (FA), mean diffusivity and volume ratio]. An increased FA peak height in MS subject NABT was observed (P = 0.02) accounting for age, gender and BPV. Removing BPV revealed additional abnormalities in NABT. The main conclusions are i) FA peak height is increased in NABT in early MS, ii) partial volume edge effects may contribute to apparent NABT histogram abnormalities, and iii) correction for brain volume differences should reduce potential partial volume edge effects.

Longitudinal evaluation of clinically early relapsing-remitting multiple sclerosis with diffusion tensor imaging

“…Baseline findings in this study are, as expected, similar to a cross-sectional analysis using a similar but not identical cohort [10] and again show FA to be a potentially N/A not applicable; FA fractional anisotropy; NABT normal appearing brain tissue; WBT whole brain tissue Fractional Anisotropy more sensitive DTI parameter at this stage. The suggestion of an increase in FA mean and peak location over time with values tending to return towards those of normal controls in patients is statistically weak.…”

“…Hence in this investigation, our image and statistical analyses tend to err on the side of caution and this may mask the emergence of a subtle disease effect, even when a robust methodology is used. This may be particularly magnified in clinically early RR MS, where abnormalities may be more subtle than those found in other MS sub-types with greater disability and disease duration [10,[34][35][36]. The subtle nature of any pathological changes at this stage of the disease is illustrated by the apparent absence of a significant measurable longitudinal abnormality in the corpus callosum.…”

“…In this paper, we present an investigation of DTI parameters with a cohort of early RR MS patients with a relatively homogenous disease duration and level of clinical disability over a two year period using a sensitive analysis technique and employing a segmentation programme used previously which has shown to provide adequate brain extraction [10]. A further registration algorithm between structural and diffusion parameter images has been added to improve the accuracy of the methodology [18,19].…”

“…Cross-sectional DTI studies have reported abnormalities both in whole brain tissue (WBT -brain tissue including lesions) and normal-appearing brain tissue (NABT -WBT with lesions extracted) in cohorts of relapsing-remitting (RR) multiple sclerosis (MS) patients, even at the earliest stages of the disease [4][5][6][7][8][9][10]. DTI changes correlate poorly with lesion volume, suggesting that the parameters are able to detect disease processes that are independent of MRI-visible white matter lesions ■ Abstract Diffusion tensor imaging (DTI) parameters such as mean diffusivity (MD) and fractional anisotropy (FA) assess aspects of structural integrity within tissue.…”

Novel diffusion anisotropy indices: An evaluation

“…FA and trace can be seriously affected by partial volume edge effects, which may contribute to apparent histogram abnormalities in normal-appearing brain tissue. However, in a previous study DTI was unable to clearly distinguish between the various pathophysiological processes involved (10).…”