Citation

Mahan MY, Rafter DJ, Truwit CL, Oswood M, Samadani U. Magn. Reson. Imaging 2020; ePub(ePub): ePub.

Copyright

(Copyright © 2020, Elsevier Publishing)

DOI

10.1016/j.mri.2020.12.012

PMID

unavailable

Abstract

PURPOSE: Mild TBI, characterized by microstructural damage, often undetectable on conventional imaging techniques, is a pervasive condition that disturbs brain function and can potentially result in long-term deficits. Deciphering the underlying microstructural damage in mild TBI is crucial for establishing a reliable diagnosis and enabling effective therapeutics. Efforts to capture this damage have been extensive, but results have been inconsistent and incomplete.

METHODS: To that effect, we set out to examine the shape of the diffusion tensor in mild TBI during the acute phase of injury. We inspected diffusivity and geometric measurements describing the diffusion tensor's shape and compared mild TBI (N = 34, 20.4-66.6 yo) measurements with those from healthy control (N = 42, 20.7-67.2 yo) participants using voxelwise tract-based spatial statistics. Subsequently, to explore associations between the diffusion measurements in mild TBI, we performed nonparametric statistics and machine learning techniques.

RESULTS: Overall, mild TBI displayed a diffuse increase in D(λ2), D(λ3), D(radial), D(mean), and C(spherical), with a diffuse decrease in A(fractional), A(mode), and C(linear), in addition to no change in D(axial) or C(planar). Most notably, our results provide evidence for D(radial) as a potential biomarker for microstructural damage, specifically its main component D(λ2), based on their performance in discriminating between mild TBI and control groups. A(fractional) was also found to be important for discriminating between groups.

CONCLUSION: Our results revealed the importance of a diffusion measurement often overlooked, D(radial), in assessing TBI and suggest differentiating diffusion measurements has the potential utility to detect variations in the underlying pathophysiology after injury.

Language: en

Keywords

Mild traumatic brain injury; Diffusion tensor imaging; Fractional anisotropy; Radial diffusivity; Voxelwise tract-based spatial statistics