Mild traumatic brain injury, which accounts for up to 90% of traumatic brain injuries, is currently diagnosed and monitored with a thorough history and physical exam. There is growing consensus in the literature that pathophysiologic changes in white matter extend past symptomatic recovery, and that these biological changes should be the target of diagnosis and monitoring. This would replace the current clinical consensus that symptom resolution and a level of functioning that allows return to work/play indicates recovery. Diffusion magnetic resonance imaging has been a critical tool for studying white matter change, and the diffusion tensor model developed in the 1990s has been the cornerstone of studying white matter in human subjects with mild traumatic brain injury. This model has several limitations, however, which in part may contribute to inconsistencies observed in the localization and direction of changes in diffusion tensor metrics such as fractional anisotropy. The past decade has brought forth scanner advancements, including stronger gradients, high angular-resolution imaging, and modeling strategies specific to these imaging data. Here, we review these contemporary techniques and apply them in a secondary analysis of data collected during the Transforming Research and Clinical Knowledge in Traumatic Brain Injury Study. Constrained spherical deconvolution and fixed-based analysis were applied to diffusion magnetic resonance imaging data collected during the study, revealing significant differences in tract-specific and global fiber-density and fiber-density cross-section measures between patient and control groups. However, results remained relatively stable over time, indicating that in this population-based sample of patients, not much change occurred in the white matter from 2 weeks to 6-months post-injury. This would seem to suggest there is a drastically different timeline for white matter recovery than was previously thought, even in injuries categorized as mild in the general population. Additionally, this reinforces the notion that multimodal tools for diagnosis and monitoring need further evaluation to create gold-standard objective classifiers of injury severity. In the short-term, future work should focus on the development of an imaging technique that when collected in the acute period following injury can predict prolonged recovery in individual patients with reasonable accuracy. Finally, characteristics which convey resiliency to poor outcomes should be investigated to facilitate development of interventions that hasten biological recovery.
Read
