A decade of DTI in traumatic brain injury: 10 years and 100 articles later

Abstract

The past decade has seen an increase in the number of articles reporting the use of DTI to detect brain abnormalities in patients with traumatic brain injury. DTI is well-suited to the interrogation of white matter microstructure, the most important location of pathology in TBI. Additionally, studies in animal models have demonstrated the correlation of DTI findings and TBI pathology. One hundred articles met the inclusion criteria for this quantitative literature review. Despite significant variability in sample characteristics, technical aspects of imaging, and analysis approaches, the consensus is that DTI effectively differentiates patients with TBI and controls, regardless of the severity and timeframe following injury. Furthermore, many have established a relationship between DTI measures and TBI outcomes. However, the heterogeneity of specific outcome measures used limits interpretation of the literature. Similarly, few longitudinal studies have been performed, limiting inferences regarding the long-term predictive utility of DTI. Larger longitudinal studies, using standardized imaging, analysis approaches, and outcome measures will help realize the promise of DTI as a prognostic tool in the care of patients with TBI.

Fig 1.

FA image (A) reveals no abnormality in a patient with TBI. Tractography (B) can be used to delineate a region of interest for analysis. In this case, the forceps major (red) appears normal, but quantitative analysis of FA within this tract showed lower FA in the TBI group compared with controls. Whole-brain voxelwise analysis (C) reveals areas of low (blue) and high (red) FA. Low FA, consistent with TAI, is present within the forceps major at the splenium of the corpus callosum, as well as elsewhere.

Fig 2.

The number of publications per year reporting DTI in TBI.

Fig 3.

The number of articles that studied patients at each timeframe and level of injury severity. Articles were only included if there was sufficient information to determine both the severity and the chronicity of individual patient injuries. Articles may be included multiple times if they studied subjects with multiple severities and/or multiple chronicities. A fully referenced version of this figure is available in On-line Table I.

Fig 4.

Thirteen studies used a longitudinal design. Numbers represent patients from all studies imaged at 2 time points. Nine studies assessed patients at both acute and subacute time points.^,,,,,,,, One study assessed patients at both acute and chronic time points. Two studies assessed patients at both subacute and chronic time points.^, One study (n = 47) assessed patients twice during the subacute period and, therefore, was omitted from the figure.