MRI-based measures of intracortical myelin are sensitive to a history of TBI and are associated with functional connectivity

Abstract

Traumatic brain injuries (TBIs) induce persistent behavioral and cognitive deficits via diffuse axonal injury. Axonal injuries are often examined in vivo using diffusion MRI, which identifies damaged and demyelinated regions in deep white matter. However, TBI patients can exhibit impairment in the absence of diffusion-measured abnormalities, suggesting that axonal injury and demyelination may occur outside the deep white matter. Importantly, myelinated axons are also present within the cortex. Cortical myelination cannot be measured using diffusion imaging, but can be mapped in-vivo using the T1-w/T2-w ratio method. Here, we conducted the first work examining effects of TBI on intracortical myelin in living humans by applying myelin mapping to 46 US Military Veterans with a history of TBI. We observed that myelin maps could be created in TBI patients that matched known distributions of cortical myelin. After controlling for age and presence of blast injury, the number of lifetime TBIs was associated with reductions in the T1-w/T2-w ratio across the cortex, most significantly in a highly-myelinated lateral occipital region corresponding with the human MT+ complex. Further, the T1-w/T2-w ratio in this MT+ region predicted resting-state functional connectivity of that region. By contrast, a history of blast TBI did not affect the T1-w/T2-w ratio in either a diffuse or focal pattern. These findings suggest that intracortical myelin, as measured using the T1-w/T2-w ratio, may be a TBI biomarker that is anatomically complementary to diffusion MRI. Thus, myelin mapping could potentially be combined with diffusion imaging to improve MRI-based diagnostic tools for TBI.

Keywords: Biomarker; Functional connectivity; Intracortical myelin; MRI; Traumatic brain injury.

Figure 1:. Mapping intracortical myelin in the human brain.

A) Illustration of the procedure for mapping intracortical myelin (Glasser and Van Essen, 2011). A T1-weighted (left) and T2-weighted (middle) image were registered to each other, and a ratio between the two images was calculated in each voxel (right), representing myelin content. Values between the pial surface (blue lines) and the gray-white border (green lines) represent intracortical myelin in the region surrounding the central sulcus (marked “CS”), and were mapped onto a 2D cortical surface. B) Distribution of the intracortical myelin contrast signal across the brain. The spatial distribution of the cortical myelin contrast calculated across 210 healthy controls (in Glasser et al., 2016; data available at https://balsa.wustl.edu/mpwM) was similar to those observed in individual subjects with few (middle) and with many (right) lifetime TBIs.

Figure 2:. Increasing number of lifetime TBIs is broadly associated with reduced intracortical myelin.

A) Partial correlations between number of lifetime TBIs suffered and the intracortical myelin contrast signal (controlling for age and blast exposure) conducted at each cortical vertex. Note the consistent and widespread negative associations between the variables. B) The distribution of the strength of TBI-myelin associations. The vast majority of these partial R values are negative. C) The distribution of median partial R values after 10,000 random permutations of subject identity. The real median partial R value (red dotted line) was significantly more negative than the permuted values.

Figure 3:. Significant local associations between number of lifetime TBIs and intracortical myelin emerged most strongly in a region overlapping the human MT+ complex.

A) The strength of the intracortical myelin contrast in a cluster in left lateral occipital cortex was significantly associated with number of lifetime TBIs (after controlling for age and blast exposure) after correcting for multiple comparisons. B) This significant cluster (outline in white dots) closely overlapped a myelin hotspot present across 4210 healthy controls (in Glasser et al., 2016; data available at https://balsa.wustl.edu/mpwM), which was identified as the human MT+ complex by Glasser and Van Essen, 2011.

Figure 4:. Intracortical myelin in the MT+ region is associated with Visual network functional connectivity.

A) The TBI-influenced region (white outline) corresponds well with an a priori parcel (from Gordon et al., 2016) in the Visual network (blue regions). B) The strength of the intracortical myelin contrast within the putative MT+ region (x-axis) was positively associated with the strength of functional connectivity between this region and the Visual network parcels (y-axis).