2.7. TABS method

TABS is an along-tract statistics analysis method for detecting WM microstructural characteristics based on streamline tractography, as described in our previous study (Liu et al., 2017, Liu et al., 2019, Mu et al., 2018). This method facilitates an easy interpretation of results by directly reporting the resulting statistics on the streamlines. Specific steps are as follows (Fig. 2):

Step 1: To ensure the statistical analysis across subjects was in a common space, we leveraged the ICBM_Mori_DTI_2mm diffusion tensor template (Fig. 2A). Whole-brain tractography (Fig. 2B) was performed within the template with a step size of 1 mm, an angle threshold of 30, and a minimum tract length of 50 mm.

Step 2: The streamlines linking the regions in the meta-analysis results were selected from the tractography data. The streamlines from the bilateral medial prefrontal cortex (MPFC) to the superior temporal gyrus (STG), bilateral MPFC to the inferior parietal lobule (IPL), and left to right hemispheres of the MPFC were selected (Fig. 2C–D). Specifically, the fiber clusters were chosen from the whole-brain tractography maps of the ICBM_Mori_DTI_2mm template using manually defined inclusion, AND, OR, and exclusion with the ExploreDTI toolbox (www.exploredti.com) (Leemans et al., 2009).

Step 3: For each fiber bundle, a representative streamline was employed from the selected streamlines with the greatest local density weighted length (Fig. 2E). The representative streamline was then modeled by using the continuous arc length function.

Step 4: To ensure all streamlines can be represented by a unified coordinate system (Fig. 2G), the coordinates in the representative streamline were transformed to all other streamlines by using the optimal point match method (Fig. 2F).

Step 5: Non-linear transformation was then used to map the FA images of all subjects (Fig. 2H) to the template FA image to obtain the transformation parameter, T, from native space to template space (Fig. 2I). The unified coordinate system calculated in step 4 was warped back into the individual’s native space to collect the diffusion measures with T. The individual diffusion measures (Fig. 2J) were then averaged along the streamline (Fig. 2K).

Framework of the tractography atlas-based analysis (TABS).