Compensation through Functional Hyperconnectivity: A Longitudinal Connectome Assessment of Mild Traumatic Brain Injury

Figure 1

Pipeline for identifying the locations of DICCCOLs on the brain of an individual. (a) Fiber tracking and tractography of the whole brain was performed via MedINRIA (http://med.inria.fr/). Box “” presents the preprocessing steps (including brain extraction, motion correction, and eddy current correction) and deterministic tractography. (a)(I) shows diffusion data of an individual brain at and some different gradient directions, and (a)(II) shows the result of tractography in 3D space in the sagittal, axial, and coronal views. (b) The transformation matrix to transfer coordinates from the subject space to the template space was obtained by registering the brain of an individual subject to the brain template. (b)(III) shows the schematic of this procedure in which box represents the transformation matrix. (b)(IV) and (b)(V) show the coronal, axial, and sagittal views of individual and template’s brains, respectively. (c) Transformation and identification of the initial location of DICCCOLs. The transformation matrix () is applied to transfer the individual surface and fiber bundles to the template space for prediction. As a result, the initial location of DICCCOLs on the individual’s brain is obtained. (c)(VI) is the surface of an individual in the subject space, and (c)(VII) is the surface of the same individual, which is transferred to the template space. The initial location of DICCCOLs on an individual’s brain was obtained by overlaying the location of DICCCOLs of the template on the transformed surface of the individual. (d) The schematic procedure of optimization in which the local neighborhood (6 mm radius) was searched in order to identify the location where the profile of connected fiber has the most similarity with the WM fiber connection profile of the DICCCOL on the template. (d)(VIII) shows the initial location of a DICCCOL, obtained from the previous step. Using the information of deterministic tractography ((a)(II)), the connected fiber at this initial location was extracted ((d)(X)). Next, the similarity between the connected fibers at this location and the connected WM fiber on the template was measured. The same procedure took place for all local neighborhoods, and the location with maximum similarity of the connected WM fibers was identified as the optimized location of the DICCCOL. Box represents the optimization procedure. (d)(IX) shows the initial and optimized locations of a DICCCOL in red and green, respectively. (d)(X) and (d)(XI) show the connected fibers at the initial and optimized locations of the DICCCOL, respectively. (e) represents the optimized locations of all DICCCOLs on an individual’s brain. (e)(XII), (e)(XIII), and (e)(XIV) show the coronal, sagittal, and axial views in 3D space, respectively.

(a)

(b)

(c)

(d)

(e)