Computational protocol: Functional segregation and integration within fronto-parietal networks

Similar protocols

Protocol publication

[…] We used diffusion tractography to identify the three branches of the SLF in 129 healthy right-handed volunteers (59 males and 70 females) aged between 18 and 79 years. For each participant, 60 contiguous near-axial slices were acquired on a 3 T GE Signa HDx TwinSpeed system (General Electric, Milwaukee, WI, USA) with the following parameters: rostro-caudal phase encoding, voxel size 2.4×2.4×2.4 mm, matrix 128×128, slices 60, NEX 1, TE 93.4 ms, b-value 3000 s/mm2, 60 diffusion-weighted directions and 7 non-diffusion-weighted volumes, using a spin-echo EPI sequence. Cardiac gating was applied with effective TR of 20/30 R-R intervals. Quality control of the data was assured using an automated analysis system (). Standard diffusion tensor tractography does not allow the reconstruction of the two most dorsal branches of the SLF because of the crossing of the dorsal association fibres with commissural and projection fibres (, , ).Crossing problems can be partially overcome by more recent methods, such as diffusion spectrum imaging (DSI) (Wedeen et al., 2008) and high angular resolution diffusion imaging (HARDI) (Frank, 2001; ; ; ). For instance, the latter estimates a distribution of possible fibre orientations in the three-dimensional space for each voxel. The result is a function, whose multi-peak shape reflects the orientation and weight of each fibre component (; Anderson, 2005; ; ). Among HARDI methods, tractography based on spherical deconvolution (SD) has been widely used to reconstruct white matter tracts in regions with multiple crossings, such as the SLF, which is object of the current investigation (, ; Catani et al., 2012; ; ; Budisavljevic et al., 2016; ). A modified (damped) version of the Richardson-Lucy algorithm for spherical deconvolution () was employed using the software StarTrack ( Algorithm parameters were chosen as previously described by our group ().Whole brain tractography selected every brain voxel with at least one fibre orientation as a seed voxel. From these voxels, we reconstructed the streamlines by sequentially piecing together discrete and shortly spaced estimates of fibre orientation to form continuous trajectories (). When entering a region with crossing white matter bundles, the algorithm followed the orientation vector of least curvature (). Streamlines were halted when a voxel without fibre orientation was reached or when the curvature between two steps exceeded a threshold of 45°. The software estimating and reconstructing the orientation vectors and the trajectories from diffusion MRI was written in Matlab 7.8 ( dissections of the SLF I, II and III were performed using a multiple regions of interest (ROIs) approach: in each hemisphere three ROIs were delineated around the white matter of the superior, middle and inferior/precentral frontal gyri, and a ROI around the white matter of the parietal lobe. In order to exclude fibres belonging to the long and posterior segment of the arcuate fasciculus, which respectively connect frontal or parietal regions with temporal regions, a no-part ROI was delineated around the temporal white matter. Further details can be found in ().For each participant, a convergence speed (CS) map of the deconvolution algorithm () was estimated. CS map quantifies how quickly the residual fitting error between the diffusion signal, and the fibre model as identified by the deconvolution algorithm decays within each voxel. CS maps better contrast white matter regions showing a smaller partial volume effect, as compared to FA or similar anisotropy maps. CS maps were registered to the MNI152 template provided with the FMRIB Software Library package (FSL, using Advance Normalisation Tools (ANTs,, which combines affine with diffeomorphic deformations (, ).Binary visitation maps were created for each tract by assigning each voxel a value of 1 or 0 depending on whether the voxel was intersected by the streamlines of the tract. Binary visitation maps of each dissected tracts were normalised to MNI space using the same affine with diffeomorphic deformations calculated above. We created percentage overlap maps using a previously published method () by summing at each point in the MNI space the normalised visitation maps from each subject; hence the overlap of the visitation maps varied according to inter-subject variability. displays the 3D rendering of the three SLFs onto the average rendering of the MNI152 template obtained using Anatomist 4.2 and BrainVISA 4.3 ( […]

Pipeline specifications

Software tools FSL, Anatomist, BrainVISA
Application Magnetic resonance imaging
Organisms Homo sapiens