Artikel
Influence of diffusion weighting factor b and voxel size on fiber tractography results for the corticospinal tract
Suche in Medline nach
Autoren
Veröffentlicht: | 13. Mai 2014 |
---|
Gliederung
Text
Objective: Tractography based on diffusion weighted (DW)-images has been progressed in order to optimize reconstruction results. However, in studies on tractography based on diffusion tensor imaging (DTI), parameters of DW-imaging schemes differ regarding for example gradient directions, b-value or voxel size. We present a prospectively conducted analysis of DTI-based tractography for the corticospinal tract with the frequently used tensor deflection algorithm, based on DTI-datasets with different voxel-size and b-value. To investigate the effects on the resulting fiber object, statistical analysis compared mean fractional anisotropy (FA)-value, mean length, number of fibers and volume of the resulting fiber object.
Method: Twenty healthy volunteers were included in this prospectively designed study. DW-images were acquired at a 3T MRI. Fiber tractography for the corticospinal tract was performed with tensor deflection algorithm, implemented in iplan cranial (BrainLab). DW-protocol included isotropic (3mm) and anisotropic voxels (1.5x1.5x3mm and 3x3x1.5mm), b-values of 700 or 1000 sm-2, resulting in 6 DTI-schemes. The starting region of interest was placed in the precentral gyrus, one include region in the mesencephal peduncle, exclude regions manually placed according to neuroanatomical knowledge.
Results: Descriptive analysis and MANOVA for the reconstructed corticospinal tracts showed different results regarding object volume and no. of fibers in favour for isotropic voxel size and higher b-values. However objects of significantly higher density and volume were obtained with isotropic 3mm voxels compared with 1.5x1.5x3mm voxel size regardless the b-value (p<0.001). A significant difference comparing 3x3x3mm and 3x3x1.5mm voxelsize did not occur. Hereby, a direct correlation between no. of fibers and the resulting fiber object was seen. There were no statistically significant differences regarding mean average FA-value or average maximum length of the resulting fiber object.
Conclusions: Despite the lack of a common consensus with a varity of DTI-schemes used in common literature, our study suggests that voxel size and b-factor result in fiber tractography objects of different quality. Although showing varied tractography results, DTI-based fiber tracking is a feasible method, while using well balanced DW–parameters.