Automatic segmentation of the pyramidal tract basing on whole-brain DTI imaging
Schoth F, Stracke P, Buergel U
A segmentation of the pyramidal tract is expected to improve neuro-navigation. We investigated 10 healthy subjects with DTI. One segmentation method used line propagation and selected paths who connected the peduncel cerebri to Brodmann area 4. The second method was voxel based and chose the voxels according to direction of the main diffusion component, the anisotropy and T1-intensity in a fuzzy logic approach. The segmentations reflect histological findings. The voxel based method showed less mean errors. Line propagation methods are expected to be more efficient in a disturbed anatomy. Both methods can also be applied for other large fiber tracts.