dipy_bundlewarp#
Synopsis#
BundleWarp: streamline-based nonlinear registration.
BundleWarp [1] is a nonrigid registration method for deformable registration of white matter tracts.
Usage#
dipy_bundlewarp [OPTIONS] static_file moving_file
Input Parameters#
static_file
Path to the static (reference) .trk file.
moving_file
Path to the moving (target to be registered) .trk file.
General Options#
--dist
Path to the precalculated distance matrix file. (default: None)
--alpha
Represents the trade-off between regularizing the deformation and having points match very closely. Lower value of alpha means high deformations. It is represented with λ in BundleWarp paper. NOTE: setting alpha<=0.01 will result in highly deformable registration that could extremely modify the original anatomy of the moving bundle. (default: 0.3)
--beta
Represents the strength of the interaction between points Gaussian kernel size. (default: 20)
--max_iter
Maximum number of iterations for deformation process in ml-CPD method. (default: 15)
--affine
If False, use rigid registration as starting point. (default True) (default: True)
Output Options#
--out_dir
Output directory. (default: current directory)
--out_linear_moved
Filename of linearly moved bundle. (default: linearly_moved.trk)
--out_nonlinear_moved
Filename of nonlinearly moved (warped) bundle. (default: nonlinearly_moved.trk)
--out_warp_transform
Filename of warp transformations generated by BundleWarp. (default: warp_transform.npy)
--out_warp_kernel
Filename of regularization gaussian kernel generated by BundleWarp. (default: warp_kernel.npy)
--out_dist
Filename of MDF distance matrix. (default: distance_matrix.npy)
--out_matched_pairs
Filename of matched pairs; streamline correspondences between two bundles. (default: matched_pairs.npy)
References#
Garyfallidis, E., M. Brett, B. Amirbekian, A. Rokem, S. Van Der Walt, M. Descoteaux, and I. Nimmo-Smith. Dipy, a library for the analysis of diffusion MRI data. Frontiers in Neuroinformatics, 1-18, 2014.