phantomfit_DD.pl

hierarchical nonlinear fitting constrained by spherical harmonics using Diffeomorphic Demons

phantomfit_DD.pl [options] source.mnc target.mnc fit_mask.mnc estimate_mask.mnc [...] output.xfm

DESCRIPTION

phantomfit_DD.pl is a variant of phantomfit.pl that uses Diffeomorphic Demons for the nonlinear registration step instead of minctracc. The deformation field is constrained by a spherical-harmonic basis to produce a smooth, diffeomorphic parametric distortion model.

The fitting proceeds through multiple resolution levels. Two masks are required: a fit mask for registration optimization and an estimate mask for distortion parameter estimation. Additional mask pairs can be provided. Like phantomfit_ANTS.pl, this variant does not expose the weight, stiffness, or similarity parameters available in phantomfit.pl, as Diffeomorphic Demons uses its own internal regularization.

OPTIONS

-verbose: Print progress information during processing.
-debug: Enable debug mode with additional diagnostic output.
-clobber: Overwrite existing output files.
-fake: Print the commands that would be executed without running them.
-init_xfm <xfm>: Use the specified transformation as the initial starting estimate.
-order <n>: Maximum order of the spherical-harmonic expansion. Default: 5.
-par <file>: Write the spherical-harmonic parameter file to the specified path.
-measure: Only measure the distortion without iterative fitting.
-step_iterations <n>: Number of iterations at each step of the hierarchical registration.
-min_step <f>: Minimum step size in millimetres for the finest level of the hierarchy. Default: 1.
-limit: Limit the magnitude of the estimated distortion field.
-keep <f>: Fraction of the distortion field to retain after filtering. Default: 1.0.
-cylindric: Use cylindrical coordinate representation for the distortion model.
-init <xfm>: Provide an initial linear transformation for the registration.
-work_dir <dir>: Specify a working directory for intermediate files.
-pca: Use PCA regularization for the distortion model.
-pcs <n>: Number of principal components to retain when using PCA regularization.

EXAMPLES

Fit a distortion model using Diffeomorphic Demons:

phantomfit_DD.pl phantom_scan.mnc phantom_model.mnc \
    fit_mask.mnc estimate_mask.mnc output.xfm

Fit with higher-order harmonics:

phantomfit_DD.pl -order 7 -min_step 0.5 -verbose \
    phantom_scan.mnc phantom_model.mnc \
    fit_mask.mnc estimate_mask.mnc output.xfm

Fit using cylindrical coordinates with parameter output:

phantomfit_DD.pl -cylindric -par distortion.par -clobber \
    phantom_scan.mnc phantom_model.mnc \
    fit_mask.mnc estimate_mask.mnc output.xfm

AUTHOR

Vladimir S. Fonov. McConnell Brain Imaging Centre, Montreal Neurological Institute, McGill University.

COPYRIGHTS