vmr.ApplyTRF

The main motivation for using this function is the ability to flexibly script spatial VMR transformation in Matlab.

 VMR::ApplyTRF  - apply transformation to VMR
 
 FORMAT:       newvmr = vmr.ApplyTRF(trf [, opts])
 
 Input fields:
 
       trf         TRF object
       opts        struct with optional fields
        .asdouble  store output as double (default: false)
        .inverse   apply inverse transformation (default: false)
        .method    interpolation, see flexinterpn_method (default: 'linear')
 
 Output fields:
 
       newvmr      transformed VMR

• using a TRF that was created in BrainVoyager (and instead of the file selector in the first two lines, this can be scripted!):

  vmr_applytrf_bvtrf.m

  %load VMR and TRF
  vmr = xff('*.vmr');
  trf = xff('*.trf');
   
  % apply transformation with cubic interpolation
  trfvmr = vmr.ApplyTRF(trf, struct('method', 'cubic'));
  trfvmr.SaveAs;
   
  % clear objects
  vmr.ClearObject;
  trf.ClearObject;
  trfvmr.ClearObject;

• using two VMRs, coregister them, then transform the first:

  vmr_applytrf_coreg.m

  % load two VMRs
  vmr1 = xff(sourcevmr);
  vmr2 = xff(targetvmr);
   
  % run coregistration (requires SPM5 or SPM8 on the path!!)
  % vmr2 is the target space (stationary), vmr1 the one to be resampled!
  trf = vmrspmcoreg(vmr1, vmr2)
   
  % resample vmr1 using sinc (lanczos3) interpolation
  vmr1_in_vmr2_space = vmr1.ApplyTRF(trf, struct('method', 'lanczos3'));
   
  % save TRF and VMR
  [sourcepath, sourcefile] = fileparts(sourcevmr);
  if isempty(sourcepath)
      sourcepath = '.';
  end
  [nullpath, targetfile] = fileparts(targetvmr);
  trf.SaveAs(sprintf('%s/%s-TO-%s.trf', sourcepath, sourcefile, targetfile));
  vmr1_in_vmr2_space.SaveAs(sprintf('%s/%s-TO-%s.vmr', sourcepath, sourcefile, targetfile));
   
  % clear objects
  clearxffobjects({vmr1, vmr2, trf, vmr1_in_vmr2_space});