This module is based on the phantom having discrete points that are readily thresholded in the deformation step to create a mask which is then used to mask the larger deformation field to avoid performing the analysis across every single voxel. This is fine for the LCTC phantom, however may not be useful more widely (including for the test phantom I'm looking t use). Is it worth in the analysis step allowing the user to input a mask for the deformation field? (Perhaps it could default to the one generated in the previous deformation step, but have the option to pull in a different user defined mask, while also allowing for no mask to be selected if preferred by the user).
So the analysis module would look like:
Registration CSV file (unmasked; auto populated)
Deformation field file (unmasked deformation field; auto populated)
Mask file for analysis subsampling (auto populated; used to mask larger csv - currently masking the csv file is done in the deformation step) {have option to not select one at all}
This module is based on the phantom having discrete points that are readily thresholded in the deformation step to create a mask which is then used to mask the larger deformation field to avoid performing the analysis across every single voxel. This is fine for the LCTC phantom, however may not be useful more widely (including for the test phantom I'm looking t use). Is it worth in the analysis step allowing the user to input a mask for the deformation field? (Perhaps it could default to the one generated in the previous deformation step, but have the option to pull in a different user defined mask, while also allowing for no mask to be selected if preferred by the user).
So the analysis module would look like: