question on deformationField
chenyangliu2018 opened this issue · 8 comments
Nice work. The output of deformationField looks like a deformed mesh grid instead of deformation vector field (DVF). May I know how to get the DVF? Thanks.
I edited the motionfield by subtracting a meshgrid with 0:1:N-1, where N is the dimension size. I was wondering is that correct?
Hi. The upstream codebase includes a command-line tool called reg_transform which can convert between different transformation formats including deformation and displacement fields. Unfortunately there is no R interface to this, but that's the route I would recommend if possible, rather than modifying the field manually.
Thanks for the reply. I will try that one. Btw, I modified the field manually just now by using the method I mentioned. It seems it works.
Sure, it may well produce sensible results numerically, but NiftyReg will not interpret it correctly after that change without some metadata being updated too. But that may not be important in your use-case.
I am kind of new learner on MRI reconstruction. If I want to incorporate the RNifftyReg in to a conjugate iteration. Could I get the transpose of the motionfield from this package. If not, any suggestion on how to deal with it. Thank you very much.
Sorry, I'm not sure what you're trying to do from that information. What is the motion field you want? Can you give a concrete example?
Hello, thank you for your response. Sorry I haven't explain my question well. The problem is if Iwant solve Mx = y, Imay need the M transpose, where M is the motionfield. I may have figured this problem out. My method is compose a sparse metrix that represent the motionfield and direct transpose it. Let me see whether it works or not.
I am assuming this is resolved – do reopen it if not, but I will need more information to be able to help.