Open laureld04 opened 1 month ago
Hello,
First of all thank you for your comments. Second, I need to ask a couple of questions.
1) Are you using the most up to date version from Github? Because over the past 2-3 months, we improved the rigid registration part SIGNIFICANTLY. It is much more robust than it used to be.
2) TORTOISEProcess has many rigid registration at different parts of the pipeline. Which one exactly are you referring to? Is it during DRBUDDI or at the final part (StructuralAlignment) which registers the DWIs to an anatomical.
3) Either way --DRBUDDI_initial_fixed_transform is not what you are looking for . That is between the up and down data, and not the structural.
4) If you are using the most up to date version, and if it still fails, after you clarify which step , I can include this option to provide a transform.
Thanks.
On Tue, Oct 22, 2024 at 1:07 PM Laurel Dieckhaus @.***> wrote:
Hello, thank you so much for this fantastic pipeline, our lab uses TORTOISE on a regular!
I had a question regarding TORTOISEProcess, specifically if its possible to provide an initial rigid instead of running the initial rigid registration in TORTOISEProcess? We have a couple of samples that fail the rigid registration during TORTOISEProcess (due to their structurals not overlapping with their corresponding DWIs). We do however have a rigid transform file from ants that registers the B0 to the T2. Is there a parameter in TORTOISEProcess that allows for this?
Apologies if this is obvious and I have missed something in the documentation, I thought it might be the --DRBUDDI_initial_fixed_transform and --DRBUDDI_initial_moving_transform but was not 100% sure.
Thank you again for any information or guidance you can provide on this!
— Reply to this email directly, view it on GitHub https://github.com/QMICodeBase/TORTOISEV4/issues/16, or unsubscribe https://github.com/notifications/unsubscribe-auth/AEFQZWCCXMAPGZG3I6J6IW3Z42A6NAVCNFSM6AAAAABQM7CZGCVHI2DSMVQWIX3LMV43ASLTON2WKOZSGYYDMMBSGA4DAMA . You are receiving this because you are subscribed to this thread.Message ID: @.***>
-- M. Okan Irfanoglu Graduate Student & GTA Computer Sciences & Engineering Ohio State University
Hello! I'm experiencing similar issues. I have a T2w image as the structural and blip up/down images, all exported to nii with associated json headers. The T2w is resampled to the same resolution and frame of reference as the DWI. However, the registration fails at some point, and I don't understand in which step, or why. I have attatched some screens shots that I hope can be helpful. In this instance I use --DRBUDDI_start_with_diffeomorphic_for_rigid_reg 1
, but I get the same issue if I set it to 0.
I installed this a couple of weeks ago following the installation guide, so I assume it is the latest version. How can I check which version I'm using?
Any help is appreciated!
Hello, thank you so much for this fantastic pipeline, our lab uses TORTOISE on a regular!
I had a question regarding TORTOISEProcess, specifically if its possible to provide an initial rigid instead of running the initial rigid registration in TORTOISEProcess? We have a couple of samples that fail the rigid registration during TORTOISEProcess (due to their structurals not overlapping with their corresponding DWIs). We do however have a rigid transform file from ants that registers the B0 to the T2. Is there a parameter in TORTOISEProcess that allows for this?
Apologies if this is obvious and I have missed something in the documentation, I thought it might be the --DRBUDDI_initial_fixed_transform and --DRBUDDI_initial_moving_transform but was not 100% sure.
Thank you again for any information or guidance you can provide on this!