BrownBiomechanics / SlicerAutoscoperM

This 3D Slicer extension enables users to perform image registration.
https://autoscoperm.slicer.org
MIT License
0 stars 3 forks source link

ENH: Hierarchical approach to 3d registration #99

Open NicerNewerCar opened 1 month ago

NicerNewerCar commented 1 month ago

Closes #94 Assuming you have already created a sequence of CT nodes and extracted the relevant volumes you wish to track, you should be able to follow the following steps to run the 3D to 3D pipeline.

Creating a Hierarchy

Navigate to the data module and select the Subject Hierarchy tab. Right click on the node list and select Create new subject, name this whatever you would like. Then you can drag the partial volumes into a hierarchy. image

Tracking

Given a sequence of n frames, this module will track a set of volumes over the frames [1,n]. This assumes that frame 0 is the neutral volume or the volume that the partial volumes were segmented from (the transform is identity).

Some small volumes, such as the pisiform, will fail to track.

https://github.com/BrownBiomechanics/SlicerAutoscoperM/blob/c6983696348ba42b1b7ef3e2b3addd9ec61feac9/Tracking3D/Resources/ParameterFiles/rigid.txt#L37-L41

Navigate to the Tracking3D module, this is under the tracking folder. Ensure that you select the subject you created in the previous step. You can also optionally select and initial guess transform. Click the apply button to launch the tracking, this will take a while. Some information will be sent to the log during the process. The apply button will transform into a cancel button if stopping the registration is necessary.

Before registration During registration
image image

image

3d_reg

Exporting Transforms as an Autoscoper TRA file

Simply ensure that the subject is selected then press the Export transforms button. All tra files will be written to the {slicer cache dir}/AutoscoperM-Pre-Processing/Tracking folder (same directory used by the AutoscoperM module).

3d_reg-aut