Transformation Update

[NicerNewerCar]

• Adds DICOM2AUT tfm
• Splits {bone}.tfm into {bone}_t.tfm and {bone}_scale.tfm
• Renames and addresses issues with Slicer2Autoscoper transform (now PVOL2AUT)

[amymmorton]

building now... and will confirm, but a comment on the commit message re: splitting the .tfm: We do want all three variants: bone.tf boneScaleOnly.tfm and boneTransOnly.tfm

[NicerNewerCar]

building now... and will confirm, but a comment on the commit message re: splitting the .tfm: We do want all three variants: bone.tf boneScaleOnly.tfm and boneTransOnly.tfm

Okay, I will update this now to include the {bone}.tfm again

[amymmorton]

• [x] Pvol 2 aut

  suggest

• [ ] export of the AUT(bone).stl (LPS)

• [ ] change bone_t to inv(bone_t)

• [ ] the DICOM2AUT tfm is combined transform of the image position to aut AND RAS->LPS - suggest naming explicitly? (if user segments in mimics and applies this tform- (staying LPS) to the bone model or a cs gen from the anatomy it will be erroneous.

[NicerNewerCar]

• [ ] change bone_t to inv(bone_t)

Should this also apply to the naming of the bone_scale (ie. inv(bone_scale))

• [ ] the DICOM2AUT tfm is combined transform of the image position to aut AND RAS->LPS - suggest naming explicitly? (if user segments in mimics and applies this tform- (staying LPS) to the bone model or a cs gen from the anatomy it will be erroneous.

Would something like DICOM(RAS)2AUT(LPS) work or were you thinking of something else?

[amymmorton]

The bone_t.tfm are in RAS not LPS (name it explicitly?) (ground truth data prev sent 7/16 vs this branch)

• [x] GT AUT mc1 stl and AUT stl generated in this branch are spatially aligned
• [x] mc1-DICOM2AUT.tfm transforms the segmented mc1 stl to AUT space in RAS -slicer

the output tra (with LPS-RAS) tform applied is correct!