We are labelling structures near joints (e.g. intervertebral, mandibular, some costovertebral) to initialise grow from seeds. I have a couple of questions about a possible process.
It could be quicker to work in the 3D view. We use painting including spheres on the 3D view but it is limited away from the surface. We use scissors in 3D as well but the background/depth etc. seems difficult to manage in 3D (variable sizes etc).
- We could paint the whole region (big sphere) and then use scissors to isolate the structures. It could be done after a couple of rotate
I see scissors can delete but is there any way to move what is selected to another segment (one now displayed at the time)?
If it could, then the convex structure could be moved (“shaved”) to structure 1 (here: skull) and the remaining would be moved to structure 2.
(image 1 below).
This segmentation process could be quite general.
Any workaround or fast similar approaches?
(besides doing the job twice : once adding and once deleting, which does not unsure the consistency)
- Volume Rendering has a ROI to crop the area of interest (e.g. joint). It would then also be easy with a couple of rotate to label a structure using scissors (especially with a convex surface) and would have the advantage of color (image 2). But unfortunately, the ROI is not considered by the scissors (image 3).
Any way to make scissors consider the ROI (=not to paint outside)? Or alternatives ?
(It seems that 1 is more important)
Thanks for any suggestion!
Image 1: scissors could be more useful if able to move to another segment not currently visible (an area could be segmented away with a couple of rotate).
Image 2: mandibular joint volume rendering with ROI. A scissors and a couple of rotate could label the skull and it is quite visual.
Image 3: however, scissors do not consider ROI