Nathan Molinier
Nathan Molinier
During the last SCT meeting (2024-10-01) we mentioned that it would be nice to have a first implementation of TotalSpineSeg before the next SCT course (beginning of december). So I...
I have a few concerns: - In the [package](https://pypi.org/project/totalspineseg/) requirements, we install a different nnunetv2 version but I believe it would work the same if we use the specified version...
Is it an issue ? I thought that we could just call the package functions using calls to bash functions such as: ```python import subprocess # Install weights subprocess.run([ "totalspineseg_init"...
Ok let me update the package then. Then, where should I add the code ?
I'm not sure I understand why it's not possible to combine Axial and Sagittal. For me the only difference is the orientation of the input, so I believe we could...
I would like to add a small comment related to how (I believe) the QC backend should look like without going into code details. For me, the complete QC code...
Maybe running the two inferences would be easier from a user perspective ? ~~~ # Label small FOV image sct_deepseg totalspineseg \ -i \ -localizer \ ~~~
@sten2lu Amazing ! Let me know if I can help with anything ! Best regards, Nathan
Move the labels outside of the segmentation: https://github.com/spinalcordtoolbox/spinalcordtoolbox/pull/5050#issuecomment-3614104862
Hey @joshuacwnewton, this PR is not complete because I was not sure how to deal with the parser, the object `QcImage` and the function `generate_qc`. It won't probably be merged...