search

rordenlab / dcm2niix

dcm2nii DICOM to NIfTI converter: compiled versions available from NITRC
https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage
Other
898 stars 229 forks source link

Incorrect value of PhaseEncodingDirection on Siemens 3D T2w SPACE scans #849

Closed mharms closed 3 months ago

mharms commented 3 months ago

3D T2w SPACE scans on Siemens acquired with a typical Sagittal orientation with A >> P phase encoding direction end up with a sidecar json entry of "PhaseEncodingDirection": "i", which implies a R/L phase encoding axis after the inherent reorientation/rotation by dcm2niix of a 3D scan into a "RAS" oriented volume. Either PhaseEncodingDirection should not be reported if the reorientation changes the volume orientation (easiest), or should be adjusted to reflect the consequence of the reorientation (harder).

3D T1w MPRAGE scans from Siemens do NOT have a reported PhaseEncodingDirection value in their json. I'm not sure if that is because the relevant DICOM field is missing from the MPRAGE (but present in the SPACE), or if the dcm2niix code ignores the value for the MPRAGE (but not for the T2w SPACE).

neurolabusc commented 3 months ago

@mharms I agree that we should only use the PhaseEncodingDirection for 2D sequences. Can you share a sample with my institutional address. It is worth noting that 3D sequences use phase encoding for two of the three dimensions (e.g. phase encoding is also used in the slice direction).

mharms commented 3 months ago

Sure, will do. Meanwhile, I'll post the following interesting observation here, related to versioning.

I'm getting a PhaseEncodingDirection tag for a T2w SPACE scan for versions v1.0.20220720 and v1.0.20240202, but not for the stable release between those two, v1.0.20230411

neurolabusc commented 3 months ago

@mharms please test latest commit to development branch and close this issue if it correctly resolves this.

mharms commented 3 months ago

Good. Thanks.