You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Suspect currently couldn't read a RDA format coming from Siemens XA60 software version.
Upon further inspection, some of the headers are changed, particularly the Siemens' VOI info.
In the old RDA, there are VoiPosition, VoiNormal, etc., but in the new RDA it was replaced by Slab and MidSlab, which are both 2D arrays.
VoiPosition, VoiNormal, etc. contain the information related to Voxel and if I recall correctly, VOI stands for "Volume of Interest". This is the same with what Siemens uses in the TWIX data. Suspect needs these VOI info in order to construct the MRS object.
When looking into accompanying DICOM files, the CSA header is also removed. Instead, it uses Slab and MidSlab info as well. As for DICOM, we can use the suspect.io.load_dicom instead of suspect.io.load_siemens_dicom, although that means we won't be able to run the to_scanner() method.
We will need to translate the voxel info in suspect's MRS object metadata to use Slab and MidSlab instead of the old VOIs. This means, we might also be able to add the transform to the general DICOM reader (suspect.io.load_dicom).
Suspect currently couldn't read a RDA format coming from Siemens XA60 software version.
Upon further inspection, some of the headers are changed, particularly the Siemens' VOI info.
In the old RDA, there are
VoiPosition
,VoiNormal
, etc., but in the new RDA it was replaced bySlab
andMidSlab
, which are both 2D arrays.VoiPosition
,VoiNormal
, etc. contain the information related to Voxel and if I recall correctly, VOI stands for "Volume of Interest". This is the same with what Siemens uses in the TWIX data. Suspect needs these VOI info in order to construct the MRS object.When looking into accompanying DICOM files, the CSA header is also removed. Instead, it uses
Slab
andMidSlab
info as well. As for DICOM, we can use thesuspect.io.load_dicom
instead ofsuspect.io.load_siemens_dicom
, although that means we won't be able to run theto_scanner()
method.We will need to translate the voxel info in suspect's MRS object metadata to use
Slab
andMidSlab
instead of the old VOIs. This means, we might also be able to add thetransform
to the general DICOM reader (suspect.io.load_dicom
).By googling, Slab and MidSlab are in the DICOM standard. See: https://dicom.innolitics.com/ciods/mr-spectroscopy/mr-spectroscopy/00189126
The text was updated successfully, but these errors were encountered: