|
|
If you open a 4D data set from a PET study, and re-sort the stack into a
hyperstack (z=1-to-Z frames by t=1-to-T frames) you get the 4D sequence.
HOWEVER I believe the the activity values from 4D PET are wrong. Intensity
that is meant to reduce over time does not. I thought this was a global
contrast issue, but just by opening a single 2d slice I saw the same
intensity values in that slice (not when part of a hyperstack).
For the same voxel, ImageJ gives different intensity values to other
software. I believe there is a problem with the DICOM reader. Perhaps there
should be some correction for the FRAME DURATION, i.e. all counts get
averaged over the acquisition time of one frame.
Perhaps other software does this automatically (recognising the modality
and the framerate from the DICOM header).
Please take care when interpreting PET data from 4D data sets
|
Locked
