A not uncommon practice in the medical world involves scanning a patient with different slice thicknesses in different parts of the anatomy, all within the same DICOM series. For example, the patient is imaged with CT or MRI at high resolution in the sup-inf direction around a tumor but lower resolution at increased distances superior or inferior. In other words, slices in the Z direction will not necessarily have a fixed pixel size for axially acquired data.
My question is around support for this variably spaced scenario.
- Does ITK support importation of a DICOM series in which the slice thickness varies between slices? My current belief is that a 3D volume of data in ITK assumes a static, consistent pixel size throughout the volume.
- If there is support for this scenario, is this supported in the ITK DICOM Importers based off of DCMTK or GDCM or both? Or other, I guess? Or is the limiting factor still the construct of the 3D volume in ITK having a fixed pixel size in Z?
- Has anyone out there come up with a workaround, assuming no support in ITK? The best that I can think of is to create a resampled version of the variably sliced data into a fixed thickness 3D ITK volume. This would be a decent amount of work, mainly involving importation of each slice individually and post processing them into a resampled, fixed thickness 3D volume. Given the relative commonality of these types of scans, I was wondering if someone had already solved the issue robustly.
Thanks.