Hello MRIcron team,

We just found a strange problem. dcm2nii is completely reversing our time series for our PET images. This is quite concerning. We tried with a 2013 version of dcm2nii, a 2016 version, and also with a GUI version on a Mac, all three programs exhibited this behavior.

I'm going to try to get some functional PET DICOMS from another scanner, to see if it's something in the way our machine is encoding the headers. But in the meantime, I thought you should be aware of the issue. Specific model is the Philips PET/CT Gemini TOF.

- Eli

I would carefully heed the warning whenever you run dcm2nii that you should use dcm2niix. dcm2nii is deprecated and will not be included in future releases. I would only recommend it for archival studies that pre-date DICOM (e.g. proprietary Elscint, Siemens and GE image formats). The vendors are continuously evolving their implementation of DICOM, so you need to continuously upgrade your DICOM software. If you have issues, file an issue on github   https://github.com/rordenlab/dcm2niix/issues
Since I do not have access to PET images, we are necessarily dependent on samples from the community.