Serious error: spatial orientation ambiguous (tag 0020,0037 not found) about dcm2niix HOT 2 CLOSED

I believe this is a duplicate of
http://www.nitrc.org/tracker/?func=detail&atid=3335&aid=8118&group_id=880
I think you tried to convert a derived image. In particular the file name "COLFA" probably refers to a derived "COLOR FRACTIONAL ANISOTROPY" image

I think you should update to version 28June2016. If I am correct it will change the warning to the less shrill and more specific:
Unable to determine spatial orientation: 0020,0037 missing (probably not a problem: derived image)
Explanation: I think this is a derived image (FA/ADC maps based on series of DWI scans) rather than a raw image. Be warned that Siemens does not preserve spatial coordinates when computing derived images. This is usually not an issue for scientists: the derived maps are stored with very poor precision (8-bit RGB) and most imaging software (FSL/SPM/AFNI) will not process RGB images. Furthermore, you can derive much nicer FA/ADC maps after you undistorted the raw images with Eddy/Topup. Personally, on our Prisma I open the protocol with the ExamExplorer, go to the "Diff" tab and uncheck the "ADC maps" and "FA maps" check boxes so the console does not generate derived images.

Despite my clarified note to the user, I do still think this is a serious error. I do suggest you follow my lead and encourage Siemens to preserve the spatial encoding when generating derived images. No DICOM viewer will be able to determine spatial orientation without this. For example, if you drop one of these images on a tool popular with clinicians such as Osirix, they are not shown any labels for spatial orientation, so the user will have to know which is the left and right side of the brain.

from dcm2niix.

THANK YOU for the detailed explanation

from dcm2niix.