questions
58 Subscribers 
questions > dcm2nii-conversion - does it work properly?
Showing 1-2 of 2 posts
Sep 23, 2016 08:09 AM | A. P.
dcm2nii-conversion - does it work properly?
Dear Mr. Rorden,
I have dicom data of different MRI sequences (scout, localizer, t1, topups and diffusion, fMRI/resting state; all measured at a Siemens 3T scanner). For each participant, I have one folder containing all dicom data of these sequences.
When I start the dcm2nii (with command in Bash/FSL), there is a lot of output during the conversion. Referring to some of the output messages, I am not sure if it is working properly or not. I hope it is fine that I attach the file with all output messages - I have stressed the messages which are questionable to me with bold lettering.
Thank you!
Best regards, Annika
I have dicom data of different MRI sequences (scout, localizer, t1, topups and diffusion, fMRI/resting state; all measured at a Siemens 3T scanner). For each participant, I have one folder containing all dicom data of these sequences.
When I start the dcm2nii (with command in Bash/FSL), there is a lot of output during the conversion. Referring to some of the output messages, I am not sure if it is working properly or not. I hope it is fine that I attach the file with all output messages - I have stressed the messages which are questionable to me with bold lettering.
Thank you!
Best regards, Annika
Sep 23, 2016 12:09 PM | Chris Rorden
RE: dcm2nii-conversion - does it work properly?
Annika-
1.) The warnings you highlighted refer to slice order. In particular, Siemens scanners tend to acquire interleaved acquisitions differently depending on whether there are an odd or even number of slices in your volume. I describe the concept here, as well as a simple test to demonstrate the behavior of your sequence:
http://www.mccauslandcenter.sc.edu/crnl/...
2.) While it looks like you do not have any DTI sequences in your set, the other thing I would caution any users exploring DTI is to validate that the b-vectors are recorded in the fashion expected by their software (e.g. FSL expects gradient directions to be reported with respect to image space, not scanner space). I include details (in particular the "dedicated document") here:
https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage#Diffusion_Tensor_Imaging
3.) As a new user, I would suggest you use MRIcroGL/dcm2niix, rather than my older MRIcron/dcm2nii. I am a single individual, and focus my efforts on the newest tools. While I think my old tools are robust and stable, manufacturers re-interpret the DICOM standard over time. In addition, I think the warning messages generated by my newer software tend to be clearer and more concise.
1.) The warnings you highlighted refer to slice order. In particular, Siemens scanners tend to acquire interleaved acquisitions differently depending on whether there are an odd or even number of slices in your volume. I describe the concept here, as well as a simple test to demonstrate the behavior of your sequence:
http://www.mccauslandcenter.sc.edu/crnl/...
2.) While it looks like you do not have any DTI sequences in your set, the other thing I would caution any users exploring DTI is to validate that the b-vectors are recorded in the fashion expected by their software (e.g. FSL expects gradient directions to be reported with respect to image space, not scanner space). I include details (in particular the "dedicated document") here:
https://www.nitrc.org/plugins/mwiki/index.php/dcm2nii:MainPage#Diffusion_Tensor_Imaging
3.) As a new user, I would suggest you use MRIcroGL/dcm2niix, rather than my older MRIcron/dcm2nii. I am a single individual, and focus my efforts on the newest tools. While I think my old tools are robust and stable, manufacturers re-interpret the DICOM standard over time. In addition, I think the warning messages generated by my newer software tend to be clearer and more concise.