DTI Phillips data

Submitted by Patrick Bedard on

Hello TORTOISE team,

Hope all is well!!

I’m trying to analyze some DTI collected at the NIH's Radiology on a Philips scanner.

I’m a bit confused about the, early, results…

after running ImportDICOM the file with the bvals (bmtxt) is empty and after dcm2niix_afni, the file with the bvals is 1 short compared to the DWI, which has 16 volumes, although dcm2niix_afni also creates a file with 17 volumes and I think they add the ADC measure but I didn’t get that with the DTI I processed for another protocol collected on a GE machine at the NIH's NMRF.


After that I tried DIFFPREP and I got his error:


Denoising DWIs...

Done denoising the DWIs.....

Gibbs ringing correction of volume: 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,


Changing the FoV to the working FoV... Please hold on...

Such a bvalue does not exist in the data... Exiting...



Is there anything special with the Phillips DTI data I should take into consideration?





Submitted by nayaka on Fri, 05/17/2019 - 12:33


Hi Patrick,

As far as my experience with Philips data, I know that there is an additional volume that is added in the data. This additional volume is the average b0 volume. This might be the extra volume you are taking about in the import. Can you please check the bvals file to see if there is an additional b0 added in i.e if the bvals correspond to the bvals you are expecting for your data? 

In addition, I would open your 4d nifti file in say, AFNI and visualize the  volumes. If you do see the extra averages B0 at the end, you can remove it and then try importing the nifti with the bvals and bvecs into TORTOISE. 

Hopefully this will solve the issue.