[Mristudio-users] NRRD format

Hangyi Jiang hjiang at jhmi.edu
Fri Nov 21 12:46:49 EST 2008


hi, 

first of all, DTI Studio does accept multiple b-values dataset. to do this, you can define the gradients in the gradient table by this format: 
#: x, y, z, b-value

e.g.:

0: 1, 0, 0, 600
1: 0, 1, 0, 600
...
8: 1, 0, 0, 800
9: 0, 1, 0, 800
...


"DTI Mapping" does not accept Analyze DWIs, for NRRD, the header should follow the NRRD-DWMRI format, not the NRRD format, they are not same. for RAW format, it is supposed that each file contains only one image volume, so that you need to put all image files into one folder.

my suggestion is that you'd better to put all image volumes into one single file (Philps REC format). before you do this,  as Susumu said, please make sure that your raw data type is in 16-bit unsigned integer (word). you can put all multiple b-value dataset into one file, or seperate them into multiple single_b-value dataset. and then process the dataset as Philips REC files or pre-edit a ".dpf" file and let program to get the image parameters from the ".dpf" file....

regards,

hangyi



>>> "[Seongjin]" <choisj70 at gmail.com> 11/21/08 9:25 AM >>>
Dear DTI studio users,

I am trying to break former multi-b-value DTI data into each data set of a
single b-value.
I saved with analyze format but it is not allowed in the DTI calculation on
DTI studio.
So I tried RAW format but it gave out a message "Number of files in this
fold is less than expected: (Img_Slices*Img_Blocks)
I retried with NRRD fromat but it did work neigther with an error message
"NRRD-DWMRI header error: not a valid NRRD-DWMRI file".
Is there any instruction on this matters? Is there anyone who has suggestion
on this case?

All the best,

-SC



More information about the Mristudio-users mailing list