Hi again,

ok, I read your first message a little more carefully now. 

First of all, I recommend using the first b=0 from your dwi data as the first volume in your input to topup. This is to ensure that your fields is in the same space as our diffusion data.

As for the 65 vs 64 slices you have two options. Either delete your top or bottom slice in all your volumes. Or you can edit the b02b0.cnf file to not attempt any sub-sampling. If you just replace any 2 on the line

--subsamp=2,2,2,2,2,1,1,1,1

with a 1 it usually works, but will take a little longer.

I would recommend deletion of your FOV is such that you have a “little space” either above or below the brain. Or if the FOV is a tight fit you can do the editing of b02b0.cnf instead.

Jesper


On 10 Feb 2017, at 13:13, Virginie Petel Légaré <[log in to unmask]> wrote:

Hello Jesper, 


When I type fslinfo AP_b0 (with one slice deleted, the last one), here is what I get.
fslinfo AP_b0
data_type      INT16
dim1           96
dim2           96
dim3           64
dim4           1
datatype       4
pixdim1        2.000000
pixdim2        2.000000
pixdim3        1.999997
pixdim4        9.300000
cal_max        0.0000
cal_min        0.0000
file_type      NIFTI-1+

fslinfo PA_b0
data_type      INT16
dim1           96
dim2           96
dim3           64
dim4           1
datatype       4
pixdim1        2.000000
pixdim2        2.000000
pixdim3        1.999997
pixdim4        9.300000
cal_max        0.0000
cal_min        0.0000
file_type      NIFTI-1+




If I type fsl info for my original DWI here is what I get  (with 65 slices): 
fslinfo DWI_65
data_type      INT16
dim1           96
dim2           96
dim3           65
dim4           65
datatype       4
pixdim1        2.000000
pixdim2        2.000000
pixdim3        2.000000
pixdim4        9.300000
cal_max        0.0000
cal_min        0.0000
file_type      NIFTI-1+

Should I delete one slice of my original DWI? 


Thank you, 

Virginie

On Feb 10, 2017, at 7:46 AM, Jesper Andersson <[log in to unmask]> wrote:

Dear Virgine,

what do you get if you type

fslinfo AP_b0
fslinfo PA_b0

?

Jesper


On 9 Feb 2017, at 18:55, Virginie Petel Légaré <[log in to unmask]> wrote:

Hello, 
I am new to FSL, and I am trying to run topup.
I have one image A>>P (65 slices) and one image A<<P (65 slices), as well as my DWI which has 1 b0  and 64 directions.

At first a message of  "Subsampling levels incompatible with image data” appeared, and I found that it was probably because I had an uneven number of slices (65), and I therefore deleted the last slices in all of my images as described here: https://www.jiscmail.ac.uk/cgi-bin/webadmin?A2=fsl;67dcb45c.1209

However, my DWI is now incompatible with both b0 images. Extracting the first image of my DWI works but I cannot merge by A_P and P_A image with them. I tried deleting the last slice of my DWI also but it does not seem to work.


fslroi DWI_65 DWI_b0 0 1
fslmerge -t b0 DWI_b0 AP_b0 PA_b0
Invconsistent orientations for individual images when attempting to merge.
          Merge will use voxel-based orientation which is probably incorrect - *PLEASE CHECK*!


Thank you.

Virginie