Hello Jesper,

Yes, all b=700 volumes are collected in AP direction while all b=2000 volumes are collected in PA direction.

Thank you!
Mahmoud

On Mon, Aug 7, 2017 at 6:40 AM, Jesper Andersson <[log in to unmask]> wrote:
Dear Mahmoud,

do you mean that all b=700 are acquired with one PE direction and all the b=2000 with that opposite direction?

If so, I have never worked with data of that exact kind so you might need some trial and error to work out which is the better option. My guess, and hope, is that option 2 would be the best.

Jesper

> On 3 Aug 2017, at 18:30, Mahmoud <[log in to unmask]> wrote:
>
> Hello All,
>
> I have dMRI data collected on a Siemens Prisma machine as follows:
>
> Two shells (b=700 and 2000) with opposite PE directions (AP and PA respectively) and a different number of diffusion gradients (89 and 90 respectively) but the number of b0 volumes are the same for both shells (7 b0 volumes collected at the beginning of each shell).
>
> I wonder which of the following ways you recommend to correct for EC, Susceptibility, and motion correction:
>
> 1-
> a) using "topup" to estimate the susceptibility map.
> b) use "eddy" with the first shell only as the input (--imain=b700).
> c) use "eddy" with the second shell only as the input (--imain=b2000).
> d) use "fslmerge" to merge the eddy corrected outputs of steps b and c.
> e) using "paste" to merge the bval and rotated bvec files from steps b and c into one bval and bvec file.
>
> or
>
> 2)
> a) using "topup" to estimate the susceptibility map.
> b) use "fslmerge" to merge the two shells as b700_2000.
> c) using "paste" to merge the bval and bvec files of the two shells into one bval and bvec file.
> d) use "eddy" with the merged shells as the input (--imain=b700_2000) and bval and bvec files of the above step.
>
>
> Thank you!
> Mahmoud
>
>