>I would not chose that order though, I would put all P->A followed by all A->P.
What is the advantage? We have been using topup and recorded 4 sets of images P>>A, A>>P, A>>P and P>>A. We did feed that order into topup and were always happy with the results.
So I was wondering why you'd recommend to put all A>>P first and then the P>>A sets (or the other way around).
Dear Lulu,
I was using eddy to pre-process my diffusion data. We collected our data according to the sequence below. But I am not sure if I have processed it right.
Our whole data has 60 orientations, but when scanning we divided them into 19/20/21 orientations and each with one b0 image (so there are three b0 images). The phase encoding direction of all these 63
files are P>>A.
In order for analysis in Topup, we collected 3 A>>P b0 images (corresponding to three 19/20/21 orientations respectively) (averages = 2). So we obtained 6 b0 images here.
Do we set the blip scanning correct? Is there something wrong?
According to our way of scanning, so I wrote my --detain = acqparams.txt file as this:
0 1 0 0.102
0 -1 0 0.102
0 -1 0 0.102
0 1 0 0.095
0 -1 0 0.095
0 -1 0 0.095
0 1 0 0.102
0 -1 0 0.102
0 -1 0 0.102
(Six negative one lines represent the blipped A>>P b0 images)
Is this the file you use for topup? This _might_ be corrected if you have put the volumes into your —imain file in the same order. I would not chose that order though, I would put all P->A followed by all A->P. Also I would probably only use one each within
the A->P pairs as they represent pretty identical information.
I also don’t understand why you have different readout times. Did you really acquire the middle 20 volumes with different readout?
The --acqp file is written as follows:
0 1 0 0.102
0 1 0 0.102
0 1 0 0.095
0 1 0 0.095
0 1 0 0.102
0 1 0 0.102
The index file is written as follows:
1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3
Typically you can use the same —acqp file as you used for topup (provided you get it right) and then set the indices appropriately. If you specify your P->A volumes first in your —datain file you can use that with the index file you have.
Good luck Jesper
Are these files and our way of scanning correct?
Could someone give me some suggestions about these?
Thanks so much!
Yours,
Lulu