Dear Gong,

Thank you very much for your so fast reply! I will try your advice.

I don’t know the readout time. But is it could be acquired from dicom files? Could you tell me how to acquire that?

I’m afraid I don’t have any experience of GE scanners, so I am not sure where/how to find it for your case. Maybe someone who has a GE scanner can chip in?

If not, just use 0.05.

And in my situation, is that alright make each value equals 1 in index.txt? Because the “data.nii" only contains 67 images, which means L>>R b0 images didn’t in it.

It depends a little. If those three b=0 scans were acquired as part of the diffusion protocol it might still be an advantage with more than just ones as it also indexes the topup output which has information about subject movement. Have a look at

https://fsl.fmrib.ox.ac.uk/fsl/fslwiki/EDDY/Faq#Why_do_I_need_more_than_two_rows_in_my_--acqp_file.3F

and hopefully it will be clear.

Good luck Jesper

Thanks again!

在 2014年3月12日，14:54，Jesper Andersson <[log in to unmask]> 写道：

Dear Gong,

{

0 -1 0 1

0 -1 0 1

0 -1 0 1

1 0 0 1

1 0 0 1

1 0 0 1

}

the problem here is the fourth column. It implies that the total readout time is 1 second, which is “unrealistic”. Currently topup doesn’t check for this, and therefore allows it, whereas eddy checks for it. I’ll make sure that they agree for future releases.

Your solution is to put a realistic readout time in the fourth column, rerun topup and then run eddy. If you know the readout time, that is the best. Otherwise pick for example ~0.05 (allowed range is 0.01–0.2).

Good luck Jesper

. And then using “topup --imain=A2P_P2A_b0 --datain=acqparams.txt --config=b02b0.cnf --out=my_topup_results --iout=my_hifi_b0” to created Hi-Fi b0 images.

Everything went well hereto. By after doing

“ fslmaths my_hifi_b0 -Tmean my_hifi_b0 “ 

“bet my_hifi_b0 my_hifi_b0_brain -m” 

and “index=“”;for ((i=1; i<=67; i+=1)); do indx="$indx 1"; done;echo $indx > index.txt” 

,I went to do eddy. But it didn’t work, error message was “msg=AcqPara::AcqPara: Unrealistic read-out time”.

I am wondering is that because our data is separated to two sequences? So I merged two dataset using fsl_merge. So I got a 4D data with 70 (67+3) images. After changed the bvals and bvecs (add 0 at the end of each row) I tried EDDY again. But it still didn’t work. Error message was “Image Exception : #22 :: ERROR: Could not open image data”.

I have no idea what to do next, could you give me some advise about how to do EDDY in our condition (b0s in two different sequence)? And how to make index.txt in this situation.

Thanks very much!

—————————————————

Sincerely，
Jinnan Gong

Best wishes for you!
TEL：＋86 18680807330
Key Laboratory for NeuroInformation of Ministry of Education, School of Life Science and Technology, University of Electronic Science and Technology of China

—————————————————

Sincerely，
Jinnan Gong

Best wishes for you!
TEL：＋86 18680807330
Key Laboratory for NeuroInformation of Ministry of Education, School of Life Science and Technology, University of Electronic Science and Technology of China