Dear FSL experts,
I am using the FSL topup tool and eddy for distortion correction and noticed significant SNR decrease in comparison to the non-corrected images.
We acquired our data using the following scheme:
90 directions B3k and 12 intermediate B0s; PE A->P
B0 volume #: 1 2 12 22 32 42 52 62 72 82 92 102
8 B0; PE P->A
For the processing I did the following steps:
1. Merge all B0 (12+8) into AP_PA_b0
2. create acqparams.txt
0 1 0 0.062 (12x)
...
0 -1 0 0.62 (8x)
3. topup
4. create index.txt
1 2(10x) 3(10x) 4(10x) 5(10x) 6(10x) 7(10x) 8(10x) 9(10x) 10(10x) 11(10x) 12
4. eddy
5. dtifit
When checking the resulting FA and V1 maps, you can see significant noise in comparison to the same data without topup/eddy, but eddy_correct. Subsequent processing, like bedpostx and probtrackx confirm these results with decreased levels of connectivity scores.
I already checked the AP_PA_b0 for accidentially included dw-images and the eddy index.txt. So far I couldn't find any mistakes. The topup resulting hifi_b0 also looks good.
I tried using just a single pair of B0s (1st B0 of AP and 1st B0 of PA) for topup and eddy without any improvements.
Do I have any mistakes in my processing or do you have any idea what this increasing noise can be caused by?
Thanks for your help.
C Melzer
|