b0dist 45

[AlbertoImg]

Dear HCPpipelines team,

Thanks for the pipeline!

I would like to ask about the reasoning about b0dist=45:

HCPpipelines/DiffusionPreprocessing/DiffPreprocPipeline_PreEddy.sh

Line 237 in 6801b5f

b0dist=45 # Minimum distance in volumes between b0s considered for preprocessing

During the preparation for TOPUP, this option collects only 2 b0 volumes per PE from our dataset with 7 b0 volumes per phase encoding direction.
So we were wondering whether to adjust this parameter to 0 so all data is considered (for pediatric datasets for example), or to keep the default value for standardization of the pipeline.

Thanks in advance
Best,
Alberto

[glasserm]

I think you can try something different if you want. I am not entirely sure why this was set this way, as I didn't implement this code.

[mharms]

I believe it was to limit the number of b0's that ended up going into TOPUP (to speed up the TOPUP stage). Nowadays, we typically use the (later added) -select-best-b0 option, in which case only single b0 volume from each polarity is selected to use in TOPUP anyway.

[glasserm]

@takuya-hayashi what was your thought on -select-best-b0? Did you find it was not helpful?

[AlbertoImg]

Hi all,

Thanks for the input so far!

I found this issue related to the version that added "-select-best-b0":
#158

What I undestood, this option would be better than having 7 b0s per PE with some of them potentially corrupted (e.g. by motion).

Then, we will move forward with the current default "-select-best-b0".

In case you have a different opinion here, please let me know.

Thanks again
Best,
Alberto