FSL eddy_correct or eddy? ENIGMA dMRI preprocessing protocole

Hello,

I have a question regarding eddy_correct and eddy.

I am currently preprocessing some DTI data following the ENIGMA protocole method. I am using the FSL command’s eddy_correct and not the more recent eddy because I have quite a lot of subjects with only 6 gradients and 7 volumes.

I have trouble understanding the difference between the two. It seems like eddy is more suited for data with lots of volumes?

There has also been a lot of litterature mentioning eddy as the better method of the two.

I wanted to know if someone could provide me some insight on the topic. Thank you!

Hi,

you are correct that you can’t easily use eddy on data with only 6 gradient directions (we usually say that you need at least 20).
One option for you would be to use eddy_correct, though you need to be careful when doing that. My experience of eddy_correct is that if you have data with not very strong eddy currents (for example by using a twice refocused sequence) and the subject doesn’t move a lot, then running eddy_correct can introduce distortions/movement that wasn’t there before into the data. So if those conditions are true for your data, you might decide to not do any distortion/movement correction at all.

Best regards Jesper

I understand. Thank you!

A