Dipy free water corrected DTI

Software Support
,
1

Hi everyone,

I’m trying to use DIPY on my diffusion MRI data to do DTI with free water correction. I was wondering if I there was a way to specify which shell should be used for the free water correction and which shell for the tensor fitting when running fwdtifit = fwdtimodel.fit(data) or if the function is selecting automatically? I’d like to use b=500 for the FW correction and b=1000 for the tensor fitting but wasn’t sure if the function was taking multiple arguments. Maybe I’m misunderstanding this and the function has to use all shells for both correction and tensor fitting, or the two things are part of the same process?
I also have b=2000 which should not be used for this so I was also wondering if that could be specified in an argument or if I had to create a dataset without that shell before running the function.
Thank you in advance for any insights!

2

Hello! This model uses data from all of the shells both for free-water estimation and for tensor fitting, which are indeed part of the same process. The function does not have an argument to avoid some of the data, so you would have to create a separate dataset without b=2,000 if you would like to avoid that shell.

For details about the model and the implementation in DIPY, you might want to read our re:science paper: [Re] Optimization of a free water elimination two-compartment model for diffusion tensor imaging