Using FMRIPREP with Tedana for echo combination

Hi all,
I want to preprocess a multi-echo dataset using FMRIPREP and combine echoes with Tedana.
Tedana docs recommend the following:

Performing slice timing correction and motion correction before tedana , and

Performing distortion correction, spatial normalization, smoothing, and any rescaling or filtering after tedana .

FMRIPREP by itself runs on the data without any issues, but it completes all preprocessing steps, including those that should be run after Tedana. Is it possible to run only this first part of FMRIPREP, then Tedana, and then run the remaining steps (e.g. spatial normalisation) in FMRIPREP? If so, would that be a good thing to do?

Thanks in advance everybody!

1 Like

Hi @human123,

We have an open PR on fMRIPrep to finish this integration and yield the optimal combination, without multi-echo informed denoising. It’s pending a code review, but you can access it now by using my branch –

You could git clone that code to your local workstation, and then build a copy of fmriprep by entering that directory with cd tedana/ and then running:

docker build -t tedana_fmriprep .

Then you could execute tedana_fmriprep as you would have executed fmriprep. So, for example:

docker run --rm -v $HOME/Desktop/ds000210:/data:ro -v $HOME/Desktop/dockerout:/out \
tedana_fmriprep /data /out/out participant \
-w /out/work/ --fs-license-file /data/license.txt --participant_label sub-01

You can optionally use a T2* driven coregistration using the --t2s-coreg flag.

Hopefully that helps !


1 Like

That’s great, thank you very much!
Could this also be run via singularity?

Yes, you’ll just need to convert the docker image to a singularity image ! Using docker2singularity the code for doing so might look like

docker run \
-v /var/run/docker.sock:/var/run/docker.sock \
-v D:\host\path\where\to\output\singularity\image:/output \
--privileged -t --rm \
singularityware/docker2singularity \

Then you would use the singularity image as described in the fmriprep docs.

Hope that helps ! Please let me know if anything is unclear.

1 Like

Thank you so much for the help!