fMRIStroke: A preprocessing pipeline for fMRI Data from Stroke patients

fMRIStoke is a BIDs application that runs on the outputs of fmriprep (www.fmriprep.org) for the preprocessing of task-based and resting-state functional MRI (fMRI) from stroke patients.

Complete information and documentation can be found at https://fmristroke.readthedocs.io/

About

https://github.com/alixlam/fmristroke/blob/main/docs/_static/pipeline.png?raw=true

Motivation

Stroke not only leads to structural damages of gray/white matter in affected patietns, but can also produce remote changes in structurally normal brain areas by a variety of different mechanisms [Siegel2017]. As a result, it is highly recommended, notably by [Siegel2017], to add specific quality checks and strategies to mitigate lesion specific confounds when dealing with stroke data especially when doing FC analysis.

To do this we propose fMRIStroke a functional magnetic resonance imaging (fMRI) data quality checks and preprocessing pipeline tailored to stroke data. It is designed to provide an easily accessible, interface that is robust to variations in scan acquisition protocols and that requires minimal user input, while providing easily interpretable and comprehensive reports. It uses fmriprep (www.fmriprep.org) outputs derivatives to generate new quality checks plots for stroke patients when lesion masks are available and computes new confounds like signals in lesion masks, and ICA based confounds (as proposed in [Yourganov2017]).

Added quality checks:

hemodynamics lagmap using the Rapidtide python tool providing output reports on the hemodynamic lags present bold series.
homotopic connectivity if freesurfer reconstruction was run.
Registration plots with lesion mask

Added confounds:

lesion: signal in lesion mask.
CSF lesion: signal in CSF + lesion combined mask.
ICA_comp: ICA based confounds [Yourganov2017].

Added outputs:

ROI masks in standardized space.
Denoised fMRI: Denoised BOLD series using the provided pipelines.
Functional Connectivity: Connectivity matrix using provided atlases and connectivity measures.

The fMRIStroke pipeline uses a combination of tools from well-known software packages, including ANTs, FreeSurfer, Rapidtide and Nilearn_

Important

This pipeline was designed to run after fmriprep. Any other fMRI preprocessing tools might not provide the required derivatives for fMRIStroke to run properly.

In summary this tool allows you to easily do the following:

Generate preprocessing quality reports specific to stroke patients, with which the user can easily identify outliers.
Receive verbose output concerning the stage of preprocessing for each subject, including meaningful errors.
Automate and parallelize processing steps, which provides a significant speed-up from manual processing or shell-scripted pipelines.

Citation

Citation.

Acknowledgements

This work makes great use of the work by the NiPreps Community. and the work done by rapidtides authors.

References

[Siegel2017] (1,2)

J. S. Siegel, G. L. Shulman, and M. Corbetta, Measuring functional connectivity in stroke: Approaches and considerations, J Cereb Blood Flow Metab, 2017. doi: 10.1177/0271678X17709198.

[Yourganov2017] (1,2)

Yourganov, G., Fridriksson, J., Stark, B., Rorden, C., Removal of artifacts from resting-state fMRI data in stroke. Neuroimage Clin 2017. doi: 10.1016/j.nicl.2017.10.027