XCP-D: A robust pipeline for the post-processing of fMRI data

Area of Science:

• Neuroscience
• Neuroimaging
• Data Science

Background:

• Functional neuroimaging relies on pre-processing pipelines for standardized data.
• Post-processing of neuroimaging data lacks standardization, hindering reproducibility.
• Existing tools may not support diverse pre-processing outputs or adhere to data organization standards like BIDS.

Purpose of the Study:

• Introduce XCP-D, a novel post-processing pipeline for functional neuroimaging.
• Address the lack of standardization in fMRI data post-processing.
• Facilitate robust, scalable, and reproducible fMRI data analysis.

Main Methods:

• Developed XCP-D through a collaborative effort using an open development model on GitHub.
• Distributed XCP-D as a Docker container or Apptainer image for accessibility.
• Implemented continuous integration testing for quality assurance.

Main Results:

• XCP-D generates denoised BOLD images and functional derivatives from resting-state fMRI data.
• The pipeline supports NIfTI or CIFTI file formats.
• XCP-D is compatible with data pre-processed using fMRIPrep, HCP, or ABCD-BIDS pipelines.
• Achieved over 5,000 downloads prior to official release, indicating high demand.

Conclusions:

• XCP-D provides a standardized, reproducible solution for fMRI data post-processing.
• Enhances the utility of functional neuroimaging data for diverse research applications.
• Promotes interoperability and data sharing within the neuroscience community.