Related Experiment Video
Updated: Jan 3, 2026
07:13
Author Spotlight: An Efficient and Robust Software for Automated Fusion of Multiple Preclinical Imaging Modalities
Published on: October 27, 2023
1.6K
Improved cortical boundary registration for locally distorted fMRI scans
Tim van Mourik1, Peter J Koopmans2, David G Norris1,2
1Radboud University Nijmegen, Donders Institute for Brain, Cognition and Behaviour, Nijmegen, The Netherlands.
Plos One
|November 19, 2019
Summary
Recursive Boundary Registration (RBR) corrects geometric distortions in echo-planar imaging (EPI) for high-resolution human brain scans. This method improves spatial specificity for laminar analysis, crucial for advanced functional MRI studies.
Area of Science:
- Neuroimaging
- Medical Physics
- Computational Neuroscience
Background:
- Advances in MRI enable submillimetre resolution in human functional imaging, facilitating detailed analyses like cortical layer investigation.
- Higher static field strengths in MRI increase spatial resolution but also exacerbate geometric distortions inherent in echo-planar imaging (EPI).
- These EPI distortions, potentially several millimetres, pose a significant challenge for precise laminar analysis in the human brain.
Purpose of the Study:
- To present a novel method, Recursive Boundary Registration (RBR), for correcting geometric distortions between anatomical and EPI volumes in human fMRI.
- To ensure topological integrity of the cortical surface during distortion correction, a prerequisite for common laminar analysis workflows.
- To automate non-linear distortion correction for routine human laminar fMRI, particularly for large field-of-view acquisitions.
Main Methods:
- Recursive Boundary Registration (RBR) applies Boundary Based Registration (BBR) iteratively on progressively smaller brain subregions.
- Registration is driven by the contrast between grey and white matter, ensuring accurate alignment.
- The algorithm explicitly preserves the topology of the cortical surface throughout the deformation process.
Main Results:
- RBR achieves submillimetre accuracy when validated against a manually distorted gold standard.
- Application to in vivo human fMRI scans demonstrates a clear enhancement in spatial specificity.
- The RBR method automates non-linear distortion correction, streamlining the fMRI analysis pipeline.
Conclusions:
- RBR effectively corrects EPI geometric distortions, enabling more precise laminar analysis in high-resolution human fMRI.
- The method's automation and accuracy represent a significant advancement towards routine clinical application of laminar fMRI.
- The open-source code and associated functions are available to facilitate further research and development in neuroimaging analysis.
