Distortion correction of functional MRI without reverse phase encoding scans or field maps.

Area of Science:

• Neuroimaging
• Medical Physics
• Biomedical Engineering

Background:

• Functional magnetic resonance images (fMRI) often have spatial distortions from magnetic susceptibility.
• These distortions can mismatch fMRI with structural images, impacting brain function analysis.
• Current correction methods require additional scans (field maps or blip-up/blip-down), which are not always available.

Purpose of the Study:

• To develop a distortion correction method for fMRI data using only acquired functional scans and a structural image.
• To enable processing of historical or limited fMRI datasets lacking specific distortion correction sequences.
• To provide a tool for accurate geometric alignment of fMRI data with structural images.

Main Methods:

• Synthesized an undistorted image with fMRI-like contrast from acquired functional data.
• Used the synthetic image as an anatomical target for distortion correction.
• Evaluated the SynBOLD-DisCo (Synthetic BOLD contrast for Distortion Correction) method.

Main Results:

• SynBOLD-DisCo produced fMRI data geometrically similar to undistorted structural images.
• The correction accuracy was comparable to methods using blip-up/blip-down acquisitions.
• The method successfully corrected distortions in fMRI data without additional specialized scans.

Conclusions:

• SynBOLD-DisCo offers an effective solution for fMRI distortion correction with limited data.
• The method facilitates the analysis of historical and retrospectively processed fMRI datasets.
• The developed tool is available for integration into standard fMRI preprocessing pipelines.