OpenFmriAnalysis

OpenFmriAnalysis corrects echo planar imaging (EPI)-induced spatial distortions between anatomical and functional MRI volumes to enable submillimetre laminar fMRI analysis.

Key Features:

• Recursive Boundary Registration (RBR): Employs RBR, which recursively applies Boundary Based Registration (BBR) across spatial scales to correct non-linear distortions.
• Multi-scale subregion registration: Performs registration on progressively smaller brain subregions to improve local alignment using grey-white matter contrast.
• Preservation of cortical surface topology: Maintains cortical surface topology during registration to support accurate laminar analysis.
• Submillimetre accuracy validation: Validated against a manually distorted gold standard, demonstrating submillimetre spatial accuracy in human in vivo scans.
• Automated non-linear distortion correction: Automates non-linear correction between anatomical and EPI volumes, enabling processing of large field-of-view acquisitions.
• High-field and high-resolution suitability: Targets distortions and specificity challenges associated with higher static field strengths and submillimetre spatial resolutions.
• Integration with existing tools: Implements wrappers for established neuroimaging tools while contributing novel methodological components.
• MATLAB implementation: Provided as a MATLAB implementation for computational workflows.

Scientific Applications:

• Laminar fMRI analysis: Enables layer-specific analyses by preserving cortical topology and aligning high-resolution anatomical and EPI data.
• High-field fMRI distortion correction: Corrects geometrical distortions in EPI acquired at higher static field strengths to improve spatial specificity.
• Submillimetre-resolution cortical investigations: Supports investigation of cortical layers and small brain structures at submillimetre resolutions.
• Preprocessing for large field-of-view acquisitions: Applies automated non-linear correction across large field-of-view functional acquisitions.

Methodology:

RBR recursively applies Boundary Based Registration (BBR) to progressively smaller brain subregions using grey-white matter contrast to correct spatial distortions between anatomical and EPI volumes while preserving cortical surface topology; validation used a manually distorted gold standard to assess submillimetre accuracy.