A new EPI-based dynamic field mapping method: application to retrospective geometrical distortion corrections

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Image Processing
• Medical Physics

Background:

• Geometrical distortions are a common artifact in echo-planar imaging (EPI).
• Accurate geometrical representation is crucial for functional MRI (fMRI) analysis.
• Existing methods for distortion correction may have limitations in dynamic imaging scenarios.

Purpose of the Study:

• To introduce a novel dynamic field mapping method for retrospective correction of geometrical distortions.
• The method is specifically designed for dynamic single-shot gradient-echo echo-planar imaging (GRE-EPI).

Main Methods:

• The proposed method requires only one additional volume acquisition.
• It extracts a field map from each phase volume by assuming temporal invariance of the echo time-independent phase component.
• Performance was evaluated through three experimental sets: prerequisite/modeling validation, time-dependent distortion assessment, and comparison with existing methods.

Main Results:

• The study validates the underlying modeling procedure.
• The dynamic field mapping method demonstrated superior noise robustness compared to other evaluated techniques.
• A theoretical framework explaining the noise performance is provided.

Conclusions:

• The developed dynamic field mapping method effectively corrects geometrical distortions in GRE-EPI.
• With a minimal increase in acquisition time, it is highly suitable for functional MRI.
• The method offers a promising direction for future prospective applications in fMRI.