Regularized joint water-fat separation with B₀ map estimation in image space for 2D-navigated interleaved EPI based...

Area of Science:

• Magnetic Resonance Imaging (MRI)
• Medical Physics
• Biomedical Engineering

Background:

• Conventional water-fat separation in MRI struggles with echo-planar imaging (EPI) due to multiple fat resonances.
• Fat signal interference is a significant challenge in diffusion-weighted imaging (DWI).

Purpose of the Study:

• Develop and validate a novel water-fat separation and B0 estimation algorithm for EPI.
• Improve fat suppression, particularly for DWI applications with confounding fat signals.

Main Methods:

• Proposed a multi-peak water-fat separation algorithm with a suitable model approximation for EPI, accounting for B0 and fat off-resonance effects.
• Validated the algorithm using simulations and in vivo spin-echo EPI in both B0-homogeneous (leg) and B0-inhomogeneous (head-neck) regions.
• Compared performance against voxel-independent separation and fat saturation techniques.

Main Results:

• The proposed algorithm demonstrated improved fat suppression compared to conventional methods, especially in B0-inhomogeneous areas.
• Addressed water-fat ambiguities observed with reference algorithms in simulations and in vivo.
• Achieved superior fat suppression for DWI data and apparent diffusion coefficient (ADC) maps.

Conclusions:

• The developed algorithm effectively suppresses multi-peak fat components in multi-shot interleaved EPI.
• Offers superior performance over conventional fat saturation and separation algorithms for EPI applications.