Phase-uncertainty quality map for two-point Dixon fat-water separation.

Area of Science:

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

Background:

• Dixon fat-water separation is crucial for MRI, but phase errors can degrade image quality.
• Accurate phase correction and region-growing segmentation are essential for reliable fat-water separation.

Purpose of the Study:

• To compare two phase correction algorithms and two quality maps (QM) for region-growing in Dixon imaging.
• To evaluate the impact of pre-processing to remove linear phase drifts on fat-water separation.

Main Methods:

• Acquisition of in-phase and out-of-phase images using a spoiled dual-gradient-echo sequence at 1.5 T.
• Processing of 97 datasets with original and hierarchical phase correction algorithms and two QMs (phase gradient-based and phase uncertainty-based).
• Evaluation of pre-processing to remove linear phase drifts associated with dual-echo acquisitions.

Main Results:

• The hierarchical phase correction algorithm demonstrated superior performance compared to the original algorithm across all applications.
• The proposed phase uncertainty QM offered a minor improvement in clinical images but showed vulnerability to flow-related phase shifts.
• Pre-processing to remove linear phase shifts had minimal impact on the overall outcome.

Conclusions:

• Hierarchical phase correction is a more effective technique for Dixon fat-water separation.
• The combination of phase uncertainty QMs and hierarchical algorithms, with appropriate pre-processing, represents the most successful approach for phase correction in Dixon imaging.