Free-breathing liver fat and quantification using motion-corrected averaging based on a nonlocal means algorithm

Area of Science:

• Magnetic Resonance Imaging
• Medical Physics
• Quantitative Imaging

Background:

• Accurate liver fat quantification is crucial for diagnosing and monitoring liver diseases.
• Traditional MRI methods can be affected by patient motion, leading to inaccurate measurements.
• Proton density fat fraction (PDFF) and $$ are key biomarkers for liver steatosis and fibrosis.

Purpose of the Study:

• To develop and validate a motion-robust chemical shift-encoded (CSE) MRI technique for accurate liver PDFF and $$ quantification.
• To enhance signal-to-noise ratio (SNR) in free-breathing acquisitions.
• To compare the proposed method against existing 2D and 3D techniques.

Main Methods:

• A free-breathing, multi-repetition 2D CSE acquisition with motion-corrected averaging using nonlocal means (NLM) was employed.
• The 2D CSE-NLM method was evaluated in a digital phantom and compared to direct averaging and single acquisition techniques.
• Patient data was used to compare 2D CSE-NLM against 3D breath-hold, free-breathing, and navigated techniques, alongside reader studies and quantitative analysis.

Main Results:

• Simulations showed 2D CSE-NLM had lower standard deviations for PDFF and $$ compared to direct averaging and 2D 1ave.
• In patients, 2D CSE-NLM demonstrated fewer motion artifacts and higher SNR than alternative methods.
• Quantitative analysis revealed comparable PDFF and $$ measurement variability to 2D direct averaging, with reduced bias in the presence of motion.

Conclusions:

• The proposed 2D CSE-NLM technique provides motion-robust and accurate quantification of liver PDFF and $$.
• This method enables reliable liver fat and fibrosis assessment during free-breathing MRI.
• 2D CSE-NLM offers an improved alternative for clinical liver imaging.