Multiecho water-fat separation and simultaneous R2* estimation with multifrequency fat spectrum modeling

Area of Science:

• Medical Imaging
• Magnetic Resonance Imaging
• Biophysics

Background:

• Multiecho chemical shift-based water-fat separation is crucial for clinical MRI, correcting field inhomogeneities.
• Existing methods use a simplified fat signal model, leading to incomplete separation and inaccurate T(2) (*) estimation.
• Fat possesses a complex multifrequency spectral signature not accounted for in simpler models.

Purpose of the Study:

• To improve water-fat separation accuracy in MRI.
• To enhance simultaneous T(2) (*) estimation in fat-containing tissues.
• To integrate a multifrequency fat model into the IDEAL technique.

Main Methods:

• Incorporated a multifrequency fat signal model into iterative decomposition of water and fat with echo asymmetry and least-squares estimation (IDEAL).
• Explored a priori fat spectrum measurement via MR spectroscopy.
• Developed novel self-calibration algorithms for in-situ fat spectrum estimation.

Main Results:

• Achieved more accurate water-fat separation compared to single-frequency models.
• Demonstrated improved T(2) (*) decay estimation in tissues with fat.
• Validated the enhanced method across diverse in vivo scans (knee, ankle, spine, breast, abdomen).

Conclusions:

• The multifrequency fat model significantly enhances IDEAL water-fat separation and T(2) (*) estimation.
• This advancement offers improved diagnostic capabilities in various clinical MRI applications.
• The method's flexibility in fat spectrum determination (a priori or self-calibrated) increases its clinical utility.