Multi-echo quantitative susceptibility mapping: how to combine echoes for accuracy and precision at 3 Tesla

Area of Science:

• Medical Imaging
• Biophysics
• Computational Neuroscience

Background:

• Quantitative susceptibility mapping (QSM) is a crucial MRI technique for assessing tissue magnetic susceptibility.
• Optimal combination of multi-echo gradient-recalled echo (GRE) data is essential for accurate QSM.
• Current methods for combining multi-echo phase information lack consensus, impacting QSM reliability.

Purpose of the Study:

• To compare different multi-echo combination strategies for MRI QSM.
• To determine the optimal method for combining multi-echo GRE signal phase information.
• To evaluate combining phase before or after Laplacian-based methods (LBMs).

Main Methods:

• Simulated and acquired multi-echo GRE data from a head phantom and healthy volunteers.
• Designed and implemented five QSM processing pipelines with varying phase combination and LBM application orders.
• Compared pipeline results using visual inspection, statistical analysis, and noise propagation assessment.

Main Results:

• Nonlinear fitting of multi-echo phase before LBMs achieved the highest regional accuracy.
• This method demonstrated the lowest phase noise propagation compared to averaging LBM-processed phase.
• Superior performance was particularly noted in high-susceptibility venous regions.

Conclusions:

• Recommending nonlinear fitting of the multi-echo signal phase prior to LBM application for QSM.
• This approach enhances accuracy and reduces noise in quantitative susceptibility mapping.
• Optimized multi-echo combination is vital for reliable QSM in clinical and research settings.