Improving parametric estimation in the brain from multispin-echo sequences using a fusion bootstrap moves solver

Area of Science:

• Magnetic Resonance Imaging
• Neuroimaging
• Quantitative MRI

Background:

• T2 mapping provides critical data for assessing brain degenerative diseases.
• Multispin-echo (MSE) sequences are standard for T2 mapping, with dictionary matching enhancing accuracy.
• Accurate T1 quantification necessitates knowledge of the transmit B1+ field.

Purpose of the Study:

• To develop a method for simultaneous estimation of the brain's B1+ field and T2 values using MSE sequences.
• To investigate the efficacy of the fusion bootstrap moves solver (FBMS) for improving B1+ accuracy and spatial smoothness.
• To compare different spatial regularization levels for B1+ and T2 estimation.

Main Methods:

• Utilized multispin-echo (MSE) sequences with dictionary matching for signal analysis.
• Applied the fusion bootstrap moves solver (FBMS), an iterative method considering pixel neighborhoods, for B1+ estimation.
• Employed total variation regularization for simultaneous B1+ and T2 estimation as a proof-of-concept.

Main Results:

• Achieved significant improvements in B1+ accuracy, with error reductions up to 45.37% in simulations.
• Demonstrated recovery of spatially homogeneous B1+ maps in in vivo 3.0T brain data, with a 16.81% decrease in B1+ error.
• Validated the effectiveness of FBMS in enhancing B1+ map quality.

Conclusions:

• Simultaneous B1+ and T2 mapping is feasible using readily available MSE sequences.
• Echo-modulation curve matching combined with FBMS offers an effective and cost-efficient approach for accurate B1+ estimation.
• The developed method enhances quantitative MRI of the brain.