Myelin water imaging from multi-echo T₂ MR relaxometry data using a joint sparsity constraint.

Area of Science:

• Biomedical Imaging
• Neuroscience
• Medical Physics

Background:

• Multiple sclerosis (MS) is characterized by demyelination, a process quantifiable via myelin water fraction (MWF) using magnetic resonance imaging (MRI).
• Current MWF imaging methods suffer from long computation times and high noise sensitivity, limiting clinical translation.

Purpose of the Study:

• To introduce a novel, efficient, and noise-robust method for determining MWF.
• To overcome the limitations of existing MWF quantification techniques for clinical application.

Main Methods:

• Developed a new method utilizing a joint sparsity constraint and a pre-computed B1+-T2 dictionary.
• Employed a single component analysis for B1+ map estimation, followed by T2 distribution determination using non-negativity and joint sparsity constraints.
• Implemented the Sparsity Promoting Iterative Joint NNLS (SPIJN) algorithm, achieving a 50-fold reduction in computation time compared to traditional methods.

Main Results:

• Simulations showed a significant decrease in absolute MWF error (0.013 vs. 0.031) at SNR=250 compared to the regularized NNLS algorithm.
• In vivo validation in healthy subjects demonstrated improved MWF quantification, particularly in frontal white matter, with reduced inter-subject variability (max std dev 0.0193 vs. 0.0266).
• Results were consistent with existing literature values.

Conclusions:

• The proposed SPIJN method offers a computationally efficient and noise-robust approach for MWF estimation.
• These advancements represent a significant step towards the clinical implementation of MWF measurements for MS assessment.