Accuracy and repeatability of joint sparsity multi-component estimation in MR Fingerprinting

Area of Science:

• Medical Imaging
• Quantitative MRI
• Biophysics

Background:

• Magnetic Resonance Fingerprinting (MRF) enables quantitative tissue characterization.
• Traditional voxel-wise MRF assumes single tissue types, limiting accuracy.
• Partial volume effects, where voxels contain multiple tissue types, pose a challenge for accurate quantification.

Purpose of the Study:

• To evaluate the accuracy and repeatability of the Sparsity Promoting Iterative Joint Non-negative least squares Multi-Component MRF (SPIJN-MRF) method.
• To assess SPIJN-MRF's capability for tissue parameter estimation and partial volume segmentation.
• To compare SPIJN-MRF's performance against conventional methods like SPM12 and FSL.

Main Methods:

• Numerical simulations using BrainWeb phantoms.
• In vivo MRF data acquisition from 5 subjects over 8 weeks.
• Comparison of SPIJN-MRF partial volume segmentations with SPM12 and FSL.

Main Results:

• SPIJN-MRF demonstrated superior accuracy in simulations, achieving Fuzzy Tanimoto Coefficients (FTC) > 0.95 compared to 0.5-0.7 for SPM12/FSL.
• In vivo data showed SPIJN-MRF relaxation times consistent with literature and minimal variation.
• SPIJN-MRF yielded low coefficients of variation (CoV) for myelin water (10.5%), white matter (6.0%), gray matter (5.6%), CSF (4.6%), and total brain volume (1.1%).

Conclusions:

• SPIJN-MRF provides accurate and precise tissue relaxation parameter estimation, effectively handling partial volume effects.
• The method generates tissue fraction maps, including myelin water, valuable for white matter disease evaluation.
• SPIJN-MRF offers a robust approach for quantitative tissue analysis in MRI.