Rapid parameter estimation for selective inversion recovery myelin imaging using an open-source Julia toolkit.

Area of Science:

• Neuroimaging
• Biophysics
• Computational Science

Background:

• Magnetic resonance imaging (MRI) is crucial for quantifying myelin content.
• Computational bottlenecks hinder the clinical application of advanced MRI techniques.
• Quantitative magnetization transfer (qMT) imaging, particularly selective inversion recovery (SIR), offers myelin-sensitive parameter estimation.

Purpose of the Study:

• To develop a computationally efficient, open-source toolkit for processing SIR MRI data.
• To enable rapid and accurate estimation of myelin-sensitive parameters like the macromolecular pool size ratio (PSR).
• To overcome computational limitations and improve the clinical feasibility of SIR imaging.

Main Methods:

• Developed an open-source toolkit in the Julia language for SIR data processing.
• Simulated SIR images with varying PSR and R1f values, adding Rician noise for accuracy testing.
• Validated the toolkit using cross-linked bovine serum albumin (BSA) phantoms and a human brain MRI scan at 3T.

Main Results:

• The Julia toolkit accurately estimated SIR parameters from simulated data (Lin's concordance correlation coefficient > 0.98).
• Parameter estimates from BSA phantoms and in vivo brain scans were consistent with published MATLAB-based methods.
• Achieved an approximate 20-fold reduction in computational time compared to previous MATLAB implementations.

Conclusions:

• A fast, open-source toolkit for SIR MRI has been developed using Julia.
• The toolkit provides rapid and accurate estimation of SIR parameters.
• Reduced computational cost enhances the accessibility of SIR parameters for clinical applications.