Macromolecule modelling for improved metabolite quantification using short echo time brain ¹ H MRS at 3 T and 7 T:...

Area of Science:

• Neuroimaging
• Magnetic Resonance Spectroscopy
• Biomarker Discovery

Background:

• Accurate metabolite quantification in brain ¹ H MRS is crucial for clinical applications.
• Macromolecular (MM) signals can interfere with metabolite quantification, especially at higher field strengths like 7 T.
• Existing methods for handling MM signals have limitations in reliability and accuracy.

Approach:

• A novel parametrized macromolecules quantification model (PRaMM) was developed for brain ¹ H MRS.
• PRaMM utilizes ratios of macromolecule peak intensities as soft constraints for improved quantification.
• The model was validated against single component MM models and MM subtraction techniques across different brain regions and field strengths (3 T and 7 T).

Key Points:

• The PRaMM model demonstrated superior performance compared to existing methods in all investigated brain regions.
• Fit quality metrics were significantly improved with the PRaMM model (p≤0.0001).
• Repeatability and reproducibility analyses showed PRaMM is reliable, with minimally detectable changes of 0.5 - 1.9 mM and coefficients of variation <10% for most metabolites.

Conclusions:

• PRaMM offers an improved method for metabolite quantification in brain ¹ H MRS.
• The model enhances the reliability and accuracy of metabolite measurements, particularly important for clinical research.
• PRaMM provides a robust approach to investigate the role of MM background signals from a clinical perspective.