Parameter optimization for reproducible cardiac ¹ H-MR spectroscopy at 3 Tesla

Area of Science:

• Cardiovascular Magnetic Resonance Imaging
• Biomedical Engineering
• Metabolic Imaging

Background:

• Cardiac proton MR spectroscopy is crucial for assessing myocardial triglyceride content.
• Optimizing data acquisition parameters is essential for reliable and reproducible measurements.
• Variability in signal intensity and phase can affect accuracy.

Purpose of the Study:

• To optimize data acquisition parameters for cardiac proton MR spectroscopy at 3 Tesla (T).
• To evaluate the intra- and intersession reproducibility of myocardial triglyceride content measurements using optimized parameters.

Main Methods:

• Acquisition parameters were optimized in 49 healthy subjects, assessing signal-to-noise ratio (SNR) and metabolite amplitude variance.
• Evaluated local vs. global power optimization, B0 shimming during free-breathing vs. breathholds, trigger delays, respiratory compensation techniques (breathholds, navigator triggering), and water suppression methods (frequency selective, CHESS, MOIST).
• Intra- and intersession reproducibility of myocardial triglyceride content was assessed by repeated measurements.

Main Results:

• Local power optimization increased SNR by 22% (P=0.0002).
• Navigator triggering and breathholds significantly reduced amplitude variation compared to free-breathing (P<0.05).
• CHESS and MOIST water suppression yielded significantly lower residual water signals than frequency selective excitation (P<0.01).
• Optimized protocol achieved intrasession limits of agreement of -0.11% to +0.04% and intersession of -0.15% to +0.9%, with coefficients of variation of 5% and 6.5%, respectively.

Conclusions:

• A reproducible cardiac MR spectroscopy protocol was developed using optimized parameters on a clinical 3T system.
• The protocol effectively minimizes signal intensity and frequency/phase variations.
• This approach enables high reproducibility for myocardial triglyceride content quantification.