Faster and More Robust CK Reaction Rate Estimation at 3T Using Acquisition-Weighted ³¹P Cardiac 1D-MRSI With...

Quantification of the creatine kinase (CK) forward reaction rate (k_f) in the human heart using phosphorus magnetic resonance spectroscopy is clinically important; however, it is limited by long acquisition times, operator subjectivity in analysis, and potential skeletal muscle contamination. This study evaluates if combining compartment-based reconstruction techniques with acquisition-weighted (AW) Triple Repetition Time Saturation Transfer (TRiST) acquisitions could overcome these challenges.

Healthy volunteers were scanned with a fully weighted (FW) TRiST protocol twice, and once with an AW TRiST protocol on a 3T MRI. The resulting spectra were reconstructed with conventional Fourier Transform (FT), as well as compartment-based reconstruction techniques: Spectroscopy with Linear Algebra Modeling (SLAM), Spectral Localization by IMaging (SLIM), and an unweighted mean of the FT spectra (ROI-FT). k_f values were calculated and compared across reconstruction methods and acquisition types.

The cardiac k_f values from FW TRiST were 0.21 ± 0.07 s^-1 (FT), 0.26 ± 0.08 s^-1 (SLAM), 0.26 ± 0.07 s^-1 (SLIM), and 0.30 ± 0.10s^-1 (ROI-FT). Corresponding values from AW TRiST were 0.27 ± 0.07 s^-1, 0.25 ± 0.05 s^-1, 0.25 ± 0.04 s^-1, and 0.24 ± 0.08 s^-1, respectively. No significant differences were observed between FW and AW results. A significant decrease in cardiac PCr/ATP ratios was observed for SLAM and SLIM reconstructed data, suggesting decreased signal contamination from skeletal muscle.

Compartment-based reconstruction techniques minimize the operator subjectivity present in the current FT method of analyzing TRiST experiments, in addition to reducing skeletal muscle contamination. When combined with an AW acquisition, scan times were reduced by 47% without compromising k_f accuracy. This method provides a more robust and efficient evaluation of in vivo cardiac metabolism.