High-resolution deuterium metabolic imaging of the human brain at 9.4 T using phase-cycled balanced SSFP spectral-spatial acquisitions

Purpose: The aim was to improve the sensitivity and robustness against B inhomogeneities of deuterium metabolic imaging (DMI) using phase-cycled balanced SSFP (bSSFP) methods at 9.4 T. Methods: We investigated two variants of phase-cycled bSSFP acquisitions, namely uniformly weighted multi-echo and acquisition-weighted chemical shift imaging (CSI) to improve the SNR of DMI in the brain after oral [6,6'-H]-glucose intake. Phase-cycling was introduced to reduce the off-resonance sensitivity of bSSFP, incurring a moderate SNR loss. Two SNR optimal methods for obtaining metabolite amplitudes from the phase-cycled bSSFP data were proposed. The SNR performance of the two bSSFP variants was compared with the SNR-optimized vendor's standard CSI. Additionally, in vivo T and T of deuterium metabolites were estimated. Results: High-resolution whole-brain dynamic DMI maps were obtained for all acquisitions. The CSI variant of phase-cycled bSSFP achieved an average SNR increase of 16% and 25% for glucose and glutamate + glutamine (Glx), respectively, compared to the SNR-optimized vendor's standard CSI. Phase-cycling improved the bSSFP metabolite estimation and provided additional spectral encoding at the cost of a 10% to 20% SNR loss. Compared to the CSI variant of bSSFP acquisition, the multi-echo variant exhibited up to 35% lower SNR performance because of uniform k-space weighting and less efficient readout. However, it achieved higher resolutions than acquisition-weighted CSI protocols and showed several qualitative improvements. Conclusion: We demonstrated the feasibility of using two phase-cycled bSSFP acquisitions for off-resonance insensitive high-resolution [6,6'-H]-glucose DMI studies in the human brain. bSSFP acquisitions have potential to improve the sensitivity of DMI despite the SNR loss of phase-cycling and other human scanner constraints.