A pharmacokinetic model for hyperpolarized ¹³C-pyruvate MRI when using metabolite-specific bSSFP sequences

Metabolite-specific balanced SSFP (MS-bSSFP) sequences are increasingly used in hyperpolarized [1-¹³C]Pyruvate (HP ¹³C) MRI studies as they improve SNR by refocusing the magnetization each TR. Currently, pharmacokinetic models used to fit conversion rate constants, k_PL and k_PB, and rate constant maps do not account for differences in the signal evolution of MS-bSSFP acquisitions.

In this work, a flexible MS-bSSFP model was built that can be used to fit conversion rate constants for these experiments. The model was validated in vivo using paired animal (healthy rat kidneys n = 8, transgenic adenocarcinoma of the mouse prostate n = 3) and human renal cell carcinoma (n = 3) datasets. Gradient echo (GRE) acquisitions were used with a previous GRE model to compare to the results of the proposed GRE-bSSFP model.

Within simulations, the proposed GRE-bSSFP model fits the simulated data well, whereas a GRE model shows bias because of model mismatch. For the in vivo datasets, the estimated conversion rate constants using the proposed GRE-bSSFP model are consistent with a previous GRE model. Jointly fitting the lactate T₂ with k_PL resulted in less precise k_PL estimates.

The proposed GRE-bSSFP model provides a method to estimate conversion rate constants, k_PL and k_PB, for MS-bSSFP HP ¹³C experiments. This model may also be modified and used for other applications, for example, estimating rate constants with other hyperpolarized reagents or multi-echo bSSFP.