Quantitative parametric images of [11C]acetate and [11C]-meta- hydroxyephedrine kinetics in the myocardium

Parametric images, showing fully quantitative parameters at the voxel level, can be used to quantify various processes with maximal spatial resolution. Parametric imaging has been validated for myocardial blood flow imaging, but still needs to be extended to other tracers and kinetic models. The aim of this study was to validate parametric images of the reversible one tissue (1T2k) model for [C]acetate and both 1T2k and the reversible two tissue (2T4k) model for [C]HED. Methods: Dynamic [C]acetate and [C]HED scans were acquired for 10 and 22 patients, respectively, using a Gemini TF PET/CT scanner. Parametric images of [C]acetate kinetics were derived using a basis function method (BFM) implementation of the single tissue compartment model. For [C]HED BFM implementations of both single and two tissue compartment models were used. Quantitative accuracy of BFM was assessed by comparing average regional parametric values with those obtained using standard non-linear regression (NLR) analysis applied to regional tissue time-activity curves. Simulations were performed to study effects of noise on both NLR and BFM. [C]acetate scans were analysed twice by three observers using two different software packages, of which one incorporates parametric image information during heart segment definition. Regional interand intraobserver variability was assessed for both software packages. Results: Both interand intraobserver variability was excellent for software incorporating parametric image information (ICC > 0.91). Quantitative accuracy of BFM was high (ICC > 0.90) for all assessed parameters although a small but significant difference (between -3 and 6%) was observed which could be addressed to heterogeneity within individual segments. In addition, BFM was equally or less sensitive to noise as compared to NLR. Conclusion: Parametric images can be generated using a basis function implementation of both single and two tissue compartment models. Obtained images were of diagnostic quality, quantitatively accurate and could be analysed with high reproducibility.