Factor analysis has been pursued as a means to decompose dynamic cardiac PET images into different tissue types based on their unique temporal signatures to improve quantification of physiological function. In this work, the authors present a novel kinetic model-based (MB) method that includes physiological models of factor relationships within the decomposition process. The physiological accuracy of MB decomposed cardiac PET images is evaluated using simulated and experimental data. Precision of myocardial blood flow (MBF) measurement is also evaluated.
A gamma-variate model was used to describe the transport of in arterial blood from the right to left ventricle, and a one-compartment model to describe the exchange between blood and myocardium. Simulations of canine and rat heart imaging were performed to evaluate parameter estimation errors. Arterial blood sampling in rats and blood pool imaging in dogs were used to evaluate factor and structure accuracy. Variable infusion duration studies in canine were used to evaluate MB structure and global MBF reproducibility. All results were compared to a previously published minimal structure overlap (MSO) method.
Canine heart simulations demonstrated that MB has lower root-mean-square error (RMSE) than MSO for both factor (0.2% vs 0.5%, MB vs MSO, respectively) and structure (3.0% vs 4.7%, ) estimations, as with rat heart simulations (factors: 0.2% vs 0.9%, and structures: 3.0% vs 6.7%, ). MB blood factors compared to arterial blood samples in rats had lower RMSE than MSO (1.6% vs 2.2%, ). There was no difference in the RMSE of blood structures compared to a blood pool image in dogs (8.5% vs 8.8%, ). Myocardial structures were more reproducible with MB than with MSO ( vs 6.2%, ), as were blood structures ( vs 5.6%, ). Finally, MBF values tended to be more reproducible with MB compared to MSO ( vs 18%, ). The execution time of MB was, on average, 2.4 times shorter than MSO due to fewer free parameters.
Kinetic model-based factor analysis can be used to provide physiologically accurate decomposition of dynamic PET images, and may improve the precision of MBF quantification.