A microscopic computational simulation of [18F]FDG transport and metabolism identifies valid regimes for compartmental analysis

Abstract

18F-fluorodeoxyglucose ([18F]FDG) positron emission tomography (PET), combined with compartmental modeling, is a powerful non-invasive imaging method for assessing cellular metabolism. However, classical two- and three-tissue compartment models assume homogeneous [18F]FDG distribution within the tissue, which needs justification, and the definition and interpretation of rate constants across these models is not always consistent. To address these issues, we develop a finite difference solver to simulate [18F]FDG transport and metabolism within a 1 mm3 tissue volume, representing the smallest volume resolvable by PET. Our simulations reveal sub-millimeter heterogeneity in [18F]FDG distribution and show that the measured PET signal is dependent not only on cellular metabolic activity but also on interstitial [18F]FDG diffusivity, vascular permeability, and vascular architecture. We further demonstrate that our finite-difference simulation reduces to a three-tissue compartment model when interstitial [18F]FDG concentration is homogeneous. Furthermore, this simplified model itself reduces to the two-tissue compartment model when vascular permeability is sufficiently high. This work quantitatively links vascular permeability, vascular architecture, cellular uptake kinetics, [18F]FDG diffusivity, and acquisition time. It also unifies the two- and three-tissue compartment models and identifies their applicable regimes. These findings deepen our understanding of [18F]FDG transport kinetics and enhance the interpretability of dynamic [18F]FDG-PET imaging.