A 5D computational phantom for pharmacokinetic simulation studies in dynamic emission tomography

• Dynamic image acquisition protocols are increasingly used in emission tomography.
• Need for computational phantoms to describe both the spatial and temporal distribution of radiotracers.
• A 5D anthropomorphic phantom was developed, for parametric imaging simulations in emission tomography.
• The phantom is based on real 4D MR data and a detailed multi-compartmental pharmacokinetic modelling simulator.
• Example applications are shown in parametric [18F]FDG and [15O]H2O PET imaging.

IntroductionDynamic image acquisition protocols are increasingly used in emission tomography for drug development and clinical research. As such, there is a need for computational phantoms to accurately describe both the spatial and temporal distribution of radiotracers, also accounting for periodic and non-periodic physiological processes occurring during data acquisition.MethodsA new 5D anthropomorphic digital phantom was developed based on a generic simulation platform, for accurate parametric imaging simulation studies in emission tomography. The phantom is based on high spatial and temporal information derived from real 4D MR data and a detailed multi-compartmental pharmacokinetic modelling simulator.ResultsThe proposed phantom is comprised of three spatial and two temporal dimensions, including periodic physiological processes due to respiratory motion and non-periodic functional processes due to tracer kinetics. Example applications are shown in parametric [18F]FDG and [15O]H2O PET imaging, successfully generating realistic macro- and micro-parametric maps.ConclusionsThe envisaged applications of this digital phantom include the development and evaluation of motion correction and 4D image reconstruction algorithms in PET and SPECT, development of protocols and methods for tracer and drug development as well as new pharmacokinetic parameter estimation algorithms, amongst others. Although the simulation platform is primarily developed for generating dynamic phantoms for emission tomography studies, it can easily be extended to accommodate dynamic MR and CT imaging simulation protocols.