Speaker
Description
Background:
Dual-tracer PET imaging exhibits substantial potential for enhancing localization of infections and assessment of inflammation-related metabolic changes over mono-tracer PET. Recently, it has been demonstrated that in vivo imaging of inflammation using 89Zr labelled antibodies has a higher selectivity index compared with the commonly used tracer [18F]FDG, positioning it as a strategic tool for investigating inflammatory processes. Furthermore, the combination of 89Zr and [15O]H2O or [18F]FDG tracer pairs in one study can provide complementary information on perfusion or metabolism, enriching the information available for clinical application. Due to the complexity of conducting simultaneous dual tracer imaging in the clinic, we have developed a simulation framework tailored to create dynamic dual-tracer images from single-tracer data, thereby facilitating the evaluation of performing simultaneous dual-tracer PET imaging in a reasonable clinical context.
Methods:
Three representative ROIs—white matter, grey matter, and skull—were derived from [15O]H2O, [89Zr]atezolizumab, and [18F]FDG patient scans and integrated into the BrainWeb phantom, with a notable hot tumour identified on the skull from [89Zr]atezolizumab. An automatic simulation pipeline was set-up: modelling and extracting kinetic parameters from patient data, assigning the recovered time activity curve to the digital phantom and then performing the analytical simulation. The PET system was modelled based on a Siemens Biograph Vision 600 PET/CT scanner, and simulated projections were reconstructed (4 iterations, 5 subsets, with time-of-flight) using Software for Tomographic Image Reconstruction (STIR). Attenuation correction, normalisation calibration, randoms, scatter and Poisson noise were also included.
Results:
Figure displays a representative slice from each reconstructed dynamic images: A, B, C showcase combinations of dual-tracer pairs ([18F]FDG, [15O]H2O, and [89Zr]atezolizumab), while D, E, F correspond to mono-tracer images. The completion of this automated simulation pipeline enables the execution of simulation studies under clinically realistic conditions. Additionally, the inclusion of other tracers (e.g. [68Ga]Ga-FAPI) may be explored in future developments.
