Source- and Geometry-Dependent Spatial Profiles in Commercial Bioluminescence Imaging: Experimental Comparison and Geant4 Simulations

In this study, we investigated the transfer of a previously established LED-based Geant4 optical-photon simulation framework to experimental bioluminescence imaging (BLI) using two commercial imaging systems. An IVIS 200 and a SmART+ BLI/CT system were used to compare spatial intensity profiles obtained from a controlled LED source and ATP-driven bioluminescence under defined phantom conditions. Direct comparison demonstrated broader spatial profiles for BLI than for LED emission, with the magnitude of the differences varying between imaging systems. To investigate the transferability of the simulation framework, a spatially extended, isotropically emitting BLI source was implemented in Geant4, including its aqueous volume and sample container. At a phantom thickness of 2.00 mm, the simulations reproduced the general characteristics of the normalized experimental BLI profiles, although the simulated spatial profiles were consistently narrower than those measured experimentally. Separate idealized two-point-source simulations were used to investigate spatial separability under controlled phantom conditions, while additional simulations demonstrated that the source-to-boundary distance influences spatial-profile formation through changes in photon entry conditions and scattering histories. Overall, these findings demonstrate the potential of source-specific Monte Carlo simulations to investigate spatial photon transport in commercial BLI systems and to support the design and optimization of experimental studies through systematic evaluation of source and phantom parameters, potentially reducing the need for extensive experimental test series.