Room: Davidson Ballroom B
Purpose: To develop a 3D digital mouse model that can be integrated with Geant4 toolkit for Monte Carlo simulations of multimodal imaging with benchtop x-ray fluorescence computed tomography (XFCT) and conventional cone beam computed tomography (CBCT).
Methods: Two sets of DICOM images of a nude mouse (pre-/post-injection of gold nanoparticles (GNPs) in vivo), were read and voxelized using Geant4. Specifically, 2D pixels (150 Î¼m Ã— 150 Î¼m) were converted into 3D voxels (150 Î¼m Ã— 150 Î¼m Ã— 150 Î¼m) and layers of DICOM files were stacked together to build the full 3D mouse models (with/without GNPs). Based on the CT numbers, the materials for the voxels were assigned from any of the following materials: Air, Lung-Inhale/Exhale, Tissue, Water, Muscle, Liver, Trabecular-/Dense-Bone, and 1 wt. % Gold. For CT simulations, 40 kVp filtered (65 Âµm Cu plus 800 Âµm Al) cone-beam was used as the primary beam and a flat panel photon counting detector was used to detect transmitted photons. For XFCT simulations, 125 kVp filtered (1.8 mm Sn) fan-beam was used to irradiate the mouse and the emitted photons were detected using two energy-resolving array detectors placed perpendicular to the beam axis at both sides of the mouse.
Results: A realistic voxelized digital mouse model was developed and tested for CBCT and XFCT simulations based on DICOM files. Gold K-shell XRF photons were detected in some crystals of array detectors, allowing XFCT to confirm the presence of 1 wt. % GNPs in the post-injection mouse model, while GNPs were found indistinguishable from bone structures using conventional CBCT.
Conclusion: Proposed digital mouse models can be used to investigate novel approaches to accomplish multimodal (as well as multiplexed) imaging based on XFCT and conventional CT implemented within the same platform, in conjunction with metal nanoprobes such as GNPs.
Funding Support, Disclosures, and Conflict of Interest: Supported by NIH grants R01CA155446 and R01EB020658