Multi-scale hybrid models for radiopharmaceutical dosimetry with Geant4

Area of Science:

• Medical physics
• Computational modeling
• Radiopharmaceutical dosimetry

Background:

• Monte Carlo (MC) simulations are crucial for accurate radiopharmaceutical absorbed dose calculations.
• Low spatial resolution of imaging techniques limits the accuracy of simulation geometries.
• Voxelization of complex models introduces trade-offs between accuracy and computational speed.

Purpose of the Study:

• To develop computational tools for creating multi-resolution organ models for MC simulations.
• To improve the spatial resolution of dose calculations in critical anatomical structures.
• To maintain or improve simulation speed while enhancing accuracy.

Main Methods:

• Development of computational tools using VTK and Geant4.
• Creation of multi-resolution organ models with varying voxel sizes based on clinical relevance.
• Validation of the multi-resolution approach against standard Geant4 voxel geometry simulations.

Main Results:

• The developed tools enable the creation of multi-resolution models for MC simulations.
• The approach allows for higher spatial resolution in specific anatomical regions.
• Proof-of-principle demonstrates feasibility and validation against standard methods.

Conclusions:

• Multi-resolution modeling offers a viable solution to enhance the accuracy of MC-based dosimetry.
• This approach balances improved spatial resolution in critical areas with computational efficiency.
• The developed tools provide a pathway for more precise radiopharmaceutical dose estimations.