Fine-resolution voxel S values for constructing absorbed dose distributions at variable voxel size

Area of Science:

• Medical Physics
• Radiotherapy Dosimetry
• Nuclear Medicine Imaging

Background:

• Accurate dosimetry is crucial for effective targeted radiotherapy.
• Existing methods for dose calculation from SPECT/PET data can be computationally intensive.
• MIRD pamphlet no. 17 provides a foundational methodology for voxel S values (VSVs).

Purpose of the Study:

• To present a revised voxel S values (VSVs) approach for enhanced dosimetry in targeted radiotherapy.
• To enable dose calculation for any voxel size and shape from SPECT or PET datasets.
• To update and improve upon the methodology presented in MIRD pamphlet no. 17.

Main Methods:

• Generated VSVs in soft tissue using the MCNPX Monte Carlo code for 9 common radionuclides.
• Developed a resampling algorithm to compute VSVs for user-defined voxel dimensions.
• Validated the revised VSV approach against direct Monte Carlo simulations and 3D-RD software for phantom and patient data.

Main Results:

• The revised VSV approach demonstrated excellent agreement with direct Monte Carlo simulations, with mean relative dose differences below 0.61% across various models.
• Calculations using the VSV approach were significantly faster, completing in under 10 seconds compared to hours for 3D-RD.
• The method showed high voxel-to-voxel dose ratio agreement (0.991-0.996) when compared to EGSnrc-based MC implementations.

Conclusions:

• The novel VSV approach provides accurate absorbed dose calculations from SPECT/PET activity maps.
• This method is substantially faster and more clinically compatible than existing direct simulation techniques.
• The revised VSV approach represents a significant advancement for dosimetry in targeted radionuclide therapy.