Integration of a dose-voxel kernel (DVK) convolution method for voxel-based dosimetry in Particle and Heavy Ion...

Area of Science:

• Medical Physics
• Nuclear Medicine
• Computational Dosimetry

Background:

• Accurate absorbed dose calculation is crucial for radiopharmaceutical therapy.
• Current methods like Monte Carlo (MC) simulations are computationally intensive.
• Dose-Voxel Kernel (DVK) convolution offers a physics-based alternative.

Purpose of the Study:

• To implement and evaluate DVK convolution within the Particle and Heavy Ion Transport code System (PHITS).
• To compare the performance of the new DVK implementation against MC simulations and clinical dosimetry methods.
• To assess the efficiency and accuracy of PHITS-integrated DVK for voxel-based dosimetry.

Main Methods:

• DVK convolution was implemented in PHITS using RT-PHITS for data conversion.
• DVKs were generated for 24 human tissue materials.
• Calculations were validated against MC simulations using gamma index, mean absorbed dose in VOIs, and voxel-wise differences.
• Results were benchmarked against Local Deposition Model (LDM) and Voxel S-Value (VSV) convolution in MIM Software.

Main Results:

• DVK calculations in PHITS were significantly faster (~40 s) than MC simulations (>9 h).
• Gamma pass fractions were high (92.1-99.9%), with mean DVK-MC dose rate differences of -1.2 ± 2.9% (SPECT/CT) and -1.0 ± 4.3% (PET/CT).
• Good agreement was observed in liver tissue, with variability in lung and liver-lung interface regions.

Conclusions:

• The PHITS-integrated DVK implementation provides an efficient and accurate voxel-based dosimetry framework.
• This physics-based approach closely matches high-statistics MC simulations.
• It offers a practical solution for patient-specific radiopharmaceutical therapy dosimetry.