LET spectra scoring for applications in proton radiotherapy

Area of Science:

• Medical Physics
• Radiation Oncology
• Computational Science

Background:

• Proton therapy utilizes mixed radiation fields in heterogeneous patient geometries.
• Dose-averaged LET (LETd) is commonly used but offers limited information.
• Advanced treatment planning requires more comprehensive radiation field characterization.

Purpose of the Study:

• To implement, validate, and demonstrate the application of fast, voxel-based single proton linear energy transfer (LET) spectra scoring.
• To provide more comprehensive data on mixed radiation fields than LETd.
• To enable accurate analysis for clinical proton therapy.

Main Methods:

• Implemented single particle spectra scoring for LET and other physics quantities in voxelized geometries.
• Utilized GPU-accelerated Monte Carlo (MC) code Fred, alongside Gate/Geant4 and Fluka.
• Validated scorers by comparing spectra from different MC codes and assessing calculation performance.
• Calculated LET spectra for an intensity-modulated proton therapy (IMPT) patient plan.

Main Results:

• Accurate implementation and excellent agreement of LET spectra across Fred, Gate, and Fluka.
• GPU acceleration enabled time-efficient LET spectra calculation (approx. 1 hour with Fred vs. tens of hours with others).
• Single particle LET spectra can differ for the same LETd in IMPT plans, highlighting potential uncertainties in RBE-based planning.

Conclusions:

• State-of-the-art MC methods enable single particle LET spectra scoring for detailed radiation effect insights.
• Fast, time-efficient calculations facilitate clinical translation of LET spectra analysis.
• This method enhances understanding of mixed radiation fields in proton therapy.