Modelling tissue specific RBE for different radiation qualities based on a multiscale characterization of energy...

Area of Science:

• Medical Physics
• Radiation Biology
• Computational Biology

Background:

• Accurate modeling of relative biological effectiveness (RBE) is crucial for optimizing radiation therapy.
• Existing RBE models may have limitations in predicting biological outcomes for different radiation types.
• Understanding the correlation between nanodosimetric quantities and DNA damage is an active area of research.

Purpose of the Study:

• To introduce the nanoCluE model for predicting RBE of protons and carbon ions.
• To investigate if nano- and microdosimetric quantities can improve the accuracy of Linear-Quadratic (LQ) model parameters.
• To compare the predictive accuracy of nanoCluE with existing RBE models.

Main Methods:

• Developed the nanoCluE model utilizing nanodosimetric and microdosimetric quantities.
• Generated track structure data using the Geant4-DNA Monte Carlo code for various proton and carbon ion energies.
• Correlated nanodosimetric and microdosimetric quantities with experimental LQ data to select optimal parameters.

Main Results:

• The nanoCluE model demonstrated a root mean square error (RMSE) of 1.57 for protons, compared to 1.31 and 1.30 for other models.
• For carbon ions, nanoCluE achieved an RMSE of 2.26, outperforming other models with RMSEs of 3.24 and 5.24.
• The model showed improved accuracy, particularly for carbon ion RBE predictions.

Conclusions:

• The nanoCluE model is feasible for predicting RBE of protons and carbon ions.
• The enhanced accuracy for carbon ions suggests further investigation is warranted.
• This model offers a potential advancement in RBE modeling for particle therapy.