Relationship between electron density and effective densities of body tissues for stopping, scattering, and nuclear interaction of proton and ion beams

TL;DR

This paper develops invariant conversion functions linking CT numbers to electron and effective densities for proton and ion therapy, enhancing heterogeneity correction accuracy in treatment planning.

Contribution

It introduces universal, invariant conversion functions for effective densities related to stopping, scattering, and nuclear interactions, improving clinical heterogeneity correction methods.

Findings

Strong correlation between electron density and effective densities.

10% difference in scattering and nuclear interaction parameters for bones.

Potential for a few percent dose calculation improvement in large bone traversals.

Abstract

Purpose: In treatment planning of charged-particle radiotherapy, patient heterogeneity is conventionally modeled as variable-density water converted from CT images to best reproduce the stopping power, which may lead to inaccuracies in the handling of multiple scattering and nuclear interactions. Although similar conversions can be defined for these individual interactions, they would be valid only for specific CT systems and would require additional tasks for clinical application. This study aims to improve the practicality of the interaction-specific heterogeneity correction. Methods: We calculated the electron densities and effective densities for stopping power, multiple scattering, and nuclear interactions of protons and ions, using the standard elemental-composition data for body tissues to construct the invariant conversion functions. We also simulated a proton beam in a lung-like geometry and a carbon-ion beam in a prostate-like geometry to demonstrate the procedure and the effects of the interaction-specific heterogeneity correction. Results: Strong correlations were observed between the electron density and the respective effective densities, with which we formulated polyline conversion functions. Their effects amounted to 10% differences in multiple-scattering angle and nuclear-interaction mean free path for bones compared to those in the conventional heterogeneity correction. Although their realistic effect on patient dose distributions would be generally small, it could be at the level of a few percent when a beam traverses a large bone. Conclusions: The present conversion functions are invariant and may be incorporated in treatment planning systems with a common function relating CT number to electron density. This will enable improved beam dose calculation while minimizing initial setup and quality management of the user's specific system.