[Paper Review] Collimator effects in proton planning

E. Matsinos|ArXiv.org|Nov 7, 2008

Solar and Space Plasma Dynamics2 references3 citations

TL;DR

This paper presents a two-step method to correct for proton beam collimator effects—specifically, block thickness and scattering—by decomposing their fluence contributions into manageable one-dimensional components. Using Monte Carlo simulations during beam configuration and interpolative reconstruction during planning, the method improves dose accuracy by up to a few percent near block edges, with validation showing substantial improvement in reproducing half-block fluence data when scattering effects are included.

ABSTRACT

The present paper pertains to corrections which are due to the presence of beam-limiting and beam-shaping devices in proton planning. Two types of corrections are considered: those which are due to the nonzero thickness of such devices (geometrical effects) and those relating to the scattering of beam particles off their material. The application of these two types of corrections is greatly facilitated by decomposing the physical effects (i.e., the contribution to the fluence) of two-dimensional objects (i.e., of the apertures of the devices) into one-dimensional, easily-calculable contributions. To minimise the time requirements in the derivation of the scattering corrections, a two-step process is introduced. The first step occurs at beam-configuration phase and comprises the analysis of half-block fluence measurements and the extraction of the one parameter of the model which is used in the description of the beamline characteristics; subsequently, a number of Monte-Carlo runs lead to the determination of the parameters of a convenient parameterisation of the relevant fluence contributions. The second step involves (at planning time) the reconstruction of the parameters (which are used in the description of the scattering contributions) via simple interpolations, performed on the results obtained during the beam-configuration phase. It is shown that the inclusion of the scattering effects leads to substantial improvement in the reproduction of the experimental data. The contributions from the block-thickness and block-scattering effects have been presented separately in the case of a simple water phantom. In this example, the maximal contribution of the block-relating effects amounts to a few percent of the prescribed dose.

Motivation & Objective

• To address the clinical omission of geometric and scattering corrections from beam-limiting and beam-shaping devices (BL/BSDs) in proton therapy planning.
• To develop a computationally efficient method for applying corrections due to nonzero block thickness and proton scattering off block materials.
• To enable real-time application of these corrections during treatment planning by pre-computing scattering parameters via Monte Carlo simulations.
• To validate the method using half-block fluence measurements, excluding scattering contributions during calibration to enable self-consistent verification.

Proposed method

• The method decomposes 2D collimator effects into 1D miniblock contributions for simplified fluence calculation.
• A two-step process is employed: first, beam configuration phase extracts a single model parameter (λ) from half-block fluence measurements; second, planning phase uses interpolation of precomputed Monte Carlo results to reconstruct scattering contributions.
• Monte Carlo simulations generate parameterized fluence contributions across varying block materials, thicknesses, incident energies, and nozzle-equivalent thicknesses.
• Scattering effects are categorized into three track types: outer tracks (OTs), bore-scattered inner tracks (BSITs), and going-through inner tracks (GTITs), corresponding to Courant's particle types.
• The scattering correction is parameterized using expansion coefficients in two geometric variables derived from the Monte Carlo output.
• The method enables real-time application of corrections during treatment planning via simple interpolation of precomputed data.

Experimental results

Research questions

• RQ1 How can the combined effects of block thickness and scattering be accurately modeled in proton therapy treatment planning?
• RQ2 Can a two-step approach using beam configuration phase Monte Carlo simulations and planning phase interpolation enable efficient and accurate correction of collimator-induced dose errors?
• RQ3 To what extent do scattering contributions improve the reproduction of experimental half-block fluence measurements?
• RQ4 What is the magnitude and spatial distribution of dose corrections due to block thickness and scattering in a clinical-relevant water phantom scenario?

Key findings

• The inclusion of scattering corrections leads to a substantial improvement in the reproduction of half-block fluence measurements, as validated by reduced χ² values.
• Block-thickness corrections reduce dose due to beam blocking, while block-scattering corrections increase dose due to protons re-emerging after scattering off the block material.
• In the water phantom case, the combined effects of thickness and scattering amount to a few percent of the prescribed dose, with the largest contributions occurring near the block border and in the entrance region.
• The entrance region exhibits significant dose contributions from low-energy scattered protons, which are particularly relevant due to the relatively low dose delivered there.
• The two-step method successfully reconstructs scattering effects via interpolation, enabling efficient application during clinical treatment planning.
• The method demonstrates that scattering effects are not negligible and must be accounted for, especially in regions adjacent to the block where dose gradients are steep.