[Paper Review] Dosimetric evidence confirms computational model for magnetic field induced dose distortions of therapeutic proton beams

Sonja M. Schellhammer, Sebastian Gantz+4 more|arXiv (Cornell University)|Sep 1, 2017

Advanced Radiotherapy Techniques30 references17 citations

TL;DR

This study provides the first experimental validation of magnetic field-induced proton beam deflection in therapeutic proton beams using film dosimetry in a 1 T transverse magnetic field. Measured lateral deflections of 1–10 mm for 80–180 MeV beams agreed with Geant4 simulations within 1 mm, confirming that magnetic field-induced dose distortions are both measurable and accurately predictable for MR-guided proton therapy systems.

ABSTRACT

Given the sensitivity of proton therapy to anatomical variations, this cancer treatment modality is expected to benefit greatly from integration with magnetic resonance (MR) imaging. One of the obstacles hindering such an integration are strong magnetic field induced dose distortions. These have been predicted in simulation studies, but no experimental validation has been performed so far. Here we show the first measurement of planar distributions of dose deposited by therapeutic proton pencil beams traversing a one-Tesla transversal magnetic field while depositing energy in a tissue-like phantom using film dosimetry. The lateral beam deflection ranges from one millimeter to one centimeter for 80 to 180 MeV beams. Simulated and measured deflection agree within one millimeter for all studied energies. These results proof that the magnetic field induced proton beam deflection is both measurable and accurately predictable. This demonstrates the feasibility of accurate dose measurement and hence validates dose predictions for the framework of MR-integrated proton therapy.

Motivation & Objective

• To experimentally validate computational models predicting magnetic field-induced dose distortions in proton therapy.
• To measure lateral beam deflection and range retraction in a clinical-relevant 1 T magnetic field using tissue-equivalent phantoms.
• To assess the accuracy of Geant4-based Monte Carlo simulations against experimental data for proton beams of varying energies.
• To establish the feasibility of accurate dose measurement in the presence of strong magnetic fields, enabling safe integration of MRI with proton therapy.
• To quantify uncertainties in beam deflection and range retraction to support clinical implementation of MR-integrated proton therapy.

Proposed method

• Irradiated a water-equivalent phantom with proton pencil beams (80–180 MeV) in the presence of a 1 T transverse magnetic field.
• Used EBT3 film dosimetry to measure 2D dose distributions with sub-millimeter spatial resolution.
• Extracted beam trajectory and Bragg peak position from lateral dose profiles via Gaussian fitting to lateral beam profiles.
• Calculated lateral deflection and range retraction by comparing Bragg peak positions with and without magnetic field.
• Performed Geant4 Monte Carlo simulations with 2.048×10^7 primary particles per simulation to achieve statistical uncertainty <0.1 mm.
• Quantified uncertainties via quadratic sum of systematic and statistical contributions, including beam energy, magnetic field map, and phantom density.

Experimental results

Research questions

• RQ1 Can magnetic field-induced proton beam deflection be experimentally measured with sub-millimeter accuracy in a clinical 1 T magnetic field?
• RQ2 How well do Geant4 Monte Carlo simulations predict the lateral deflection and range retraction of proton beams in the presence of a 1 T transverse magnetic field?
• RQ3 What are the dominant sources of uncertainty in measuring and simulating beam deflection and range retraction in a magnetic field?
• RQ4 To what extent do measured beam deflections agree with simulated predictions across a range of proton beam energies (80–180 MeV)?
• RQ5 Is the magnetic field-induced dose distortion in proton beams predictable and measurable with sufficient precision for clinical implementation in MR-guided proton therapy?

Key findings

• Lateral beam deflection ranged from 1.0 mm (80 MeV) to 10.1 mm (180 MeV), with measured values of 1.0±0.4 mm and 10.1±0.5 mm, respectively.
• Simulated deflections (1.0±0.1 mm at 80 MeV, 9.3±0.3 mm at 180 MeV) agreed with measurements within 1 mm across all energies.
• Range retraction was minimal, with measured values ranging from -0.5 mm (100 MeV) to +0.4 mm (180 MeV), and simulated values within ±0.3 mm.
• The largest source of uncertainty was the magnetic field map (up to 0.2 mm at 180 MeV), contributing approximately 2% to the deflection uncertainty.
• Systematic uncertainties were below 0.3 mm for deflection and 0.2 mm for range retraction, with statistical uncertainties <0.1 mm.
• The agreement between simulation and measurement within 1 mm validates the predictive capability of Geant4-based models for MR-integrated proton therapy.