Encrypted login | home

Program Information

Evaluating Accuracy of PET-Based Pencil-Beam Range Verification in Carbon Ion Therapy Via Monte Carlo Simulations

no image available
Y Chi

Y Chi*, N Qin , Z Tian , Y Shao , W Lu , X Jia , The University of Texas Southwestern Medical Center, Dallas, TX


SU-H1-GePD-T-1 (Sunday, July 30, 2017) 3:00 PM - 3:30 PM Room: Therapy ePoster Lounge

Purpose: Range uncertainty is a major concern in carbon ion therapy. Positron Emission Tomography (PET) can be used to verify the beam range by detecting positron emitting nuclei (PEN) distribution. Recently, online-PET for range verification of a pencil beam has been proposed. Because of the known beam-central axis, 1D activity can be reconstructed, which may improve verification accuracy compared to conventional approach by reconstructing a 3D distribution. In this study, we used GPU-based Monte Carlo (MC) tools to perform large-scale simulations to evaluate the accuracy of online-PET-based pencil-beam range verification.

Methods: We used a GPU-based carbon-ion therapy MC tool to calculate PEN distribution induced by a carbon-ion pencil-beam with initial energy 300 MeV/u and 1.8e6 particle numbers corresponding to a pencil beam in a typical treatment. Subsequent simulation of PEN decay, positron annihilation, photon transport and photon signal score at PET detectors was achieved using an in-house developed GPU-based photon/electron MC tool. An effective PET detector response was assumed. Based on detected photon signals, we reconstructed 1D PEN distribution along the known pencil-beam central axis via a back-projection approach. Repeated simulations were performed to study impacts of system configurations on the accuracy of predicted PEN range.

Results: As the PET detector angular coverage reduced from 360ᴼ to 45ᴼ, error (difference between estimated and ground truth PEN ranges) (range defined at 60% peak activity) increased from 0.04±0.04mm to 0.27±0.20mm. Under 45ᴼ coverage, among the range definitions of 70% to 40% peak activity, the smallest error was 0.24±0.16mm at 50% peak activity. When PET data acquisition window reduced from 10min to 2min post irradiation, error increased from 0.30±0.15mm to 0.74±0.54mm.

Conclusion: Range verification for the 1D pencil-beam case may be achieved with sub-millimeter accuracy using 45ᴼ PET detector coverage, the definition of 50% peak activity, and 2 min data acquisition time.

Contact Email: