Line-Of-Response (LOR)-Based Mid-Range Probing—towards Online Proton Range Verification and Plan Adaptation
L Ma1*, M Chen1, D Yang1, Y Zhong1, X Jia1, X Gu2, Y Shao1, W Lu1, (1) University of Texas Southwestern Medical Center, Dallas, TX, (2) Stanford University, Stanford, CA
Presentations
TU-D930-IePD-F7-5 (Tuesday, 7/12/2022) 9:30 AM - 10:00 AM [Eastern Time (GMT-4)]
Exhibit Hall | Forum 7
Purpose: In-vivo range measurement is critical to proton therapy and plan adaptation. As a continuation of the mid-range probing strategy, this work 1) performed an end-to-end Monte Carlo (MC) simulation from beam delivery to PET acquisition, 2) proposed a non-image-based range measurement workflow and 3) answered the key question for on-line measurement: how much dose and how long acquisition is needed to collect enough signal and to achieve millimeter-level range measurement accuracy.
Methods: The end-to-end MC simulation was performed in two steps. First, we simulated proton transportation with gPMC_v3.0, an in-house GPU-accelerated proton MC package, which outputs dose and positron emitter distributions. Second, we simulated PET acquisition with GATE to obtain the detected lines-of-response (LOR).We proposed an LOR-based range measurement without PET image reconstruction. A “surrogate” was defined as the point closest to the LORs near planned distal end position. With simulated range variation scenarios, we can establish the correspondence from surrogate to beam range.To guide spot and acquisition time selection, we further calculated the uncertainty of measured range by repeating the simulation for each experiment setting (combination of spot dose and measurement time).
Results: We performed MC simulation for proton transportation in brain and acquisition of proton-induced LORs for a customized brain PET scanner. 0.04~0.08 Gy Dmax corresponded to 10~20 O15 per mm3. We established a calibration curve that converts the measured LORs to beam range (dose R50) and conducted the uncertainty analysis. For a spot with Dmax of 0.4Gy and 60-second acquisition, 95% confidence interval of measured range was ±1.5mm.
Conclusion: With end-to-end MC simulation, we established a basic benchmark and workflow for PET-based on-line proton range measurement.
Funding Support, Disclosures, and Conflict of Interest: This work was supported in part by NIH grants (R01 CA235723, R01 CA218402).
Keywords
Not Applicable / None Entered.
Taxonomy
Not Applicable / None Entered.
Contact Email