271 - Clinical Integration of a Novel Snout-Mounted Quad-Camera Prompt Gamma Imaging System for In Vivo Proton Range Verification
Presenter(s)
F. Safavi1, S. W. Peterson2, S. Mossahebi3, V. R. Sharma4, A. R. Chalise5, E. Shakeri6, H. Lewis6, M. K. Gobbert6, J. Polf7, and L. Ren8; 1University of Maryland, Baltimore, Baltimore, MD, 2Department of Physics, University of Cape Town, Cape Town, South Africa, 3Department of Radiation Oncology, University of Maryland School of Medicine, Baltimore, MD, 4Department of Medical Physics, University of Wisconsin-Madison, Madison, WI, 5University of Maryland School of Medicine, Baltimore, MD, 6University of Maryland Baltimore County, Baltimore, MD, 7M3D, Inc, Ann Arbor, MI, United States, 8University of Maryland, School of Medicine, Radiation Oncology, Baltimore, MD
Purpose/Objective(s): To report the first clinical integration of a snout-mounted quad-camera prompt gamma imaging system and to evaluate its millimeter-scale accuracy for in vivo proton range verification during routine beam delivery. We hypothesized that this integrated configuration would enable reproducible detection of clinically relevant proton range shifts across varying gantry angles, beam energies, delivery modalities, and controlled range perturbations.
Materials/Methods: A snout-mounted quad-camera prompt-gamma imaging system consisting of four position-sensitive solid-state Compton camera modules was integrated on a clinical proton therapy gantry for real-time detection of prompt-gamma emissions during beam delivery. The modules were arranged in a rectangular geometry centered on the beam axis to enable multi-angle detection. Cameras were synchronized for simultaneous acquisition and connected via Ethernet to a remote workstation for real-time monitoring and data processing. Measurements were performed under clinical pencil-beam scanning delivery, including single-energy spots and spread-out Bragg peak fields at 170 and 245 MeV, gantry angles of 90° and 270°, and doses of 2 and 7.5 Gy. Prompt-gamma events were reconstructed using a GPU based kernel weighted back projection algorithm with Compton filtering. The primary endpoint was distal falloff shift relative to introduced physical range shifts (0–10 mm). Reproducibility was evaluated across repeated deliveries. Range-shift sensitivity and profile agreement were quantified using linear regression and root-mean-square error analysis.
Results: The snout-mounted quad-camera PGI system operated stably during clinical proton beam delivery without workflow disruption or interference with treatment geometry. Controlled range shifts (0–10 mm) produced corresponding and reproducible distal prompt-gamma profile displacements across all runs. A strong linear relationship was observed between introduced and measured range shifts across tested energies and gantry angles. Detection of shifts =6 mm was achieved reliably under clinical beam conditions. Multi-camera acquisition enabled consistent localization of emission profiles across repeated measurements, demonstrating stable performance during realistic treatment delivery.
Conclusion: This study demonstrates the first clinical integration of a snout-mounted quad-camera Compton prompt-gamma imaging system and its ability to detect millimeter-scale range shifts during clinical beam delivery. The system enables real-time multi-camera acquisition and monitoring under routine treatment conditions. These results support the feasibility of in vivo proton range verification using prompt-gamma imaging and establish a foundation for clinical translation toward reducing proton range uncertainties.