Author: Zahra Ahmadi Ganjeh, Stefan Both, Peter Dendooven, Jeffrey Free, Brian Zapien Campos 👨🔬
Affiliation: Department of Radiation Oncology, University Medical Center Groningen, University of Groningen, Particle Therapy Research Center (PARTREC), Department of Radiation Oncology, University Medical Center Groningen, University of Groningen 🌍
Purpose: Positron emission tomography (PET) has been investigated for range monitoring in intensity-modulated proton therapy (IMPT), but real-time feedback been limited due to the focus on long-lived positron emitters imaging. Our previous studies demonstrated sub-second imaging of 12N (T1/2=11 ms) activity for range verification at individual beam spot levels in single-field irradiations. This study aims to evaluate the feasibility of 12N imaging for range verification in multi-field irradiations, which brings the approach closer to a more realistic clinical scenario. Furthermore, we assess the accuracy to detect in real-time treatment deviations due to range shifted scenarios.
Methods: A CIRS-731 HN phantom was irradiated using a clinical proton beam. A 4-cm square target in the head was irradiated with 4 GyRBE from two fields delivered at gantry angles of 90º and 0º, with a 100-ms delay in-between the beam spots to enable 12N imaging using a dual-head Siemens PET scanner. Alongside to the nominal irradiation, modified scenarios were irradiated using 2- and 5-mm solid water slabs in order to mimic treatment setup deviations. Data analysis consisted of 2D image reconstruction and calculation of positron activity range (PAR) at beam spot and energy layer levels.
Results: The PAR uncertainty of the first layer (1.1×109 protons of 152 MeV) of the first field was 0.8-1.2 mm, increasing to 1.2-1.8 mm for the first layer (2.2×109 protons of 150.3 MeV) of the second field. Range shifts detected were 2.1 and 1.8 mm for the 2-mm slab and 4.4 and 4.8 mm for the 5-mm slab.
Conclusion: 12N imaging approach was successful for online range verification in multi-field IMPT. Range precision decreased for the second field by a factor of 1.5 with respect to the first field. The method successfully detected range deviation within sub-second time scales, demonstrating the potential for its implementation in clinical workflows.