A commissioning protocol for portal imaging-based radiotherapy in vivo dosimetry systems

Background and Purpose: With the availability of commercial electronic portal imaging detector-based in vivo dosimetry (EPID-based IVD) solutions, many radiotherapy departments are adopting this technology. However, comprehensive commissioning guidance is lacking. This study aims to provide a protocol for testing the accuracy and sensitivity of EPID-based IVD systems. Material and methods: The protocol was tested across four institutions using two different systems. Accuracy was evaluated with homogeneous slab phantoms using different square regular fields, and clinical plans in a CIRS lung phantom. Multiple forward and back-projected algorithm implementations were assessed for different energies. Sensitivity analysis in the lung phantom examined responses to setup errors, anatomical variations, and delivery errors. Results: In homogeneous phantoms, over 85 % of pixels passed the 5 %/2mm gamma criteria, except for the 2x2 cm² field. In the lung phantom, all systems and implementations achieved over 95 %-pixel pass rates at the 2 %/2mm criterion for volumetric modulated arc therapy (VMAT) plans. For conformal radiation therapy (3DCRT) plans, one system implementation showed poor accuracy, with over 90 % agreement only at the 5 %/2mm criterion. Considering all systems and implementations, average sensitivity and specificity for CRT plans ranged from 0.92 and 0.42 (at 2 %/2mm) to 0.71 and 0.52 (at 5 %/2mm), while for VMAT plans ranged from 0.41 and 0.81 (at 2 %2mm) to 0.37 and 0.81 (at 5 %/2mm). Conclusion: We successfully developed a protocol to commission EPID IDV systems. It was found that not all systems and implementations achieved satisfactory accuracy and sensitivity, emphasising the need for thorough commissioning and benchmarking.