Author: Michael Paul Barnes, Lana C. Critchfield, John J. DeMarco, Ellie Durussel, Nilendu Gupta, James Irrer, Peyton Jensen, Grace Gwe-Ya Kim, Cory S. Knill, Seng Boh Gary Lim, Justin K. Mikell, Jean M. Moran, Seyyedeh Azar Oliaei Motlagh, Mario Perez, Michael Pillainayagam, Claire Zhang 👨🔬
Affiliation: University of Michigan, Department of Radiation Oncology, University of Michigan, Department of Medical Physics, Memorial Sloan Kettering Cancer Center, Department of Radiation Oncology, The Ohio State University Wexner Medical Center, Washington University School of Medicine in St. Louis, University of California San Diego, Royal North Shore Hospital, Calvary Mater Hospital Newcastle, Cedars-Sinai Medical Center 🌍
Purpose: EPIDs are widely used in linac QA due to their efficiency and accuracy. As EPIDs age, performance degradation may occur, potentially compromising reliability of EPID-based automated QA results. This study investigates whether performance degradation on an EPID significantly affects QA measurements by comparing results obtained immediately before and after panel replacement.
Methods: A 4.5-year-old as1200 EPID, exhibiting clinically significant panel degradation from heavy portal dosimetry use, was calibrated one month before and immediately after panel replacement. EPID-based QA tests were conducted to evaluate beam output, symmetry, MLC offset, dose constancy at variable dose rate/gantry speed, and MLC leaf speed using the Automated Quality Assurance (AQA) suite. Seven measurements were acquired pre-replacement, and five were collected post-replacement. Beam output was compared to daily QA results, and statistical analysis was performed using a Student’s t-test, along with calculations of mean and standard deviation, to assess the significance of observed differences between pre- and post-panel replacement measurements.
Results: Beam output measurements showed minimal differences, with no significant changes except for 16MV (p < 0.02), corresponding to only a 1.1% output change, undetectable in daily QA. MLC leaf speed showed a significant difference (p < 0.01) post-panel replacement, while dose rate/gantry speed tests showed no significant variation, suggesting minor machine performance changes rather than EPID degradation. MLC leaf offset and transverse symmetry demonstrated significant differences (p < 0.01), but changes were small, with maximum mean differences of 0.04mm and 0.08%, respectively. Axial symmetry showed no significant differences. Overall, panel replacement had minimal impact on QA performance.
Conclusion: While differences were statistically significant for EPID QA results with age, these changes may not reach clinical significance when EPIDs are calibrated within one month of data acquisition. Future work will involve simulating artifacts to identify thresholds where performance degradation begins to affect QA results.