Statistical process control: machine performance check output variation

Background: The aim of this study was to illustrate and evaluate the use of different statistical process control (SPC) aspects to examine linear accelerator daily output variation through machine performance check (MPC) over a month. Methods: MPC daily output data were obtained over a month after AAPM TG-51 were performed. Baseline data were set, and subsequent data were conducted through SPC. The Shewhart chart was used to determine the upper and lower control limits, whereas CUSUM for subtle changes. Results: The upper and lower control limits obtained via SPC analysis of the MPC data were found to fall within AAPM Task Group 142 guidelines. MPC output variation data were within ±3% of their action limits values and were within 1% over thirty days of data. The process capability ratio and process acceptability ratio, Cp and Cpk values were ?2 for all energies. Potential undetected deviations were captured by the CUSUM chart for photons and electrons beam energy. Conclusions: Control charts were found to be useful in terms of detecting changes in MPC output.