P143] Quality assurance of tomoedge dynamic jaws: A retrospective analysis with the statistical control charts approach

Abstract

Purpose TomoEDGETM refers to the sliding window motion of the jaws along the superior-inferior edge of the treatment target. An automatic procedure, (“Dynamic Jaws Sweep” DJS), tests the movement functions with specific exercises and measures the alignment of the linac photon source in relation to jaw collimation evaluating the longitudinal dose profiles. The results of the DJS were retrospectively analyzed using the statistical control charts. Methods From June 2013 to September 2016 the DJS was performed, on a weekly basis, on a TomoHDATM system for a total amount of 130 runs. All measurements were analyzed with TomoTherapy Quality Assurance (TQA) software that provides: the fluence output factor (FOF) for the three clinical jaws aperture (J20, J14, J7), the time skew and the linac photon source position along the longitudinal direction (IECy SP). Qualitative graphs related to specific exercises are also produced. For each quantitative parameter the Individual (ICC) and the Moving Range chart control (MRCC) were implemented. For both charts the Central value, the Upper and Lower Control Limit are based on the average and on the statistical variance of the first 10 runs. The control chart results were compared to the TQA limits, suggested by the vendor. Results The FOF values outside the limits overlap with erroneous qualitative graphs, pointing out a real critical issue of jaw movements. Both the Individual and the Moving range control chart showed analogous results; indeed, with the ICC the FOF were outside limits in 12.3%, 13.1% and 20% of their measurements for J20, J14 and J7 jaw settings respectively and the time skew parameter only in 4.6%, while, with the MRCC, the corresponding values were 9.2%, 10.1%, 14.3% and 9.0% for the time skew parameter. The IECy SP, while remaining within vendor specific values, rapidly diverged from its initial value with an increasing trend, eventually approaching the zero value. Conclusions Control chart approach seems to be an useful tool for studying in depth the results of the DJS procedure. Much more stringent TQA limits, based on the own machine behavior, are useful to easily detect critical issue related to the dynamic jaws movements.