摘要
The purpose of this study is to evaluate the performance of an adaptive gating method, which is designed to accommodate the beam-to-beam and day-today variation of the internal/external correlation, as well as the real tumor position during respiratory-gated fractionated radiotherapy. We define a two-step procedure: (1) before each treatment, target positions are detected and synchronized with an external surrogate for establishing the internal/external correlation model and determining the position of the gating window, and then (2) during the delivery of the treatment beam, the gating is triggered by an external signal based on the updated internal/external correlation and window position. This correlation is described by a linear-quadratic model including a time shift between the internal and external signals. To simulate the proposed method, data of tumor motion in the superior-inferior direction synchronized with an external surrogate during hypo-fractionated radiotherapy from five lung patients are analyzed retrospectively. Duty cycle (DC), target coverage (TC) and the average distance (AD) between the internal target position and the edge of the gating window for all false positives are calculated as evaluative criteria. Under a 5 mm gating window, the average TC is 88.9%, with a DC around 45% and a mean AD of 0.7 mm. A daily update is also simulated for comparison, and it is found that beam-to-beam updating is superior. In conclusion, the combined updating of internal/external correlation and the gating window for each beam can improve the accuracy and reliability of respiratory-gated radiotherapy.