From phase-based to displacement-based gating: A software tool to facilitate respiration-gated radiation treatment

Abstract

Background: The Varian Real-Time Position Management (RPM) system allows respiratory gating based on either the phase or displacement (amplitude) of the breathing waveform. A problem in clinical application is that phase-based gating, required for respiration-correlated (4D-CT) simulation, is not robust to irregular breathing patterns during treatment, and a widely used system version (1.6) does not provide an easy means to change from a phase-based gate into an equivalent displacement-based one. We report on the development and evaluation of a robust method to convert phase-gate thresholds, set by the physician, into equivalent displacement-gate thresholds to facilitate its clinical application to treatment. Methods: The software tool analyzes the respiration trace recorded during the 4D-CT simulation and determines a relationship between displacement and phase through a functional fit. The displacement- gate thresholds are determined from an average of two values of this function, corresponding to the start and end thresholds of the original phase gate. The software tool was evaluated in two ways: first, whether in-gate residual target motion and predicted treatment beam duty cycle are equivalent between displacement-gating and phase-gating during 4D-CT simulation (using retrospective phase recalculation); second, whether residual motion is improved with displacement-gating during treatment relative to phase-gating (using real-time phase calculation). Residual target motion was inferred from the respiration traces and quantified in terms of mean and standard deviation in-gate displacement relative to the value at the start of the recorded trace. For retrospectively-calculated compared with real-time-calculated breathing traces, we evaluate the inaccuracies of real-time phase calculation by measuring the phase gate position in each trace as well as the comparison of the mean in-gate displacement and standard deviation of the displacement. Results: Retrospectively calculated data from 10 patients were analyzed. The patient-averaged mean ± standard deviation displacement was reduced from 0.16 ± 0.14 cm for phase-gating under simulation conditions to 0.12 ± 0.08 cm for displacement-gating. Evaluation of respiration traces under treatment conditions (real-time phase calculation) showed that the average displacement gate threshold results in a lower in-gate mean and residual motion (variance) for all patients studied. The patient-averaged mean ± standard deviation displacement was reduced from 0.26 ± 0.18 cm for phase-gating (under treatment conditions) to 0.15 ± 0.09 cm for displacement-gating. Conclusions: Real-time phase gating can lead to inaccuracies when the breathing trace is irregular or because of other unidentified reasons. In such cases, the in-gate displacement differed by as much as 2 mm between phase-based and displacement-based gating for the real-time phase calculation. Displacement-based gating is equivalent to phase-based gating for retrospectively-calculated phase information.