Dosimetric impact of hysteresis on lung cancer tomotherapy: A moving phantom study.

PURPOSE: To investigate the dosimetric impact of hysteresis on lung cancer tomotherapy.

METHODS: Measurements were acquired using MapCheck with an XY4D motion simulation table. Six hysteresis states (0, π/32, π/16, π/8, 3π/16, and π/4) were considered with sinusoidal motions in the superior-inferior and left-right orientations. The measured data were analyzed both globally (from all detectors) and structure-by-structure in the measurement plane. The dose difference (DD) analysis method with local normalization in the absolute dose mode with a DD threshold of 6 cGy was adopted to analyze each hysteresis vs. static state (H(∗)S) and nonzero vs. zero hysteresis (H(∗)0). The threshold was 10% for all analyses. Wilcoxon signed rank tests with significance level p = 0.05 were used for statistical analysis.

RESULTS: The DD analysis of each H(∗)S mostly indicated that the passing rate differed between structures but was similar between hysteresis states. The DD analysis of H(∗)0 showed that the passing rate decreased with increasing hysteresis. The differences between larger hysteresis (≥3π/16) and other states were significant for comparisons between global, left lung, chest wall, and target. Both analyses showed that the DD distribution changed with hysteresis.

CONCLUSIONS: Hysteresis difference causes the DD distribution to change. Structural difference had more impact than hysteresis state difference on hysteresis motion vs. static comparisons. Remarkable effects on nonzero vs. zero hysteresis comparisons were only seen for structures closely related to the target at large hysteresis. Small organs at risk that are close to the target need to be considered further.