Abstract
PURPOSE: The Synchrony tracking system of Radixact is capable of real-time tumor tracking by building a correlation model between external light-emitting diodes on the patient's chest and an internal marker. A phase shift between the chest wall and a lung tumor has been reported. Hence, this study focused on evaluating the accuracy of the tracking system, especially under a patient-specific breathing pattern with respiratory phase shifts. METHODS: A phantom containing fiducial markers was placed on a moving platform. The intrinsic delivery accuracy was verified with a patient-specific breathing pattern. Three patient-specific breathing patterns were then implemented, for which phase shifts, φ, were introduced. Phase shifts with +0.3 s and +1 s were tested for dosimetric aspects, whereas ±0.3, ±0.6, and ±0.8 s shifts were used for tracking accuracy. The resultant dose distributions were analyzed by γ comparison. Dose profiles in the superior-inferior and lateral directions were compared. Logfiles of the tracking information were extracted from the system and compared with the input breathing pattern. The root mean square (RMS) difference was used to quantify the consistency. RESULTS: When the φ value was as large as 1 s, a severe inconsistency was observed. The target was significantly underdosed, down to 89% of the originally planned dose. γ analysis revealed that the failed portion was concentrated in the target region. The RMS of the tracking difference was close to 1 mm when φ was ±0.3 s and approximately 4 mm when φ was ±0.8 s. Tracking errors increased with an increase in the degree of phase shifts. CONCLUSION: Phase shifts between the patient chest wall and the internal target may hamper treatment delivery and jeopardize treatment using Synchrony Tracking. Hence, a larger planning target volume (PTV) may be necessary if a large phase shift is observed in a patient, especially when an external surrogate shows a lag in motion when compared with the tumor.