Room: AAPM ePoster Library
Purpose: To develop a method to verify, in real-time, the reproducibility of each breath-hold for liver and lung SBRT patients during radiation therapy treatment delivery.
Methods: Five patients treated with SBRT to the liver dome or lower lobe of the lung using the Active Breathing Coordinator (ABC) breath-hold device were included in this work. kV planar images acquired with each breath-hold and every 60° of the gantry rotation during VMAT delivery were used to manually trigger the beam on/off based on the location of the diaphragm or GTV position with respect to physician drawn contours from three repeated CT simulation scans. Post-treatment, all triggered images were exported from the treatment console and the diaphragm or GTV was manually delineated. The distance from this contour on the triggered images to the physician drawn contour was calculated and compared to clinically acceptable positional tolerances.
Results: For these patients, there were multiple instances where the beam was manually triggered off due to breath-hold irregularity, demonstrating the clinical value of this technique. For liver SBRT, the maximum and mean distance from the manually drawn and physician contours of the liver was greater than the 5-mm PTV margin value for 37% and 11% of all triggered images. For lung SBRT, the GTV from the triggered images was outside of the physician drawn PTV contour for 42% of all triggered images with an average deviation outside of the PTV of 0.4-mm.
Conclusion: The verification technique for breath-hold reproducibility presented in this work was able to identify breath-hold variations that were in excess of planned margins according to the internal anatomy of the liver and lung for SBRT patients during treatment. Future work will increase the size of the data while improving contour definition and evaluating the dosimetric impact of the breath-hold variability in these patients.