Room: Exhibit Hall | Forum 6
Purpose: To determine the effect field-of-view (FOV) of CBCTs for head and neck (H&N) cancer patients has on accuracy of dose calculation in evaluating planning target volume (PTV) coverage and hence deciding adaptive replanning need.
Methods: Ten H&N patients had CBCTs taken on their replanning day. The patients were divided into head scan and thorax scan groups, based on the CBCT scanning protocol used. The planning CT was registered to the replanning CT and the CBCT so the treatment beams could be copied to them. The beams calculated on the replanning CT provided the reference dose. The beams were also recalculated on the CBCT, with electron density set to water within the external contour (homogeneous calculation). In the head scan group, the external contour is limited by the FOV, which misses the shoulders and posterior neck. The doses calculated on the CBCT were compared to the reference dose using 3D gamma. Gamma was calculated within the combined high and low dose PTV (tumor plus nodes).
Results: The average 3D gamma within the combined PTV is 90.6%, 96.1% and 98.0% for the thorax scan group, using tolerances of 3%/1mm, 4%/1mm and 5%/1mm, respectively. The agreement using the head scanning protocol is only 79.8% at 5%/1mm. The accuracy of the CBCT homogeneous dose calculation is implied by the required tolerance of 5%/1mm to get an acceptable gamma of 98%. Our method could therefore detect PTV coverage reduction below 95%. Comparing nodal PTV coverage directly, the difference from the reference is 2.3% for thorax, 4.3% head.
Conclusion: Using a complete external contour is critical to accurate dose calculation on CBCTs. The thorax scan with its larger FOV is preferred over the typical head scan, especially when the PTV extends inferiorly. The PTV coverage can be sufficiently estimated using a homogenous dose calculation.