Purpose: We evaluated the AAA and AcurosXB(AXB) algorithms for varying Hounsfield-Units(HUs) in lung SBRT patients to investigate the dosimetric impact of an extremely low-HU caused by Chronic Obstructive Pulmonary Disease(COPD) on target coverage and Organs-at-Risk.
Methods: SBRT lung plans were generated using a Quasar Lung Phantom with a prescription of 12Gy×5 to the PTVs located centrally and peripherally. The lung average-HU was varied from -990 to 650 to simulate the low-HU region caused by COPD. The AAA plans were calculated with varying HUs in lung using the Eclipse 15.6-AAA and normalized to cover 95% of the PTV with the prescription dose. With the same HU and monitor-unit, the AXB plans were generated. The dose-to-medium in AXB plans was compared with the dose-to-water in AAA plans with the dosimetric indices including D99% for GTV/PTV, maximum dose of spinal-cord and rib, and lung V20. PTV coverage in AAA plans was analyzed for varying HUs caused by COPD, and evaluated with respect to the PTV coverage in AXB plans.
Results: The AAA overestimated the PTV coverage of > 10% for both targets in the HUs below -900. The central target was less impacted by varying HUs compared to the peripheral target. For an extremely low-HU of -990, the AAA overestimated the PTV D99% up to 45% compared to the AXB. However, the two dose algorithms were comparable in the normal lung HU range from -750 to -650. Spinal-cord and rib doses were relatively insensitive to the variation of HU for both targets. As the HU decreased below -900, the AAA overestimated the lung V20 and skin dose, and lung V20 was strongly dependent on varying HUs.
Conclusion: For COPD SBRT cases, the AAA overestimates the PTV coverage over 10% in the HUs below -900 and the AXB is more accurate than the AAA.