Program Information
Experimental Verification of Eclipse AAA and Acuros Lung Dose Calculations
S Shiraishi*, M Grams , J Antolak , S Park , K Olivier , Y Garces , L Fong de los Santos , Mayo Clinic, Rochester, MN
Presentations
PO-BPC-Exhibit Hall-10 (Saturday, March 18, 2017) Room: Exhibit Hall
Purpose: To experimentally validate dose calculation accuracy in a lung-like material for the Eclipse AAA and Acuros algorithms. Shrinking lung tumors were simulated to investigate the resulting changes in dose distributions.
Methods: A water phantom using cylindrical cork sections as lung surrogates was fabricated. The lung sections fit within the acrylic water phantom which simulated a chestwall. Large and small tumors were machined from solid water and implanted separately within the lung insert. An insert without tumor was made to simulate a “healthy” lung. The inserts were split along their long axis and Gafchromic EBT3 film was placed within the lung and tumors to measure a two-dimensional dose distribution. Calculation accuracy was investigated using AP/PA fields delivering 200cGy to the “healthy” lung insert using various field sizes. VMAT treatment plans were then delivered to the phantom containing the tumors. To simulate the largest possible change in dose distribution where the tumor completely resolved, the plan designed for the largest tumor was delivered to the “healthy” lung insert. Differences between measurements and calculation were evaluated by dose profile comparison and 3%/2mm gamma passing criteria.
Results: The AP/PA fields showed that AAA generally underestimated the dose by 2-3% while Acuros agreed within 1%. The VMAT plans showed AAA slightly underestimated dose in the GTV to PTV margin while Acuros agreed well with the measurements. For the large tumor plan, AAA and Acuros gamma passing rates were 100%, and for the small tumor plan, the passing rates were 95% and 100%, respectively. When the large tumor plan was delivered to the “healthy” lung insert, dose within the PTV increased by only 5% compared to the original plan.
Conclusion: Both AAA and Acuros are capable of accurately calculating dose in a lung-like material. Tumor regression alone does not substantially change the original isodose distribution.
Contact Email: