ESTRO 2023 - Abstract Book

S1445

Digital Posters

ESTRO 2023

The AAA v. 16.1 calculation model was configured for the 6 MV FFF energy, including the PDIP add-on (AAA-PDIP). By keeping T= 0.01, optimal DLG value for the AAA-PDIP algorithm was achieved by comparing predicted versus measured portal images for the well stablished Chair pattern. Twenty 6 MV FFF clinical plans (70 fields in total) were retrospectively verified using the AAA-PDIP algorithm. VMAT technique was used in 13 cases, and IMRT in the remaining cases. These plans showed 3D gamma passing rates (GPRs) ≥ 94% when were verified using our primary system (PTW Octavius 4D) and using the 3% global/2 mm (with a 10% cut dose) criteria recommended by the AAPM TG-218 report. The plans were also verified using the PDIP v. 16 algorithm, with non-tunable option of the DLG and T values. Same acquired portal images were analyzed with AAA-PDIP and PDIP. The PDIP was configured using the van Esch package [J Appl Clin Med Phys. 2013 Nov 4;14(6):4286]. 2D GPRs and average gamma indexes (AGIs) resulted from the AAA-PDIP and PDIP-based verifications were compared. The 3% global/2 mm (with a 10% cut dose) were applied. A 2 tailed Student t-test ( α = 0.05) was used to evaluate whether there is a significant difference between both algorithms. Results 1) Graph 1 shows the DLG tuning. A value of DLG= 0.1 mm provided the highest GPR and lowest average gamma index for the Chair test. The DLG= 1.25 mm used by the Acuros algorithm resulted in the worse GPR for the AAA-PDIP model. 2) Over the 70 fields corresponding to the 20 cases, GPRs of 99.0% ± 2.7% and 93.7% ± 9.7% were obtained for the AAA PIDP and PDIP models, respectively. AGIs of 0.2 ± 0.1 and 0.4 ± 0.2 were found for the AAA-PIDP and PDIP models, respectively. AAA-PDIP showed significantly (p< 0.001) better GPR and GAI values than the PDIP algorithm. Graph 2 shows the GPR and AGI values averaged over the fields involved in each case.

Conclusion