ESTRO 36 Abstract Book

809 / 1096

S793 ESTRO 36
_______________________________________________________________________________________________
Material and Methods
Five preliminary sensitivity test scenarios were created
from a baseline plan modifying each MLC segment by
introducing increasing shifts in leaves positions (0.5 mm -
2 mm). The baseline and test plans were delivered to an
Octavius 1000SRS array (PTW) as well as to EBT3 films. An
average correction was applied to 1000SRS results to
account for the response dependence on source-detector-
distance (SDD) [O. Blanck . Phys Med 2016]. The same
five test plans were delivered a second time to the
1000SRS re-orienting all beams perpendicularly to the
array (nominal position) to eliminate SDD and angle
dependence. As a second step 40 clinical MLC plans
optimized for various treatment sites (liver, spine,
prostate) were delivered to the liquid ion chamber array
for patient-specific QA using both the clinical beam
orientations and the beam “nominal position”. For the
latter only a subset of segments(18-21) was selected.
Finally, for 15 out of 40 clinical plans a film-based DQA
was also performed. All results were analyzed using (2%,
2mm),( 3%, 1mm) and (2%, 1mm) gamma index criteria [O.
Blanck . Phys Med 2016].
Results
The pass-rate reductions from the baseline,obtained
delivering the five test plans, are shown in fig.1 for (3%,
1mm) gamma criteria. The Octavius 1000SRS showed a
good sensitivity to simulated delivery errors with pass-rate
reductions increasing from 1.7% to a maximum of 43% with
increasing leaves shifts (0.5 mm - 2 mm). Similar
sensitivity was observed when the beams were re-oriented
in the nominal position geometry. The pass-rate
reductions observed with films showed a more irregular
trend, and the maximum reduction was 16%. The average
pass-rates obtained over clinical plans are shown in fig.2,
for the three gamma index criteria. The mean values
obtained by the 1000 SRS array, using both the clinical and
nominal beam geometry, and by film-dosimetry are all
above 92%, when using 3%, 1mm criteria. Differences
among the mean pass-rates observed for the three
measurement modalities were not statistically significant
(p> 0.1, t-test)
Conclusion
The results confirm that the 1000SRS array is reliable for
pre-treatment QA of CyberKnife MLC plans. The test
scenarios highlighted a higher sensitivity to small leaves
shifts than what observed by film dosimetry. The gamma
pass-rates obtained for clinical plans DQA were
comparable to film pass-rates. The possibility to use the
beam nominal position was validated and can be an
alternative to eliminate SDD and angle dependence.
EP-1484 Validation of ptw’s diamond as alternative
method for the imrt-vmat pretreatment verification
A. Niebla Piñero1, B.C. Portas Ferradás1, P. Rosa
Menéndez 1, J.A. Merino Gestoso 1, M. Rodríguez Pazos1, D.
Pardo Calvo1, B. Sobral Aspérez1
1Hospital Universitario Nuestra Señora de Candelaria,
Medical Physics, Santa Cruz de Tenerife, Spain
Purpose or Objective
The aim of the study is the validation of the software
DIAMOND, as alternative method to ion chamber point
dose measurements to verify prostate’s IMRT-VMAT plans.
For this purpose, we have selected 109 IMRT and 65 VMAT
treated plans between the 12/09/2013 to 08/16/2016. We
have compared the results using Diamond with the ion
chamber results in the pretreatment verification. Using a
ROC analysis we have obtained the new tolerances to
apply in our QA program.
Material and Methods
We have selected 109 IMRT step & shoot plans calculated
with CMS XiO and 65 VMAT calculated with CMS MONACO.
This plans were calculated over the own patient’s CT and
also over the I’mRT phantom’s CT used for the point dose
verification. These plans were sent to DIAMOND to make
the recalculation in two points: One of them in a high
dose-low gradient region (P1), and the other in a high