911 / 1082
S895
ESTRO 36 2017
_______________________________________________________________________________________________
planning strategies were explored: 1) A new optimized
IMRT plan on each weekly MR image (WeeklyAdapt) 2) A
new optimized IMRT plan on week four MR image only
(OneAdapt) and 3) Calculating the plan from the planning
MR on each weekly MR image (NoAdapt). The PTV coverage
for all ViewRay MR-IGRT plans were such that 95% of the
PTV received 100% the prescription dose. The differences
between accumulated doses on planning MR for all three
ART strategies were evaluated using the dose-volume
constraints for targets and critical structures.
Results
For PTV70, PTV60 and PTV54, as compared to the D95
coverage on planning MR, the differences in D95
accumulated doses between three ART strategies were
<2%. The maximum dose to both cord and brainstem
between WeeklyAdapt and only OneAdapt were very
similar and <1% as compared to the planning MR
values. However, NoAdapt Dmax for brainstem and cord
were >27% and 25% than the original planning MR Dmax
values respectively. The mean left and right parotid dose
remained very similar both WeeklyAdapt and OneAdapt
strategies although there were up to 10.7% increase in the
mean dose to the parotids with NoAdapt strategy.
Conclusion
This study demonstrated that no significant differences in
accumulated doses were observed between weekly ART
and only one ART at week four during MR IGRT of H&N
cancer patients. Further studies are needed to evaluate
benefits of daily online ART during MR IGRT.
EP-1668 Dose calculation accuracy using CBCT images
for head and neck VMAT
M.A. Carrasco Herrera
1
, B. Quintana
2
, J.M. Nieto
2
, F.J.
Luis Simon
1
, C. Santa Marta Pastrana
3
1
Hospital Universitario Virgen del Rocio, Medical Physics
department, Sevilla, Spain
2
Hospital Universitario Virgen del Rocio, Radiation
Oncology department, Sevilla, Spain
3
Universidad Nacional de Educación a Distancia UNED,
2UDepartment of Mathematical Physics and of Fluids,
Madrid, Spain
Purpose or Objective
Assessment of the differences between CT and CBCT based
dose calculation for a volumetric modulated arc therapy
(VMAT) in head and neck radiotherapy treatment.
Material and Methods
CBCT images of an Alderson RANDO phantom with tissue-
equivalent material were acquired in a Varian`s On-Board
Imager OBI (v1.5) installed on a Varian DHX accelerator,
using its standard-dose head protocol (100 kV and 2.5 slice
thickness). On the other hand, planning CT images were
acquired in a Toshiba Aquilion LB using the same phantom
and with our own clinical head and neck protocol (120 kV
and 3 mm slice thickness). Different OAR (Body, spinal
cord, parotids, mandible, oropharynx, dermis, an inner
ring with 1 cm thickness and shoulder) and a PTV were
delimitated. Additionally, reference points were inserted
over all these structures. All defined structures and points
were registered with the CBCT images by means of the
Varian rigid registration software. Both the delimitation of
volumes and the design of the treatment plan have taken
into account the limited field of view of CBCT (length 16
cm, diameter 25 cm). A head and neck VMAT plan has been
calculated in Eclipse (v10) using both sets of images. For
CT images, we only used a standard calibration curve and
3 different calibration sets of curves for CBCT images, i.e.,
standard, measured with a CATPHAN 504 phantom and
measured with a CIRS 062M head phantom placed between
head and neck RANDO slices. Dose and HU were calculated
in all reference points as well as dose-volume-histograms
for the anatomical locations for both CT and CBCT. A
gamma analysis was used for HVD comparison.
Results
The mean HU differences are less than 50 UH and the
relative dose differences are less than 3% for all the
calibration curves (Table 1) on every reference point over
all the structures.
The gamma (2%, 2 mm) DHV analysis shows an excellent
agreement for almost all the structures (>95%) (Image1).
Ring and dermis have gamma >85% .The non-pass regions
correspond to very low dose regions. The worst gamma
(>50%) corresponds to the left parotid because it is a very
small structure (10 cc) into a high gradient dose zone.
Furthermore, there is a difference of 1.8% on its volume
as measured on the CBCT and the CT images, probably due
to interpolation errors. These results are similar for all the
calibration
curves analysed.