892 / 1082
S876
ESTRO 36 2017
_______________________________________________________________________________________________
were ≤3mm (SD≤4mm) in the three directions, while the
vector shift was 8.5±4.4mm (range: 0.0-24.2mm),
revealing the necessity of performing a daily CBCT.
Differences between the planned PTV and the
recalculated PTV coverage (V95%) were within 3.5%
proving the feasibility of VMAT in breast cancer treatment
(figure 2).
Figure 2: Dose recalculation on a DIBH-CBCT. Similar
results were obtained in all cases.
Conclusion
DIBH-CBCT was shown to be feasible with adequate image
quality, making possible to evaluate the efficacy of VMAT
skin flash tool approach also in DIBH condition.
Considering the relative short image acquisition time
required, a protocol for DIBH breast SIB treatments in 15
fractions with DIBH-CBCT daily position is starting in our
department.
EP-1638 Comparison dual image registrations for SBRT
treatment in central and peripheral tumour lung cancer
D. Esteban
1
, M. Rincón
2
, J. Luna
1
, A. Sánchez-
Ballesteros
2
, A. Ilundain
1
, L. Guzmán
1
, D. Gonsálves
1
, W.
Vásquez
1
, J. Olivera
1
, I. Prieto
1
, J. Vara
1
1
Hospital Universitario Fundación Jiménez Díaz,
Radiation Oncology, Madrid, Spain
2
Hospital Universitario Fundación Jiménez Díaz, Physical
Radiation Oncology, Madrid, Spain
Purpose or Objective
Four-dimensional (4D) Cone Beam CT (CBCT) became
available in clinics and has been used to quantify
localization precision and intrafraction variability of lung
tumour. Dual image registration is an option for the 4D-
CBCT image registration: a clip-box and mask registration.
The clip-box registration is the same as those in 3D CBCT,
while the mask registration is a new feature of Elekta XVI
CBCT system, which is a soft-tissue registration using a
soft-tissue volume called mask.
Published study of 3D CBCT using clip-box with different
landmarks led to different registration precisions for
peripheral and central lung tumours. The study aims to
compare if the different location of the tumour
(peripheral and central) has an impact on the precision of
the Elekta 4D-CBCT based automatic dual image
registrations using clip-box and mask for SBRT treatment
of lung cancer.
Material and Methods
A sample of 29 patients with diagnosed lung cancer and
treated with SBRT from 2014 to 2016 was obtained by
purposive sampling. Tumours were classified depending
upon his location. They were scanned on a Philips 4D CT
scanner and 8-phase images were reconstructed. GTV was
contoured on each phase and ITV was generated by
including the GTVs. PTV was created from ITV with a 5mm
margin. Then a clip-box (place on the spine) and mask
(margin of 0.5 cm on the PTV) were created allowing
translations.
Elekta XVI CBCT system using a 4D module was used to
acquire 4D-CBCT scans of the patients in treatment
position prior to radiation delivery and a total of 101 were
obtained. These images were registered with reference 4D
CT images using dual image registration and a "balance
clip-box mask" was generated (Imagen 1). The
registrations were reviewed by physicians, choosing the
treatment position in which better coverage of PTV is
performed. Possible treatments positions may be clip-box,
mask
or
intermediate
values
of
these.
Descriptive analysis and Student t test was performed.
Translational shifts calculated from dual automatic
registrations were compared with the different standard
registrations chosen by the physician, and the possible
impact of the different location of the tumours on the
precision was
studied.
Results
There are forty 4D-CBCT from central and sixty from
peripheral tumours. Differences of translational shifts
between automatic dual registrations and the standard
(difference=automatic
registration-standard)
was
obtained. The Table 1 lists the mean and standard
deviation of translational shifts directions.
Conclusion
Although the clip-box is created from a localized midline
structure, the mask is more accurate for central tumours.
The outcomes showed that using different dual image
registrations resulted in different registration precisions
for peripheral and central lung tumours, being favourable
in all cases for the mask with statistically significant
difference. This difference may be explained because clip-
box registration made gross alignments whilst the mask
made fine
adjustments.
EP-1639 Does intrafraction motion between vmDIBHs
during breast cancer treatment impact on delivered
dose?
M. Kusters
1
, F. Dankers
1
, R. Monshouwer
1
1
Radboud university medical center, Academic
Department of Radiation Oncology, Nijmegen, The
Netherlands
Purpose or Objective
Intrafraction motion simulations were performed in the
treatment planning system (TPS) to determine the
robustness of the VMAT comprehensive locoregional breast
planning technique for residual breathing motion. The
simulations are used to determine a threshold for the
warning signal of the in-room monitoring system for
excessive intrafraction motion deviations.
Material and Methods
Our current treatment plan technique consists of 6
opposing 24
o
VMAT beams in latero-medial and medio-
lateral direction. A CTV-PTV margin of 7 mm is used and
plans with a PTV coverage of V95% > 95% are clinically
accepted. IGRT is based on online CBCT.
The influence of the intrafraction breath hold motion on
PTV/CTV dose coverage and OARs dose is simulated in
Pinnacle TPS (Philips, Madison, WI, USA) with for each
beam and each fraction a random shift in the anterior-
posterior (AP) direction of each isocenter (with 6 beams