ESTRO 36 Abstract Book

150 / 1082

S137 ESTRO 36 2017
_______________________________________________________________________________________________
OC-0267 Automatic contour propagation of breast and
heart for re-planning in breast cancer radiotherapy
G.G. Sikkes1, L.T.C. Meijers1, A.N.T.J. Kotte1, I.E. Van
Dam1, B. Van Asselen1, H.J.G.D. Van den Bongard 1
1 UMC Utrecht, Radiation Oncology Department, Utrecht,
The Netherlands
Purpose or Objective
Adaptive radiotherapy (RT) can improve the radiotherapy
treatment when target volume, patient anatomy or
treatment position has changed during treatment.
Automatic, non-rigid, contour propagation in a clinical
setting is expected, in comparison with rigidly transferred
contours, to reduce delineation time and can increase the
efficiency of adaptive radiotherapy. For breast cancer
patients, adaptive radiotherapy is mainly performed due
to breast contour changes (e.g. oedema) or reduction in
seroma volume. The aim of this study was to determine
the applicability of non-rigid contour propagation of
breast (clinical target volume, CTVbreast) and heart for
re-planning in breast cancer RT.
Material and Methods
25 breast cancer patients, each with a RT planning-
CT(CT1) and a subsequent CT(CT2) taken before or during
(fraction 0-9) treatment, were selected for this
prospective study. For 11 of the 25 patients the changes
visible on CT2 were clinically acceptable, consequently,
no re-planning was needed. In 14 patients, 15 new RT
plans (including patient with bilateral breast cancer) were
performed, due to changes in target (n=11), treatment
position (n=2) or switch from Breath hold technique to free
breathing CT (n=1).
Heart and CTVbreast were manually delineated on CT1 by
a radiation oncologist according to clinical delineation
guidelines. Contours on CT1 were transferred to CT2 using
a rigid registration technique, equal to the standard
clinical procedure. The CT1 contours were also
transferred to CT2 via deformable automatic contour
propagation using ‘Advanced Medical Image Registration
Engine’(ADMIRE research software v1.12/v1.13.3/v1.13.5,
Elekta AB, Sweden).
The rigidly transferred contours as well as the deformable
propagated contours were compared to the clinical
contours on CT2 using comparison measures target
volume, DICE, Hausdorff distance and mean distance.
Results
The volume, DICE, Hausdorff distance and mean distance,
presented in table1 and figure1, depict differences
between the deformable propagated and rigidly
transferred contours relative to the clinical contours. 14
of the 15 patients receiving new RT plans (including
patient with bilateral breast cancer) had a higher DICE, a
smaller Hausdorff distance and mean distance and had less
volume differences between the deformable propagated
CTVbreast and heart contours relative to the clinically
used contours in contrast to the rigidly transferred
contours.
Conclusion
These results shows that the deformable propagated
CTVbreast and heart contours are, as expected, more
close to the clinical used contours than the r igidly
transferred contours. Therefore, deforma ble contour
propagation can reduce delineation time and can be used
to optimize the workflow of adaptive RT for breast cancer.
OC-0268 Volumetric comparison between PET/CT and
CT simulation for target delineation in esophageal
cancer
E. Jimenez-Jimenez 1, P. Mateo s2, N. Aymar1, I. Ortiz1, R.
Roncero1, M. Vidal1, M. Gimenez3, F. Romero4, S.
Sabater5
1Hospital Universitari Son Espases, Radiation Oncology
Department, Palma de Mallorca, Spain
2Hospital Can Misses, Medical Physics Department, Ibiza,
Spain
3Hospital Universitari Son Espases, Nuclear Medicine
Department, Palma de Mallorca, Spain
4Hospital Universitari Son Espases, Medical Physics
Department, Palma de Mallorca, Spain
5Complejo Hospitalario Universitario de Albacete,
Radiation Oncology Department, Albacete, Spain
Purpose or Objective
FDG-PET/CT has proven to be useful in the staging process
of esophageal tumors. However, evidence supporting the
use of FDG-PET/CT in the tumor delineation process and
radiotherapy planning is limited.
Our objective was to compare the volumes defined by PET-
CT vs. CT simulation in esophageal planning.
Material and Methods
Nineteen esophageal carcinoma patients were referred for
concomitant radio-chemotherapy with radical or
neoadjuvant intent.
Each patient underwent CT and FDG-PET/CT for
simulation treatment in the same treatment position. Two
separate GTVs were defined; one based on CT data alone
(GTV-CT) and another based on combined PET/CT data
(GTV-PET/CT). Volume sizes for both data sets were
compared and the spatial overlap was assessed by the Dice
similarity coefficient (DSC), which represents the ratio of
volume overlap between 2 contours.