ESTRO 36 Abstract Book

916 / 1082

S900 ESTRO 36 2017
_______________________________________________________________________________________________
DECT measurements of typical tissue-surrogate phantoms
and evaluated its uncertainty.
Results
The methodological uncertainty of electron-density
assessment for the alpha-blending method was found to
be below 0.15% for arbitrary mixtures of human tissue. In
the case of small abundance of high-Z elements, electron-
density results are positively biased, e.g. 0.5% for thyroid
containing 0.1% iodine (Z=53) by mass, which is due to the
K edge of the photoelectric effect. The calibration
parameters obtained from various published data sets,
showed very little variation in spite of diverse
experimental setups and CT protocols used. The
calibration uncertainty was found to be negligible for soft
tissue while it was dominated by beam hardening effects
for bony tissue.
Conclusion
The alpha-blending approach for electron-density
determination shows universal applicability to any mixture
of human tissue with a very small methodological
uncertainty (< 0.15%); and a robust and bias-free
calibration method, which is straightforward to
implement. We conclude that further refinement of
algorithms for DECT-based electron-density assessment is
not advisable.
EP-1674 Experimental investigation of CT imaging
approaches to deal with metal artefacts in proton
therapy
S. Belloni1,2, M. Peroni1, S. Safai 1, G. Fattori1, R. Perrin 1,
M. Walser1, T. Niemann3, R.A. Kubik-Huch3, A.J. Lomax1,
D.C. Weber1,4,5, A. Bolsi1
1 Paul Scherrer Institut, Center for Proton Therapy,
Villigen PSI, Switzerland
2 University of Bologna, Department of Physics and
Astronomy, Bologna, Italy
3 Cantonal Hospital Baden, Department of Radiology,
Baden, Switzerland
4 Inselspital, Radiation Oncology, Bern, Switzerland
5 University Hospital Zurich, Radiation Oncology, Zurich,
Switzerland
Purpose or Objective
Metal implants are challenging for proton therapy, mainly
because of beam hardening artefacts severely
compromising image quality of the planning CT. In fact,
they result in non-negligible uncertainties in Stopping
Power (SP) evaluation and significantly affect VOI
delineation accuracy. The aim of this study was to
compare different approaches to minimize the artefacts:
a manual approach based on delineation of the visible
artefacts, which was developed and is used clinically at
the Center for Proton Therapy (PSI), and the new tools
recently introduced in CT, such as SIEMENS Iterative Metal
Artefact Reduction (iMAR) and Sinogram Affirmed Iterative
Reconstruction (SAFIRE). Moreover, an experimental
verification of direct SP calculation from Dual Energy (DE)
images with iMAR has also been considered.
Material and Methods
A clinical treatment of a cervical chordoma patient was
reproduced on a head and neck anthropomorphic
phantom, which presents metal implants (titanium screws
and cage) in the area where the PTV was defined. An IMPT
plan with two anterior oblique and two posterior oblique
fields (dose per fraction 2 GyRBE) was optimized and
calculated on 7 different CTs which corresponded to the
different imaging approaches: no correction of artefacts,
manual correction, iMAR (each of these reconstructed
using Filtered Back Projection (FBP) and SAFIRE) and DE
together with iMAR. The delivered dose was measured
with EBT3 Gafchromic films, inserted in three sagittal
planes of the phantom included in the PTV area, and was
compared with the dose calculated on the different CTs
from machine log files. Local dose differences and gamma
maps were used to evaluate the results, taking into
account residual positioning errors, daily machine
dependent uncertainties and film quenching.
Results
We restricted the analyses to the 50% isodose and defined
A+10% and A-10% as the percentage area having percentage
differences higher (lower) than 10% (-10%). In general,
A+10% between calculated and measured dose distributions
were below 10% for plane 1 and 2 with the DE approach
combined with iMAR (Table 1). Maximum differences were
mainly located in the areas of steep dose gradients.
Focusing on the SAFIRE algorithms, the three methods
showed comparable results to the corresponding FBP
algorithms for plane 2 and 3. For plane 1, A +10% increased
to 24.8% for uncorrected approach, but SAFIRE was again
comparable to FBP when iMAR is used.
Conclusion
DE combined with iMAR shows potential for predicting SP
values and reducing metal artefacts. However, all
approaches provided comparable, and clinically
acceptable, results in terms of dosimetry accuracy. This
could be related to the uncertainties in the experimental
setup and in the measurements method (mainly use of
gafchromic films), which might be comparable to the
differences introduced by the metal artefacts correction
approaches. The planning approach with multiple fields
was robust against errors introduced by metal implants.
EP-1675 Influence of CT contrast agent on head and
neck VMAT dose distributions
L. Obeid1, J. Prunaretty1, N. Ailleres1, L. Bedos1, A.
Morel1, S. Simeon 1, P. Fenoglietto1
1Institut Régional du Cancer de Montpellier,
Radiotherapy, Montpellier, France
Purpose or Objective
Intravenous contrast agent injection during the patient CT
simulation facilitates radiotherapy contouring in the case
of head and neck cancers. However, the image contrast
enhancement may introduce discrepancy between the
planned and delivered dose. The aim of this retrospective
study is to quantify the variations of Hounsfield unites
(HU) and to investigate their effect on Volumetric
Modulated Arc Therapy (VMAT) dose distributions.
Material and Methods
Ten patients previously treated by VMAT techniques with
identical dose levels (70/60/50 Gy) were selected. For
each patient, two CT scans were performed, 2 min. (CT inj)
and 12 min. (CTdelay ) after Iomeron® 350 biphasic
intravenous injection (60 mL, 1mL/s followed by 90 mL, 2
mL/s after 30 s). The treatment planning (optimization
and calculation) was performed with CTinj using the Eclipse
TPS and two calculation algorithms (AAA® and Acuros
XB®). Two other treatment plans were recalculated with
the same parameters and CTdelay. The mean HU and the
iodine distribution were compared between the two scan
images in the PTV50, the parotids and the thyroid. A
dosimetric comparison using dose-volume histograms in
target volumes and OAR (thyroid, parotids) was