51 / 146
was identified for each patient considering 5 mm margins and stored to
be used during treatment.
A treatment plan was individually optimized for each CT series and the DIBH
plan’s advantage was evaluated in terms of heart mean dose and other do-
simetric constraints before approving DIBH plan.
The imaging procedure consists of FB CBCT acquisition with online cor-
rection and AP (MV) – LL (kV) imaging during DIBH (first 3 fractions). After
systematic error correction, at the 4th fraction and every 6 fractions, portal
image of DIBH treatment fields is acquired.
It was calculated mean shifts resulting from online match along AP, CC, LL
axes, for CBCT (105 acquisitions) in FB, for 2D-images of setup-fields (86)
and treatment fields (59) in DIBH.
Results:
By analysis advantages and disadvantages, 17 pts of 24 eligibles
were sent to DIBH treatment. The benefit in terms of heart mean dose results
to 1.39 Gy (range 0.35–4.87 Gy).
FB CBCT shows shifts (in mm) along AP, CC, LL axes respectively of 0.1
±
4.9,
−
3.2
±
4.5,
−
0.1
±
4.5.
Analyzing DIBH setup-fields images, mean difference (in mm) along AP
results 1.8
±
3.3, CC
−
2.2
±
4.4, LL
−
0.9
±
2.9; the root mean square of the
SD’s of shift of all pts represents the random error and results along AP
3.0 mm, CC 4.1 mm and 2.4 mm along LL.
The analysis of the treatment field images show mean shifts (in mm) along
AP of 2.5
±
2.6, CC of
−
0.5
±
4.6 and LL of
−
2.1
±
2.6; the random error is
2.6 mm (AP), 3.4 mm (CC) and 1.8 mm (LL).
Conclusion:
The width of the gating window used for DIBH treatments,
as revealed by analysis of all available images, has proved suitable for our
setup procedures.
http://dx.doi.org/10.1016/j.ejmp.2016.01.158A.155
INFRARED-GUIDED PATIENT SETUP FOR BREAST CANCER PATIENTS:
COMPARISON WITH CBCT
S. Naccarato
*
, a ,G. Sicignano
a ,F. Ricchetti
a ,S. Fersino
a ,A. Fiorentin
o a ,N. Giaj Levra
a ,R. Mazzola
b ,F. Along
i a ,R. Ruggier
i a .a
Radiation Oncology
Sacro Cuore – Don Calabria Hospital, Negrar, Verona, Italy;
b
Radiation Oncology
School, University of Palermo, Palermo, Italy
Introduction:
To quantify the setup discrepancies between a non radio-
graphic Infrared System and CBCT in breast cancer patients receiving IMRT
with SIB.
Materials and Methods:
In 9 breast cancer patients the ExacTrac Infrared
System (ET-IR) detected translational (anterior-posterior AP, superior-
inferior SI, left-right LR) and rotational (Yaw, Roll, Pitch) setup errors, guided
by 5–8 reflective markers positioned on the thorax of the patients. Such
computed shifts where applied to an ExacTrac 6D-robotic couch. Then, a
verification CBCT on a TrilogyTM (Varian Inc) linac was acquired with low-
dose protocol. The patients’ 3D CBCT data were retrospectively analyzed
to determine translational and rotational residual shifts after an automat-
ic registration Such registration was focused on the soft tissues included
in a box which contained sternum and the ribs, and by using the optimal
brightness/windows level to detect the skin edge of the irradiated breast.
The discrepancies between ET-IR and CBCT were quantified in term of sys-
tematic (S) and random (σ) errors. The non parametric Wilcoxon matched
pair test was used to test the setup differences (D) caused by the two dif-
ferent systems. Statistical significance was considered at p
<
0.05.
Results:
Translational discrepancies in terms of S and σ were
<
1.4 mm and
2.8 mm respectively, while rotational discrepancies in term of (S) and (σ)
were
<
1.1° and 1° respectively. Significant (median
±
SD) were found to be
statistical significant only in the AP direction D_AP
=
(1.3
±
2.2) mm
(p
=
0.001), and for Roll, D_Roll
=
(–0.6
±
1.3)°mm (p
=
0.002) and Yaw,
D_Yaw
=
(0.2
±
0.9)° (p
=
0.002) rotations. However, also in this cases the
magnitude of the differences was always below our Non Action thresh-
olds (
<
2 mm;
<
1°).
Conclusions:
This study indicates that there is a good agreement between
ET-IR system and CBCT. Hence, for breast cancer patients, the IR-ET guided
setup represents a reliable and without added dose alternative to radio-
graphic imaging.
http://dx.doi.org/10.1016/j.ejmp.2016.01.159A.156
TOWARD THE DEFINITION OF A DOSIMETER RELATED GAMMA METRIC
TO IDENTIFY UNACCEPTABLE VMAT PLANS
A. Negri
*
, A. Scaggion, M.A. Rossato, M. Paiusco.
Istituto Oncologico Veneto
IOV-IRCCS, Padova, Italy
Introduction:
VMAT pre-treatment QA is usually performed comparing mea-
sured and calculated dose distributions in phantom by means of gamma
index (GI) metric. This procedure assumes that the differences between mea-
sured and calculated dose distribution reflect those in patient regardless
of the QA dosimeter. The aim of this study was to compare the gamma index
passing rate (GP%) evaluated in different QA phantoms to those evaluated
in patient and to establish a phantom related GP% acceptance level.
Materials and Methods:
For 30 H&N cancer patients, VMAT delivered dose
distribution was simulated modifying the RTPLAN DICOM file to contain
delivery errors tracked by MLC and gantry log files. Planned and deliv-
ered dose distributions were calculated in patients and in different
mathematical phantoms reproducing the geometries of some widely avail-
able QA systems (planar, cylindrical surface and fully 3D cylindrical). Mean
GI and GP% evaluated in patient were correlated with those in each single
phantom. ROC analysis was applied to define a phantom specific GP% thresh-
old reflecting a predefined acceptance level set in patient geometry.
Results:
Mean GI and GP% strongly depend on the dosimeter geometry. In
case of planar devices the inaccuracies detected within the phantom do
not absolutely reflect those occurring in patient. Moreover the GI map
changes with the depth of measurement. Conversely, pre-treatment QA per-
formed with volumetric and cylindrical 2D geometry shows a strong
correlation with gamma analysis performed in patient. ROC analysis reveals
that acceptable dose differences in patient result in different values of GP%
depending on the considered phantom.
Conclusions:
The choice of the QA phantom deeply influences the gamma
metric. A general a priori criterion for the acceptance of a pre-treatment
QA cannot be used. The study presents a method to express preferences
on the phantom geometry for patient specific QA and to evaluate the optimal
geometry related GI threshold.
http://dx.doi.org/10.1016/j.ejmp.2016.01.160A.157
PRELIMINARY STUDY FOR COMPARISON BETWEEN TWO TREATMENT
PLANNING SYSTEMS FOR VOLUMETRIC MODULATED ARC THERAPY
R. Nigro
* , a , b ,M.C. Pressello
a , c ,E. Bellan
d .a
Casa di Cura ARS Medica, Roma,
Italy;
b
Department of Radiotherapy Oncology, San Camillo de Lellis, Rieti, Italy;
c
San Camillo Forlanini, Roma, Italy;
d
Ospedale S. Maria Misericordia, Rovigo,
Italy
Purpose:
The aim of this study is to investigate the modeling of two treat-
ment planning systems for VMAT treatment with different modulation.
Materials and Methods:
Two different plans were selected with different
modulation request. Plan 1 was a simpler volume and Plan 2 needed a high
level modulation. Both VMAT plans were generated either with Monaco
TPS v.3.20 (CMS-Elekta) or with Pinnacle TPS v.9.8 (Philips Medical Systems).
Both TPS were clinically validated according to AAPM task group 119.
The plans were optimized to obtain the same planning objectives on PTVs
and OARs and were delivered using 6 MV photons with an Elekta Synergy
linear accelerator equipped with Agility MLC. A dosimetric validation was
performed using the Delta4 diode array phantom (Scandidos, Uppsala,
Sweden). The 3D gamma method, with criteria 3% 3 mm, was used to
compare the measured and calculated dose distributions.
Results:
Plan 1 exhibited a dose deviation with a mean value of 0.3% in
Monaco plan delivery and
−
1.3% in Pinnacle. The quality control results gave
a passing rate
>
98% for the gamma index for Monaco and a passing
rate
>
92% for Pinnacle. Plan 2 has a dose deviation with a mean value of
−
1.8% in Monaco plan delivery and
−
2.7% in Pinnacle. Quality control results
presented a passing rate
>
96% for the gamma index for Monaco and a
passing rate
>
86% for Pinnacle.
Conclusion:
Monaco provides better dosimetric agreement than Pinna-
cle. The lowest dosimetric agreement results were obtained for the plan
where the highest dose-gradients were observed. The quality control results
gave a high passing rate
>
97.4% for the gamma index for Pinnacle in 90%
of VMAT treatment nevertheless observed differences could be reduced with
e46
Abstracts/Physica Medica 32 (2016) e1–e70