ESTRO 36 Abstract Book

894 / 1082

S878 ESTRO 36 2017
_______________________________________________________________________________________________
5mm is significant but less pronounced. In particular, the
gain for the OARs is small when moving from 3 to 0 mm
while a 3mm margin increases the average D98 by 0.51Gy.
EP-1641 Intra-fractional CBCT validation of a 6D couch
to facilitate precision RT of head and neck cancer
A. Bertelsen1, C.R. Hansen1,2, M. Nielsen1, J. Eriksen2,3,
N. Gyldenkerne3, J. Johansen2,3, S.L. Krogh1, J.
Westberg1 , C. Brink1,2
1 Odense University Hospital, Laboratory of Radiation
Physics, Odense, Denmark
2 University of Southern Denmark, Department of Clinical
Research, Odense, Denmark
3 Odense University Hospital, Department of Oncology,
Odense, Denmark
Purpose or Objective
Small planning margins for head and neck (H&N) cancer
patients reduces toxicity and can be assisted by stable
fixation equipment and IGRT including online corrections
for translations and rotations (6D). This clinical study uses
intra-fractional CBCT (iCBCT) to investigate the precision
and, accuracy of 6D couch usage, and if the patients
compensate for the rotational corrections.
Material and Methods
H&N patients receiving standard fractionated IGRT were
included in this study. The inclusion occurred in a three
month period after an upgrade of the 6D couch system
(Elekta, Hexapod). After online couch correction based on
a pre-treatment CBCT (pCBCT) additional setup
verification was performed using iCBCT acquired during
single arc VMAT.
The residual uncertainty of the 6D couch correction was
assessed by comparing the pCBCT and iCBCT registration
of the spinal cord.
The residual setup error of the target volume ( from the
iCBCT) was used to calculate the population-based
systematic and random uncertainty given as the standard
deviation (SD) of patient mean values and SDs.
Correlations between residual errors and initial setup
errors as well as patient specifics (age, weight, BMI,
performance status (PS)) were tested using Spearmann’s R
or Kruskall-Wallis (PS) test. For correlations the standard
5% statistical significance level was Bonferroni corrected.
Results
In total 44 patients were included resulting in 1174
iCBCTs.
On average, the initial setup errors prior to use of the 6D
couch were small, but had statistically significant mean
values different from zero (table 1). Deviation in Z is due
to standard adjustment of the fixation. The X rotation
value is likely caused by the Hexapod calibration
performed without load on the couch. Large SD was
observed in the initial Y rotation.
The residual error shows that all setup uncertainties are
statistically significant reduced (Leveens test, p<0.001).
However, a large systematic residual error of 0.4mm was
observed in the Y-direction. This might be related to a
significant strong correlation between initial X-rotation
and residual Y-translation (figure 1). Small SDs for both
rotations and translations (table 1) indicate that patient
compensation for rotational corrections is modest. The
larger SD values for X and Z are related to significant
strong correlations to the initial Y rotation (R=0.77 and
0.68, respectively).
No significant correlations were observed between
patient-specific parameters and pre or post setup errors.
Based the translational systematic and random errors a
minimum CTV to PTV margin of 2mm is required (Van Herk
et al. IJROBP 2000). However this estimate does not
include compensation for rotations of elongated targets.
Conclusion
The Hexapod system performed well and patients did not
seem to compensate for rotational corrections. However
due to related uncertainties rotations should be kept at a
minimum. The present setup reduces the systematic and
random uncertainties which make use of small treatment
margins plausible.
EP-1642 Patient-specific transperineal ultrasound
probe setups for image guided radiotherapy
S. Camps1,2, F. Verhaegen3, P. De With1, D. Fontanarosa2,3
1Eindhoven University of Technology, Faculty of
Electrical Engineering, Eindhoven, The Netherlands
2Philips Research, Oncology solutions department,
Eindhoven, The Netherlands
3GROW - School for Oncology and Developmental Biology,
Department of Radiation Oncology MAASTRO,
Maastricht, The Netherlands
Purpose or Objective
Despite the many advantages of ultrasound (US) imaging
(e.g. safety and high-contrast soft tissue imaging) the use
of US in image guided radiotherapy (IGRT) workflows is not
widespread. This can primarily be attributed to the need
of a skilled operator for US volume acquisition. We
introduce an algorithm that provides the operator with a
patient-specific probe setup that allows good anatomical
structure visualization based on clinical requirements.