961 / 1082
S945
ESTRO 36 2017
_______________________________________________________________________________________________
Results
The
basic
characteristics
of
couch-bending,
reproducibility and stability were within 0.10mm within
the treatment volume (see Fig.2 a, b).
The relative spatial deviation was smaller than 0.40mm
for rotations ranging up to 200° (see Fig.2 c). For different
vertical positions, the couch drifted less than 0.25mm for
2 different loads and rotations.
The differences between the prescribed and measured
absolute position were evaluated in terms of histograms
showing the overall 3D deviation. In 95% of all
measurement points the 3D accuracy was better than 0.63
mm (see Fig.2 d).
Regarding the weight-induced couch-bending no
correlation between the accuracy and payload could be
found.
Conclusion
The results show that the important spatial properties of
the patient positioning system are well within the
acceptable clinical tolerances. The very high
reproducibility of the PPS allows further optimization of
the absolute position. The measured datasets serve as
new input for a high accuracy calibration.
EP-1742 Optimisation and implementation of brain
CBCT templates; an institutional pilot study.
S. Petkar
1
, N. Lalli
2
, F. Solda'
1
, C. Gillies
2
, S. Moinuddin
1
,
N. Fersht
1
1
UCLH NHS Foundation Trust, Radiotherapy, London,
United Kingdom
2
UCLH NHS Foundation Trust, Radiotherapy Physics,
London, United Kingdom
Purpose or Objective
Volumetric imaging (CBCT) in brain has facilitated the use
of volumetric delivery techniques and reduction of
uncertainty margins by quantifying bone set-up variation,
and, by inference, OAR position.
The aim of this work is to optimise the current imaging
practice on a Truebeam Linac STx(2.5) for adult brain
tumour patients by investigating the relationship between
CBCT parameters, patient dose and image quality for both
bone and soft tissue.
Material and Methods
In our institution we currently image this cohort with daily
kV imaging and weekly CBCT. A daily ‘shift to zero’ is
applied for set-up variations of less than 3mm with
residual error evaluated on a weekly basis via a standard
Full-Fan, partial arc CBCT template (Table 1-Template 1)
The CBCT image quality was evaluated by comparing three
progressive dose-reduction trial imaging templates +/-
length reduction (Table1-Templates 2, 3, 4) with the
standard CBCT template.
Initial tests were carried on Catphan
®
and Rando
®
phantoms to assess image quality and scan artefact
limitations. Three consecutive skull-base meningioma
patients were then imaged with the four templates on
sequential weekly imaging sessions during treatment.
Each CBCT was reviewed by an expert group of a clinician,
two radiographers and two physicists to evaluate, by
consensus, the discernibility of the optic nerves,
ventricles, cranial bones, temporalis muscle, and the
external contour. In addition, the fidelity of the co-
registration to the skull-base anatomy (ROI) was assessed.
A standard threshold was applied throughout the
investigation.
Results
A total of 15 CBCT images were acquired and reviewed.
All structures were visible for each template except for
the ventricles which were assessed as indistinct with
templates 3 and 4 (Figure 1).
The fidelity of registration was satisfactory for each
template.
Conclusion
A length-reduction template can be utilised in brain
tumours providing the skull-base is included in the CBCT
ROI. In cases where the PTV is distant to the skull base,
length reduction should be used with care.
A concomitant dose reduction is also feasible unless the
discernibility of the ventricles is essential, such as when
changes in ventricular volume/position are of concern.
In our institution we will adopt a new imaging strategy for
brain tumour patients employing template 4 for all benign
skull based tumours and retaining template 1 (standard)
for all other lesions.
Further similar work will be carried out to optimise the
CBCT protocols in other cohorts of patients (e.g.
paediatrics).
EP-1743 Evaluation of proton grid therapy in
challenging clinical cases
T. Henry
1
, A. Valdman
2
, A. Siegbahn
1
1
Stockholm University, Department of Medical Physics,
Stockholm, Sweden