890 / 1096
S874
ESTRO 36
_______________________________________________________________________________________________
respectively. DIFGI is benchmarked against Varian’s RPM
until final validation of the device, but it is compatible
with all treatment units and CTs.
Fifteen left-sided breast cancer patients have been
recruited until now. If heart constraints can’t be fulfilled
in free-breathing (FB), then patients are trained and
undergo a second CT scan in DIBH using the DIFGI.
The stability, repeatability, reproducibility and reliability
of the method are studied. Two radiopaque markers, one
on the mediastinum tattoo and another along the back,
serve as a reference to measure breath amplitude (Fig.2).
The stability and repeatability are measured on the DIBH
CT scan. The reproducibility mean value, systematic, and
random errors are determined by using daily kV images
and weekly CBCTs. The reliability of the device is
calculated as the failure ratio compared to RPM.
We also analyse Dmean, V30 (cm
3
), and V25 (%) for the
heart in both techniques.
Results
Stability and repeatibility are below 1.7 and 3.3 mm in all
cases, respectively. Repoducibility mean value is 1.7 mm,
systematic error is 0.5 mm, and random error is 0.9 mm.
DIFGI reliability is 95%. All failures are human errors
occurred during the learning period.
Dosimetric benefits compared to FB for the heart are: 3.0
vs 6.7 Gy for mean dose, 14.9 vs 53.4 cm
3
for V30, and 2.8
vs 9.5% for V25.
Conclusion
DIFGI is a simple, friendly, low-priced external
respiration-monitoring device compatible with all
treatment units and CTs. The preliminary results of the
stability, repeatability, reproducibility, and dosimetrical
benefits are encouraging. The reliability of the device
depends on human intervention so we plan to interlock it
with the treatment unit.
EP-1617 Reproducibility and stability of vmDIBHs
during breast cancer treatment measured using a 3D
camera
M. Kusters
1
, F. Dankers
1
, R. Monshouwer
1
1
Radboud university medical center, Academic
Department of Radiation Oncology, Nijmegen, The
Netherlands
Purpose or Objective
To accurately perform voluntary moderately deep
inspiration breath hold (vmDIBH) radiation therapy it is
essential to determine the position of the chest wall at the
start of treatment and to monitor deviations during
treatment.
An in-house developed real-time automated monitoring
system of the respiratory motion is implemented to verify
the reproducibility and stability of the vmDIBH during
breast cancer treatments.
Material and Methods
Patients with left sided breast cancer are guided to
perform vmDIBH assisted by verbal instructions and an
additional aid called the ‘breathing stick’ [1].
A 3D Kinect v2 camera (Microsoft, USA) was mounted in
the treatment room to visualize the patient on the
treatment couch. Software was developed to track and
visualize the anterior–posterior motion of a small area of
the surface of the thoracic wall in real time, allowing RTTs
in the treatment room to verify the reproducibility and
stability of the breath holds during treatment.
The data of ten patients was analysed for reproducibility
and stability. The formulas for reproducibility and stability
were derived from Cerviño et al. with minor adaptations
[2]. For reproducibility the standard deviation of the mean
of each DIBH level was used. For stability all breath holds
were fitted by first order polynomials, the slopes were
multiplied by their breath hold lengths to find a range and
all these ranges were averaged.
Results
A typical example of reproducible and stable vmDIBHs
during a treatment fraction of a patient is shown in Figure
1.
Figure 1:
Typical example of five reproducible and stable
vmDIBHs for one treatment fraction
The analysed reproducibility and stability of vmDIBH
treatment for then patients using the breathing stick is
shown in Figure 2. The median reproducibility and stability
were 0.9 mm and 1.1 mm, respectively.
Figure 2:
Clinical measurements of the reproducibility and
stability of the vmDIBHs for ten patients
Conclusion
The reproducibility and stability of the chest wall can be
accurately measured using the in- house developed
monitoring system. vmDIBH in combination with the
breathing stick shows good stability and reproducibility
which are comparable to the results in the study of
Cerviño et al. [2].
In this work the current results are limited to ten patients;
we continue to, acquire more data for future analyses.
The breathing stick is routinely used in our clinic;
currently we use the breath hold monitoring system to test
whether using this tool has an added value.