Error detection thresholds of diode detector arrays in SRT patient specific QA
PO-1555
Abstract
Error detection thresholds of diode detector arrays in SRT patient specific QA
Authors: Mark Tutty1, Marco Langhans2, Ann-Kathrin Stedem3, Christoph Kleefeld4, Andreas A. Schönfeld5
1The Beacon Hospital, Radiotherapy , Dublin, Ireland; 2Hospital of Wilhelmshaven, Radiation Oncology, Wilhelmshaven, Germany; 3Heinrich-Heine Universität Düsseldorf, Science, Düsseldorf, Germany; 4Nation university of Ireland, Galway, Medical Physics, Galway, Ireland; 5Sun Nuclear Corp, Medical Physics, Melbourne, USA
Show Affiliations
Hide Affiliations
Purpose or Objective
The aim of
this study was to compare the SRS MapCHECK and ArcCHECK detector arrays (both Sun Nuclear Corp., Melbourne, FL, USA), operated with and
without high density merge, in their ability to recognize
artificially introduced MLC errors in SRT treatment plan deliveries and to
establish recommended gamma criteria for each device.
Material and Methods
Various
treatment plans with target volumes ranging between 1.5 cm and 3 cm diameter were
created in Eclipse TPS (Varian Medical Systems, USA) at The Beacon Hospital,
Dublin. These treatment plans were modified by introducing predetermined MLC
and/or jaw errors according to Kim et al. (2014). The introduced MLC and jaw
offsets ranged from 0.5 mm to 2 mm.
To determine the
clinical impact of these misalignments, dose volume histograms (DVHs) of the
modified plans were calculated and compared to the DVHs of the respective
unmodified plans.
The modified
treatment plans were then irradiated onto the SRS MapCHECK array contained
within the StereoPHAN End-to-End phantom (also Sun Nuclear Corp.) as well as
the ArcCHECK array. Additional, shifted measurements were taken to double the
effective spatial resolution of the arrays (high density merge). The measured
dose maps of the modified treatment plans were compared to the TPS calculated
dose map of the unmodified treatment plan.
The accompanying
SNC Patient software was used to establish the gamma passing rates. The gamma
analysis was carried out using various gamma criteria ranging from 1%/1mm to
3%/3mm, to see which were sensitive enough to pick up on the misalignments and
which may be oversensitive. The passing criteria was set to be 95% for a plan
to be determined viable for treatment. The results each array was compared to
each other to see which array was successfully able to pick up on the smallest,
clinically relevant misalignment.
Results
By comparing
DVHs of modified plans to the original plans, it could be seen that misalignments
as small as 1 mm induced in the system can have a detrimental effect on the PTV
coverage, which could potentially lead to an under-dosage to the treatment
volume, therefore any gamma criteria selected would be required to be able to
pick up on shifts of 1 mm.
The SRS MapCHECK
was able to pick up all introduced MLC errors using Gamma criteria of 2%/1mm. Doubling
the spatial resolution of the SRS MapCHECK did not affect the error detection
capability or passing rates. The ArcCHECK could detect introduced errors in
deliveries of treatment plans with larger target volumes of about 3 cm using
gamma criteria of 3%/1mm. However, the ArcCHECK was insensitive to introduced
errors in deliveries of treatment plans with small target volumes of about 1.5
cm.
Conclusion
The SRS MapCHECK
successfully detects clinically significant errors in patient specific QA of
SRT treatment plans. The ArcCHECK has shown to struggle in detecting errors in very
small fields, but performs well in larger field sizes.