FREQUENTLY ASKED

QUESTIONS
Patient-Specific Quality Assurance

Contents
Identifying Risks To Patient Safety................................................................................ 2
Log File-Based QA............................................................................................................. 2
EPID-Based PerFRACTION™............................................................................................ 3
Workflow Efficiency........................................................................................................... 5
Practical Concerns............................................................................................................ 5
IT & Setup Considerations............................................................................................... 7
Technical Considerations................................................................................................ 8
References.......................................................................................................................... 9
 Frequently Asked Questions

Identifying Risks to Log File-Based QA

Patient Safety Q: Does PerFRACTION™ use log files?
A: When configured for EPIDs, PerFRACTION uses
Q: What are the most common sources of machine log files for some information, but primarily
error during treatment? in conjunction with EPID measurements. Specifically,
A: Most clinicians agree that the patient is the largest monitor chamber dose rate and output data (via log
source of uncertainty in any treatment delivery. One study files) are used with independent MLC leaf position
confirmed that patient set-up, anatomy changes, and measurements acquired using the EPID.
isocenter placement were the most common sources of Only in cases when the EPID is not deployable for
error.1 The same study showed in-vivo EPID analysis to specific patient treatment beam(s), the imaging
be an efficient and effective method for detecting these technique for EPID transit dosimetry is not compatible
patient-related errors. with 3D reconstruction, or EPID data collection has not
been configured will PerFRACTION perform log-based
Q: Can rigorous machine QA negate the need 3D dose reconstruction (if enabled).
for patient-specific measurement QA?
A: Rush University demonstrated that even when linacs Q: Can PerFRACTION be PATIENT QA:

pass stringent TG-142 monthly QA, there can still be large configured to use only WHAT LOG
FILES MISS

log files?
INTRODU

dose errors (greater than 10%) in patient specific QA.2 The

CTION
With the
rapid adva
introduce ncement
d into the of radiation
Assuranc process, therapy treat
e. In the drive placing incre ment mod
available, for efficiency ased dem alities, new
and on med com
including , many com ical physicists plexities are

study suggests patient-specific QA is still necessary, and

value, but measuremen mercial optio
techniques t and calcu ns for patie and Qual
using log lation-based ity
summari nt-specific

A: Yes. PerFRACTION
es of publ file analy methods. QA are
alone woul ications and sis only can Both meth
d not have clinical case miss signi ods provide
identified s where the ficant error
issues with use of log s. Following
potential file-based are four
impact to patient-sp
treatmen ecific QA
“Mo t.
nitoring daily

can be configured to use

errors usin MLC position
g trajector al

that, “Unacceptably large changes in dose delivered are

EPID mea y log files “A clinically
sure and obs
VMAT deliv ments for IMRT and between ima erved discrepancy
eries” ge-based
Agnew et MLC position and log-base
al.,
Care Centre Radiotherapy Physic s” d

only log files for 3D dose

, Belfast Health s, Northe Neal et al.,
Med. Biol. and Social rn Ireland Depart
59 (2014) Cancer Virginia, Charlo ment of Radiat

possible… despite the machine passing routine QA.” Their

N49-N63 Care Trust,
UK. Phys. ttesville, Virgini ion Oncology,
a. Med. Phys. University of
43 (6), June
2016

WHAT THE
EPID SEES Illustration
of leaf position

reconstruction, but users

errors over Test pattern
time analyzed
using both images showing
EPID and

conclusion stated, “The cumulative effect of many small

file data. Errors log 1.3mm leaf
WHAT LOG
FILES SEE seen in the
EPID data displacem
are not seen ent
the log file in
analysis
ABSTRACT
This paper
SUMM ARY
ABSTRACT presented
treatment a clinical

should be aware of the

This study
SUMM ARY deviating
imaging
identif
case in which
from its progra ied an MLC leaf real-time
investigated An EPID-b to intra-
positioning ased exit-flu mmed and logged be consistently

errors can, in worst case scenarios, produce large ones.

accuracy the differe
electronic determined nces between MLC capture cine ence position by
portal imagin using either identified images during dosimetry system >1mm.
possibility g devices log files treatment. was
of
control via reducing time spent (EPID) and assess or detected
a suspec
ted MLC The author used to
ed the by leaf displac visually
was develo
phantom-less on patien
t-specific EPID image other means. The ement that
ped and validatmethodologies. quality s leaf
for the treatmwas measured and position as record not
was
with rotatio In-house
nal and static ed to track MLC software ent in questi the log ed on the

limitations of using only log

electronic
portal image picket fence tests
positional
accuracy
daily MLC
patterns on, the prior files were analyz
daily MLC acquired day’s ed
performance . This software was using an integrated Whereas on the treatm treatment and

This amalgam should be considered as part of the QA

TrueBeam used to monito the ent days. for
systems, over a one-ye
ar period r and record log file reported no
positions with results ed positio difference
determined directly compaon two Varian the leaf to ns,
be 1.3 ±0.1m image-based measu
between
planned
MLC positio using leaf red to MLC
ning errors trajectory The offset
lower than log files. was confirm m medial from the rements showed
the related identified by the log Average ed using planned positio
the EPID-b MLC positio file test pattern
ased
multiple MLC software. Over ning errors analys is were CONCLUSIO irradiations. n.
obtained N

files and consider this in

the duratio
software positional
errors were n of the study,using The author
s concluded,
but these detected log-file derived “It has been
trajectory same errors using the

process.”
log files. were not EPID position by leaf positio clinically
detected >1mm, and ns can differ observed
using the the actual from their that
CONC LUSION leaf positio therefore cannot actual
MLC positio ns.” Furthe be consid
The author ns r, “Frequent ered to
s noted, “In precondition through indepe verification be
trajectory
logs this study, imaging provid to trust log-file ndent means of
did not detect created during it was found records. Intra-tris a necessary
the deliver that the planned positioes a method to
capture departeatment EPID

terms of clinical use and

an EPID.” leaf positio y of a picket
They also nal errors fence test ns.”
should not noted, “…this that were ures from
detected MLC
be
detect system solely relied upon study shows that using
for QC as [log
atic machi
ne faults. they do not files]
” always

Q: Is a “similarity index” comparison between interpretation of results.

CBCT and CT a suitable method for detecting For more information on the
patient set-up errors or anatomy changes? limitations of log file QA, see Sun Nuclear’s Patient QA:
What Log Files Miss document
A: A similarity index can be used to make the user aware
that there is potential for an issue due to differences in the Q: Are log files measurements?
patient set-up and/or anatomy. However, the dosimetric
A: Not exactly. Log files contain data reported by
impact of these issues is not known unless the dose
motor encoders and onboard systems. The MU/output
computation is performed using the CBCT. Relying on a
data in a log file is a measurement from the monitor
CBCT image to detect patient issues will also not detect
chamber. However the MLC leaf position information
issues related to patient movement during treatment.
in a log file is not a physical measurement; it is merely
Additionally, in cases where a CBCT is not performed, there
the feedback from an electrostatic motor.
is no ability to detect potential patient issues unless an
EPID is used. The log file-based approach relies on machine
information as the foundation for “self-reporting” of
errors. But, the accuracy of machine information for
the detection of errors remains firmly in question. 7,8,17

21 I| www.lightages.com
SUN NUCLEAR CORPORATION // sunnuclear.com
 Patient-Specific Quality Assurance

EPID-Based PerFRACTION
Q: Can log files be considered independent? Q: Does PerFRACTION include both 2D and
A: No. Log files cannot be considered independent. 3D functionality?
They accumulate machine interlock encoder data A: Yes, when using EPID data.
defined and recorded by the linac, and they do not
measure the radiation fluence distribution. Q: What is the difference between 2D and 3D
Log files provide great precision (different from analysis functionality in PerFRACTION?
accuracy) and support automation 4,5. However, log A: Both functions are included with PerFRACTION.
files remain a reflection of what the machine “thinks” PerFRACTION 2D (only available when using EPID data)
happened during treatment, and have been shown provides the following:
to be unreliable as the sole source for patient QA,
1. A
 utomatic Fraction 0 pre-treatment IMRT QA using
missing many common errors that can be detected
absolute dose (Included in optional Dosimetry
by measurement using EPIDs. In addition, log files
Package)
cannot detect patient related issues. 4,6,7,17
It is important to understand that log file data is 2. A
 utomatic capture and 2D comparison of daily
a direct extension of the machine interlocks, and treatment EPID images.
therefore provides no ability to identify critical errors
3. A
 utomatic detection of failures related to patient setup,
above and beyond the early warnings and shut-
patient movement, and anatomical issues including
offs already provided by the interlock system of the
weight loss and tumor growth/shrinkage, via transit
treatment unit.
image analysis.
Relying solely on data obtained from the linac control
system or treatment planning system cannot be 4.Automatic email of failed results.
considered independent QA.17 PerFRACTION 3D adds the following:
Q: Can log files detect MLC or dose errors? 1. A
 utomatic reconstruction of 3D dose on the patient CT
A: It has been demonstrated that log files can detect (or on daily CBCT with optional Dosimetry Package).
severe MLC errors, but routinely miss critical MLC
2. A
 utomatic dose/volume analysis, including Clinical
positioning errors and drift due to a variety of issues,
Goals and 3D gamma results for the total volume and
including T-nut or motor failure, encoder error, or mis-
structure by structure.
calibration. 6,7,17 A 2014 linac manufacturer field safety
notice underscored this point, announcing dose rate 3. A
 utomatic Point Dose analysis for composite dose to
errors up to 5% had been missed by the recorded Points of Interest and per-beam calculation points.
monitor chamber units. This error was detected by a
4. A
 utomatic email notification summarizing pass/fail
measurement device.8
verification of clinical objectives.
Q: But, aren’t log files easy to analyze?
Q: What planning data is needed by
A: Log files are easy to access and work with.
Recognizing this appeal, and in response to customer
PerFRACTION?
feedback, PerFRACTION can be configured as a A: For 2D Planar Analysis: DICOM RT Plan
calculation (i.e.log file) only-based option. However, For 3D Dose Reconstruction: DICOM RT Plan, RT Dose, RT
ease of use for log file-based analysis should be Structure Set, and CT Image (Planning CT or CBCT).
understood in the context of the inherent limitations
and risks of relying on it alone for pre-treatment and
in-vivo patient QA. It is important to know EPIDs are
easy to work with as well, and include the benefit of
providing independent information. This is why an
increasing number of QA products have focused on
using the EPID to its full potential.

SUN NUCLEAR CORPORATION // sunnuclear.com | 3

 Frequently Asked Questions

EPID-Based PerFRACTION Continued

Q: Does PerFRACTION calculate absolute or Q: A recent publication found that a high
relative dose? percentage of errors can be detected during
A: Both. 2D Absolute Dose for Fraction 0 is a powerful the first fraction. How can PerFRACTION
option for pre- treatment IMRT/ VMAT QA. 3D dose detect these errors if the comparison
reconstruction is always absolute. 2D transit measurements baseline is the first fraction?
for Fraction n are (currently) relative.
A: An absolute dose functionality for Fraction n, which
Q: Is PerFRACTION a true in vivo dosimetry generates a predicted planar dose based on the patient’s
plan and CT, is currently under development.
solution?
A: Yes. PerFRACTION provides a true in-vivo dosimetry Q: Does PerFRACTION recalculate onto
solution by harnessing the power of simultaneous 2D transit CBCT?
measurement and 3D absolute dose reconstruction and
A: Yes, PerFRACTION has the option to retrieve and
analysis. When the additional power of CBCT-based dose
recalculate dose onto the CBCT instead of the treatment
reconstruction is added, all potential sources of error are
planning CT. This is an automated process and easy to set
accounted for.
up in the system preferences. The system will default to the
This is consistent with the accepted definition in the most recent acquired CBCT for dose reconstruction and
literature as presented in a comprehensive review by van analysis when this feature is active.
Elmpt, et al, 2008:18
Q: Is the use of PerFRACTION a good way to
“in-vivo dosimetry: measurement or determination of the spend my limited QA time?
dose inside the patient. Measurements performed during
A: Yes. A 2015 study from University of Washington
treatment can be performed invasively, i.e. inside the patient,
reviewed 30 months of failure mode data. From the
or non-invasively, i.e. on or at some distance from the
analysis, 343 incidents were rated as “potentially severe” or
patient, whereby the in vivo dose at the point of interest is
“critical.”1 Of these incidents:
obtained by extrapolation.”
• 6% were detected by EPID-based pre-treatment QA
An in-vivo solution that utilizes back-projected EPID data is
simply using different inputs to provide the representation • 74% were detected by EPID-based in-vivo QA for the
of the delivered patient dose. It is noteworthy that such first fraction
solutions actually make it harder to discern sources of
• 20% were detected through EPID-based in-vivo QA
potential error since all sources — linac output, MLC leaves,
following the first fraction
jaws, and patient — are combined in the EPID signal.
Interested readers are referred to the SNC white paper On Clearly, routine measurement-based patient QA using the
the Matter of Forward Versus Back Projection for further EPID can improve your ability to discover and reduce the
details. impact of the 94% of potentially severe common treatment
errors not detected by pre-treatment QA alone, thus
Further considerations for the use of EPIDs for in-vivo
improving patient safety.
dosimetry is presented in the AAPM vision 20/20 paper.
Furthermore, PerFRACTION’s automated processing
provides these benefits with minimal footprint on your daily
workflow (see below).

SOURCE R&V or TPS

RT Plan

For a detailed overview of process

PerFRACTIONª
flows for PerFRACTION 2D and 3D
modes, please see the inside back
DATABASE

cover of this document.

41 I| www.lightages.com
SUN NUCLEAR CORPORATION // sunnuclear.com
 Patient-Specific Quality Assurance

Workflow Efficiency Practical Concerns

Q: Do I have to use PerFRACTION for every Q: What about fields that are larger than the
fraction? EPID, fields that require an EPID shift, or a
A: It’s up to you. PerFRACTION has been designed to field (i.e. vertex) where the EPID cannot be
handle automated pre-treatment phantom-less IMRT QA deployed?
and automated in-vivo QA for every fraction, but actual A: PerFRACTION’s dose reconstruction algorithms are
frequency can be selected to suit departmental needs. capable of handling fields where part of the field is
As described earlier in this document, a 2015 study from outside the EPID sensitive area. Panel shifts and sag are
the University of Washington found 74% of errors rated as managed using the Auto-Align functionality within 2D
“potentially severe” or “critical” were detected when a first- analysis. For Fraction 0, the EPID can be deployed at a
fraction in-vivo QA was added to the pre-treatment QA,1 closer SDD to better fit the field sizes.
and a further 20% when used for later fractions.
In cases where the EPID is not used, is impractical, or
Q: 2D image to image comparison for images are missing, PerFRACTION can use the log files
to reconstruct dose.
pre-treatment QA can be done with my
Varian linac with Portal Dosimetry. How is Q: What linac/R&V configurations are
PerFRACTION different? currently supported by PerFRACTION?
A: In addition to full independence, PerFRACTION offers A: The following configurations are supported by
several important advantages in efficiency and clinical PerFRACTION:
value: • Varian configurations:
Automated. Portal Dosimetry requires manual initiation - TrueBeam™ / ARIA®, TrueBeam / MOSAIQ®,
of the calculation/comparison. With PerFRACTION, the C-Series / ARIA, C-Series / MOSAIQ®
analysis and alerts happen automatically. • Elekta configurations*:
Flexible. Point dose, 2D, and 3D analysis tasks can be - Elekta digital accelerators (includes Versa) with
simultaneously performed using different methods and iViewGT 3.4 or 3.4.1 / MOSAIQ
settings. This streamlined, automated workflow makes it *Mosaiq 2.5+ and XVi 5.0+ are required for CBCT
more likely that analysis can and will be performed daily functionality.
for all patients.

Comprehensive. PerFRACTION is part of an integrated

patient QA workflow and user interface that provides
quality and clinical goals tracking from secondary
calculations (DoseCHECK™) to pre-treatment QA
(PerFRACTION Fraction 0) through in-vivo monitoring
(PerFRACTION Fraction n) and exists within the larger
integrated environment of the SunCHECK Quality
Assurance platform

SUN NUCLEAR CORPORATION // sunnuclear.com | 5

 Frequently Asked Questions

Practical Concerns Continued

Q: With phantom-less pre-treatment QA, is 2. Commissioning – Arrays are essential tools for
there any need for arrays like MapCHECK® 2 efficient and stringent commissioning of new
accelerators and treatment modalities. Arrays
and ArcCHECK®?
can collect data on a wide range of parameters
A: Patient QA arrays remain the gold standard for fully
that EPIDs and machine log files cannot. Medical
independent AND rigorous QA. Arrays offer a variety of
Physics Practice Guideline 5, created by AAPM
benefits in combination with phantom-less QA:
Task Group 244, recommends the use of arrays
1. Audit QA – Arrays should be used for periodic audit with volumetric capabilities (such as MapCHECK or
QA (every nth patient) to ensure issues are not ArcCHECK® with 3DVH®) as a step in a rigorous test
present which the EPID or log file may miss (gantry of TPS commissioning16.
rotation, sag, etc). Periodic array measurements test 3. Troubleshooting – Arrays are uniquely valuable
the system in more than one manner. Array audit for troubleshooting unusual machine behavior and
frequency can be a function of the complexity or inconclusive results. Their physical independence
clinical familiarity with the QA case type. There are from other systems offers flexibility and stringency
many examples of audit QA in use at clinics. For to test the system end to end. Arrays are credited
instance, a daily device (e.g. Daily QA™ 3) is used for with discovering systematic and systemic errors
morning output checks, but a different device (e.g. that phantom-less methods will not see.
PROFILER™ 2, MapCHECK™2) is used for monthly 4. Post-Service / Upgrade Verification – Arrays
output checks, and yet another device (e.g. 1D enable completely independent verification (and
SCANNER™, PC ELECTROMETER™, IC PROFILER™) base comparison) following any machine or TPS
is used for annual output checks. These devices service or upgrade.
check the same QA parameters in different ways 5. Backup – Arrays offer an invaluable backup for
and serve as an overall audit of other QA devices. occasions where a phantom-less method may be
off- line or unavailable.

61 I| www.lightages.com
SUN NUCLEAR CORPORATION // sunnuclear.com
 Patient-Specific Quality Assurance

IT & Setup Considerations

Q: Is SunCHECK (including PerFRACTION, Q: How does PerFRACTION communicate
DoseCHECK, SNC Machine, and SNC with ARIA?
Routine) a Cloud application? A: PerFRACTION communicates with ARIA through the
A: SunCHECK applications are ‘cloud enabled’ because system’s DICOM Query/Retrieve service to automatically
they are accessed from anywhere on your clinical find and process new PerFRACTION data.
network via the web browser. By running locally, these
Q: How does PerFRACTION communicate
applications provide more automation and faster
with MOSAIQ?
processing performance than a purely Cloud application.
A remote Cloud data storage service will likely be A: PerFRACTION monitors the MOSAIQ import folder and
available in the future. automatically retrieves relevant data once it is found.
CBCT data is retrieved using a SQL database query.
Q: How do I access PerFRACTION in my clinic?
A: Simply direct a supported web browser to the network Q: What is the process for installation and
location where PerFRACTION is installed. training? What is required from Elekta and
Varian?
Q: I am a consulting medical physicist; how
will I use these applications at my client A: Upon receipt of your order, Sun Nuclear Installation
sites? and Support will contact you to schedule a date for
installation and training. Sun Nuclear will collect
A: For optimal performance, SunCHECK applications
information about your facility and guide you through
such as PerFRACTION, DoseCHECK, and SNC Machine
the process of understanding what will be required from
are designed to run on a local network.
Elekta and/or Varian for your specific configuration.
This is also required for proper function of the Sun Nuclear will work with you to ensure that all the
automation architecture. Consulting physicists may pre-requisites are configured prior to coordinating
access these applications by connecting to the network PerFRACTION specific installation and training.
hosting them.
Q: What kind of involvement from our
Q: Can I install PerFRACTION on a server center’s IT department does PerFRACTION
that I provide? require?
A: Yes. Documentation is available upon request to A: During the pre-installation process, Sun Nuclear will
help ensure user provided servers meets the required work closely with your IT department to ensure that the
specifications. server(s), linac(s), TPS and R&V are configured correctly
prior to PerFRACTION installation and training.
Q: How does PerFRACTION’s automation
function?
A: PerFRACTION utilizes the information contained in
DICOM headers to enable automated data retrieval and
processing. When a DICOM RT Plan is first received
by the system (usually via DICOM export from the
Treatment Planning System, or TPS) patient information
such as name and medical record number, as well as
plan details such as plan name and number of fractions,
are read from the DICOM header. The patient record is
automatically created within the system and it begins
monitoring for any further incoming information related
to that patient and plan.

SUN NUCLEAR CORPORATION // sunnuclear.com | 7

 Frequently Asked Questions

Technical Considerations
Q: What method does PerFRACTION use to Q: How does PerFRACTION handle beam
analyze 2D results? modeling?
A: Analysis methods for available for 2D (relative and
A: PerFRACTION uses a standard library of beam models
absolute dose) include gamma, percent difference,
covering most commercial linear accelerator energy/
composite evaluation (DTA), gradient compensation, and
MLC configurations. The PerFRACTION beam model
Diff-to-DTA (Sun Nuclear exclusive based on ICRU 83
library uses beam data that is more specific and accurate
Appendix A).
than universal/golden beam data provided by linac
Q: What algorithm does PerFRACTION use to manufacturers.
calculate & evaluate 3D results? Q: Can the beam model be customized for
A: GPU-accelerated collapsed cone convolution/ my machine?
superposition exclusively licensed from Johns Hopkins
A: Sun Nuclear can provide a custom beam model in
University. 3D dose is evaluated using Point Doses, 3D
situations where this is determined necessary.
Gamma, Clinical Goals, Dose Volume Histogram (DVH),
and isodose images. Q: Is EPID drift a concern?
A: PerFRACTION 3D dose reconstruction is immune
Q: How is the EPID calibrated for absolute
to EPID drift or changes in the EPID because the
dose?
proprietary leaf-edge detection algorithm does not rely
A: PerFRACTION generates a calibration RT Plan that is on absolute values from the EPID image. When using the
specific to the linac, MLC, EPID panel, energy, and SID. This EPID for absolute dose analysis in 2D-mode, the EPID
plan can be exported directly to the Record & Verify system requires calibration, the process for which is included in
for delivery. Once images are collected, PerFRACTION PerFRACTION.
automatically retrieves them and compiles the calibration.
Q: By using PerFRACTION frequently, will the
Q: What accuracy studies exist on lifetime of my EPID be reduced significantly?
PerFRACTION? A: Not likely. The shift toward EPID dosimetry over the last
A: There have been several accuracy studies performed on decade has fueled innovation in the design of EPIDs, so
both PerFRACTION 2D10, 12, 13 and 3D results14, 15. These with newer EPIDs radiation lifetime has been improved.
papers have found that PerFRACTION 2D is “sensitive
enough One OEM notified customers that the dose tolerance of
their EPID is 5M cGy (or 50 kGy) in one year. In order to
to detect small positional, angular, and dosimetric errors exceed that number, 195 patients per day would have to
within 0.5mm, 0.2 degrees, and 0.2% respectively,”13 and be treated with the EPID extended for every field for the
that PerFRACTION’s dose calculations are accurate to entire year on one linac. PerFRACTION also allows the use
within 1% of other treatment planning systems.14 of log files and/or intermittent EPID measurements for
Q: How does PerFRACTION handle electron transit/in-vivo dose monitoring as a way to manage EPID
life expectancy.
density corrections?
A: PerFRACTION provides the ability to enter CT-to-
electron (CT-to-ED) density values for CT scanners used
for treatment planning. These CT-to-ED values are

automatically applied using information from the DICOM

header in the CT image during 3D dose calculation. Default
tables based on published literature are also available.

81 I| www.lightages.com
SUN NUCLEAR CORPORATION // sunnuclear.com
 Patient-Specific Quality Assurance

References
1. C. Bojechko, et al., “A quantification of the effectiveness of EPID dosimetry and software-based plan
verification systems in detecting incidents in radiotherapy,” Med Phys. 42, 5363 (2015)
2. A. Templeton, et al., SU-E-T-273, “Do Task Group External Beam QA Recommendations Guarantee Accurate
Treatment Plan Dose Delivery?,” Med. Phys. 42, 3395 (2015)
3. Automated MOSAIQ processing requires manual export of DICOM files from MOSAIQ to the SNC Machine
DICOM listener.
4. D. Rangaraj, et al., “Catching errors with patient-specific pretreatment machine log file analysis,” Practical
Rad. Onc. 3(2), 80-90 (2013)
5. A. Stell, et al., “An extensive log-file analysis of step-and-shoot intensity modulated radiation therapy segment
delivery errors,” Med. Phys., 31(6), 1593-1602, (2004)
6. J. Monroe and C. Bull, “Study of Dosimetric Leaf Gap and Transmission Factor Variations Affecting Common
Clinical QA Tools,” Med. Phys. 42, 3500 (2015)
7. A. Agnew, et al., “Monitoring daily MLC positional errors using trajectory log files and EPID measurements for
IMRT and VMAT deliveries,” Phys. Med. Biol., 59, N49-63 (2014)
8. V. Tran, “Unexpected 6MV Beam Output Variations,” Urgent Field Safety Notice, CP-12459, Varian Medical
Systems, June 17, 2014, pp 1-3
9. B. Mijnheer, et al., “Current status of 3D EPID-based in vivo dosimetry in The Netherlands Cancer Institute,”
Journal of Phys.: Conf. Series, 573 (2015)
10. S. Dieterich, et al., SU-E-T-133, “Assessing IMRT Treatment Delivery Accuracy and Consistency On a Varian
TrueBeam Using the Sun Nuclear PerFRACTION EPID Dosimetry Software,” Med. Phys. 42, 3362 (2015)
11. A. Mans, et al., “Catching errors with in vivo EPID dosimetry,” Med. Phys., 37, 2638 (2010)
12. SU-E-T-139: Automated Daily EPID Exit Dose Analysis Uncovers Treatment Variations, A Olch, Med. Phys. 42,
3363 (2015)
13. SU-C-BRD-06: Sensitivity Study of An Automated System to Acquire and Analyze EPID Exit Dose Images, A
Olch, Med. Phys. 42, 3193 (2015)
14. Real-time dose computation: GPU-accelerated source modeling and superposition/convolution. Jacques, et
al. Med Phys. 2011 Jan;38(1):294-305.
15. Towards real-time radiotherapy: GPU-accelerated superposition/convolution. Jacques, et al. Comput
Methods Programs Biomed. 2010 Jun;98(3):285-92.
16. J. Smilowitz, et al., “AAPM Medical Physics Practice Guideline 5.a.: Commissioning and QA of Treatment
Planning Dose Calculations – Megavoltage Photon and Electron Beams, ” J. App. Clin. Med. Phys., 16, 5768
(2015)
17. B. Neal, et al., “A clinically observed discrepancy between image-based and log-based MLC positions,” Med.
Phys. 43, 2933 (2016)
18. W. Elmpt, et al., “A literature review of electronic portal imaging
for radiotherapy dosimetry,” Radiotherapy and Oncology 88 (2008) 289–309

SUN NUCLEAR CORPORATION // sunnuclear.com | 9

 Frequently Asked Questions

Process Flows: PerFRACTION™ Patient-Specific Quality Assurance

Process Flows:™PerFRACTION™ Fraction n™

Fraction n - In-Vivo Monitoring

1 I | www.lightages.com
10 SUN NUCLEAR CORPORATION // sunnuclear.com
 Patient-Specific Quality Assurance

Fraction n 3D mode includes the option of using DoseCHECK calculated dose as a reference
dose (dashed line). DoseCHECK as a pre-treatment secondary check sold separately.

sunnuclear.com // +1 321 259 6862

Corporate Headquarters: 3275 Suntree Boulevard, Melbourne, FL 32940 USA
erFRACTION_2PG_072016 All data used is best available at time of publication. Data is subject to change without notice. ©2016 Sun Nuclear Corporation. All Rights Reserved.

SUN NUCLEAR CORPORATION // sunnuclear.com | 11

 Corporate Headquarters
3275 Suntree Boulevard
Melbourne, FL 32940 USA

+1 321 259 6862

sunnuclear.com

©2016 Sun Nuclear Corporation. All Rights Reserved. All data used is best available at time of publication. Data is subject to change without notice. PF_ Patient-Specific Quality Assurance _041317