helical tomotherapy quality...
TRANSCRIPT
Helical Tomotherapy Quality Assurance
Cheng Shi PhDChengyu Shi, PhDDepartment of Radiation Oncology
Di i i f M di l Ph iDivision of Medical Physics
Southwest Chapter of the AAPM, Oct 9, 2009
ObjectivesObjectives• To introduce our five-year clinical experience
ith helical TomoTherapwith helical TomoTherapy• To introduce our five-year research experience
with helical TomoTherapy
There is no conflict of interestThere is no conflict of interest
IGRT SolutionIGRT SolutionImaging
Ionizing Non-ionizing
CT Ultrasound MRIX-ray
PETSPECT RF Laser
OthersOthers
IGRT QAIGRT QA• IGRT solution introduces precise targeting
• However, it may also introduce more uncertainties and increased risk to miss the targetg
• Therefore, comprehensive QA procedures are necessary
• Helical Tomotherapy has its own characteristics• Helical Tomotherapy has its own characteristics, and requires dedicated QA protocol
Helical TomoTherapy QAHelical TomoTherapy QA
QA
System PatientSystem Patient
Hardware Software Planning Delivery
Helical Tomotherapy ProcessHelical Tomotherapy ProcessR i d i
MVCTRegistered image
MVCT Guided
Image registration
Helical CT Delivery
Helical Tomotherapy QA DesignHelical Tomotherapy QA Design
Component-based: QA each component, h MLC d t t h tsuch as MLC, detector, couch etc.
System based: QA the whole systemSystem-based: QA the whole system performance
I am NOT good I am good
Helical Tomotherapy QAHelical Tomotherapy QA
Chester Ramsey, et al. "Mechanical and Dosimetric Quality Assurance for Helical Tomotherapy," International Journal of Radiation Oncology Biology Physics, 1 November 2007 (Vol. 69, Issue 3, Page S699)
Helical Tomotherapy QAHelical Tomotherapy QA
Chester Ramsey, et al. "Mechanical and Dosimetric Quality Assurance for Helical Tomotherapy," International Journal of Radiation Oncology Biology Physics, 1 November 2007 (Vol. 69, Issue 3, Page S699)
Helical Tomotherapy Target IssueHelical Tomotherapy Target Issue
Eroded target Rusted on-board ion chamber
Special MLC QA for Helical Tomotherapyp py
S h b d i dSeven test patterns have been designed
A. Box in BoxB. Modified Checkerboard. od ed C ec e boa dC. MLC Interleaf LeakageD IEC X gradientD. IEC-X gradientE. IEC-Y gradientF C l Fi ld A (2 t t )F. Complex Field A (2 tests)G. Complex Field B (2 tests)
Pattern B. Modified checkerboxPattern B. Modified checkerbox• Can be thought as small box in box testg• Initially used for Mimic MLC test• Modified for tomotherapy MLC testpy
Pattern D. IEC-X GradientPattern D. IEC X GradientPNMW depthnormwedge PNMW ••=
PNMW ••= depthnormwedge PNMW ••=
Tomotherapy Patient-Specific QATomotherapy Patient Specific QA
QA Equipment
Generate QA procedure
Deliver QA procedure
Analysis QA procedure
Evaluate QA procedure
What if the QA fails?Q
Find out why?
Phantom resolution? High dose gradient? Small thin PTV? Machine output? …
Repeat the QA procedure
Example-Overweightp g
•Patient with prostate cancer (weight 370 lb)•Tomotherapy couch can support ~440 lb.•Patient plan is homogeneous for PTVp g•Phantom size is very different from patient size resulting to non-homogenous dose at thesize, resulting to non-homogenous dose at the point of the chamber
Delta 4 Dosimeter for Tomo
• Purchased September 2008TomoTherapy OptionRapidArc Option
Fi t f th• First few monthsResearchSBRT QASBRT QA
• Mid‐January 2009Fully clinical for tomotherapy DQA
• Absolute calibrate on T ThTomoTherapy
• Phantom MVCT image for planningplanning
Delta 4 Dosimeter ResultsDelta 4 Dosimeter Results• 140 DQAs delivered Q
on Delta4
• 34 SBRT patients with gamma index > 95% (3%/3mm)
• Pass criteria3%/3mm
90%>90% gamma
• To date,Mean gamma: 97 1%Mean gamma: 97.1% (85.9‐100)
DiscussionDiscussion
T th i li bl l l ti lt d it h ld• Tomotherapy gives reliable calculation results, and it should be within 3% tolerance about dose measurement for most of casescases
• EDR2 is a good tool for relative and absolute dose measurementmeasurement
• Gamma factor is a good evaluation parameter
• Some QA cases need more analysis
• QA is a little hard to do in some cases such as spine andQA is a little hard to do in some cases such as spine and lung SBRT
Tips or Problems you may meetTips or Problems you may meet
• Scanning QA phantom with different resolution and extension
P t i h b i t h it i f th PTV• Put ion chamber into homogeneity region of the PTV
• Use high resolution phantom for smaller or thinner PTV
• You can extend EDR2 film range up to 600~700 cGy, however, that is for relative dose measurement only. For absolute dose measurement using film, you can do it under 400 cGy, and it should be good for normal treatment procedure.
• Delta 4 dosimeter is a good candidate for Tomo QA
Design algorithmDesign algorithm
Hardware related• Target changeTarget change• Jaw• MLC• MLC…Software related• Composite plan• DVH…Clinical/Research need• Archive patientArchive patient…
Design languageDesign language
Excel VBA for clinicalMatlab for researchMatlab for researchVisual c++ for fast applicationXML, html
Tools we have developedTools we have developed
ToPinnacleTomoBinTomoBinTomoAnnualTomoGkPinnacleTomo-XML-DOMTomo-XML-DOMExcel sheets
ToPinnacleToPinnacle
Need: Physicians need plan comparison between Tomotherapy and Pinnacle andbetween Tomotherapy and Pinnacle and composite plan
Design: Visual c++ program to transfer dose matrix data from Tomotherapy tomatrix data from Tomotherapy to Pinnacle
TomoBin
Need: Tomo procedure needs sinogram d bi MLC d QAand binary MLC needs QA
Design: Matlab program to design tomotherapy MLC patterntomotherapy MLC pattern
How were the patterns designed?How were the patterns designed?
Consider dose and mechanical factorsSeven patterns (9 tests) have been designeddesignedA software platform is also designed to
t d l th ttgenerate and analyze the patternsThe patterns can be transferred easilyThe patterns can be transferred easily to other centers with a helical tomotherapy unittomotherapy unit
TomoAnnualTomoAnnual
Need: How to compare commission data with annual QA data? How to predict ptarget failure?
Design: Matlab program to use commissionDesign: Matlab program to use commission data as baseline for monthly and annual
fprofile comparisonC Shi S Stathakis N Papanikolaou “Development and application of aC Shi, S Stathakis, N Papanikolaou, Development and application of a platform for analyzing helical tomotherapy annual and monthly profiles,” the XVth International Conference on the Use of Computers in Radiation Therapy Toronto Canada June 4 7 abstract # 6102 (2007)Therapy, Toronto, Canada, June 4-7, abstract # 6102 (2007).
TomoAnnualTomoAnnual
Profiles collected for theprevious target a-d and after thetarget replacement etarget replacement e.
The angle difference between two profiles for d) is around 1 degree.
TomoGkPinnacleTomoGkPinnacle
Need: for research use; how to transfer data from one planning station to another?p g
D i M tl b t t dDesign: Matlab program to convert dose matrix between different TPS
C Shi, J Penagaricano, Y Yan, N Papanikolaou, “Application of the Post-Processing Dose Tool (PPD) to Dosimetrically Compare Gamma Knife andProcessing Dose Tool (PPD) to Dosimetrically Compare Gamma Knife and Hi Art Tomotherapy,” SU-FF-T-101, AAPM 48th Annual Meeting (2006)
TomoGkPinnacle->APT4DTomoGkPinnacle >APT4D
Lan Lin, Chengyu Shi, Gregory Swanson, Nikos Papanikolaou, "Development of a Novel Post-processing Treatment Planning Platform for 4D Radiotherapy," Technology in Cancer Research and Treatment (April 2008)Technology in Cancer Research and Treatment (April 2008)
Tomo-XML-DOMTomo XML DOM
Need: how to extract necessary information from tomotherapy plan to perform a 2nd check?p
Design: XML language to extract archived patient informationarchived patient information
T Roland CY Shi N Papanikolaou "Development and application of aT Roland, CY Shi, N Papanikolaou, Development and application of a robust toolkit, Tomo-XML-DOM, for helical tomotherapy system," Submitted to 50th AAPM Annual Meeting (2008)
Excel sheetsExcel sheets
Need: daily, monthly, annual QA, patient setup, QA…p,
D i l VBADesign: excel VBA
Where do we go?Where do we go?
• System-based QA vs. component-based QA
• Internet QA with intra/inter institution comparison
• Risk estimation for failure rate and prediction
S lf QA hi d QA• System self QA vs. third-part QA
SummarySummary
• Current Helical Tomotherapy QA Status is component-based QA and varies from center to p Qcenter
• Future implementation of Helical Tomotherapy QA maybe system-based, internet-based, risk-Q y y , ,estimation, and system-self QA
• AAPM TG 148 will provide recommendations for QA tests and frequency of testingQ q y g
AcknowledgementsAcknowledgements
• Dr. Papanikolaou• Dr GutierrezDr. Gutierrez• Dr. Lan Lin• Dr. Fan-Chi (Frances) Su• Dr Vikren SarkarDr. Vikren Sarkar• Roland Teboh• Weihong He• Luis Vazquez Quino• Luis Vazquez Quino
E il hi @ h dEmail: [email protected]