02/04/2013
AdvancesinBrachytherapyDose
Calculations,PartII
Firas Mourtada, Ph.D.
Chief of Clinical Physics
Helen F. Graham Cancer Center
Christiana Care Health System
Newark, Delaware
Adjunct Associate Professor
MD Anderson Cancer Center
Houston, Texas
Disclosures
• Iam amember oftheAAPM/ESTRO/ABGWorking
GrouponModelͲbased DoseCalculation
Algorithms.OurWGis working with all
brachytherapy TPSvendors.
1
02/04/2013
LearningObjectives
• 1.Identify keyclinical applicationsneeding advanced
dosecalculation inbrachytherapy.
• 2.Provide anoverview ofthealternativestoTG43
• 3.Explains inpractical terms therecommendations of
TG186
• 4.Identify relevantcommissioning processes forsafe
clinical integration.
• 5.Provide practical clinical examples using a
commercially available system
2
02/04/2013
SectionVofTGͲ186Motivations
• Nopracticalguidanceonthecommissioningfor
clinicaluseforMBDCAforBrachytherapy
• Thereisaneedtoestablishuniformcommissioning
proceduresforvariousMBDCAvendorsandacross
institutions
• Nottoforgetthecurrentguidelines…
3
02/04/2013
BrachytherapyTPSCommissionProcess
• AAPM TG-40: Comprehensive QA for Radiation Oncology
• AAPM TG-43: Dosimetry of Interstitial Brachytherapy
sources
• AAPM TG-53: Quality Assurance for Clinical
Radiotherapy Treatment Planning
• AAPM TG-59: High Dose-rate Brachytherapy Treatment
Delivery
• AAPM TG-56: Code of Practice of Brachytherapy Physics
• AAPM TG-64: Permanent Prostate Seed Implant
Brachytherapy
Two General Categories
Commission tasks broken down into two general
categories:
– Non-dosimeteric tests
– Dosimetric tests
4
02/04/2013
“Non-dosimetric” Tests
•
•
•
•
•
•
Image Input tests (CT, MR, PET, film)
3D anatomical structure input test
Contour tests
Image use and display test
Accurate transfer to afterloader TCU
Computer storage and database backup
“Dosimetric Tests”
•
•
•
•
•
•
•
Dose calculation methodology/algorithm tests
Dose display tests
DVH tests
Source strength decay
Source localization tools
Impact of grid size
Etc.
5
02/04/2013
Ex1: Oncentra
Single mHDR
source Test (TG43)
Name
P1
P2
P3
P4
P5
P6
P7
P8
P19
P9
P10
P11
P12
P13
P14
P15
P16
P17
P18
Norm.
no
yes
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
X(mm)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Y(mm)
0
0
0
0
0
0
0
0
0
5
10
15
5
10
Ͳ5
Ͳ10
Ͳ15
Ͳ5
Ͳ10
Z(mm)
5
10
15
20
25
30
35
40
100
0
0
0
5
10
0
0
0
5
10
Abs.Dose(cGy)
(OCBͲPlato)/
OCB
Plato
(Plato)
1940.31
1940.31
0.00%
500.00
500.00
0.00%
224.20
224.20
0.00%
126.88
126.88
0.00%
81.36
81.36
0.00%
56.53
56.53
0.00%
41.49
41.49
0.00%
31.69
31.69
0.00%
4.46
4.46
0.00%
1548.19
1548.19
0.00%
329.52
329.52
0.00%
144.89
144.63
0.18%
987.14
986.69
0.05%
244.38
244.11
0.11%
1413.56
1413.56
0.00%
305.50
305.50
0.00%
135.24
134.40
0.63%
988.55
987.97
0.06%
244.95
244.73
0.09%
Ex2: Oncentra two
mHDR sources Test
(TG43)
Name
P1
P2
P3
P4
P5
P6
P7
P8
P19
P9
P10
P11
P12
P13
P14
P15
P16
P17
P18
Norm.
no
yes
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
no
X(mm)
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
0
Y(mm)
0
0
0
0
0
0
0
0
0
5
10
15
5
10
Ͳ5
Ͳ10
Ͳ15
Ͳ5
Ͳ10
Z(mm)
5
10
15
20
25
30
35
40
100
0
0
0
5
10
0
0
0
5
10
Abs.Dose(cGy)
(OCBͲPlato)/
OCB
Plato
(Plato)
1726.42 1726.42
0.00%
500.00
500.00
0.00%
243.72
243.72
0.00%
146.29
146.29
0.00%
97.99
97.99
0.00%
70.37
70.37
0.00%
52.99
52.99
0.00%
41.31
41.31
0.00%
6.37
6.37
0.00%
1553.28 1552.87
0.03%
457.74
457.53
0.05%
231.14
230.93
0.09%
948.16
947.62
0.06%
275.26
274.88
0.14%
1445.94 1445.88
0.00%
439.04
438.86
0.04%
223.51
222.76
0.34%
949.45
948.77
0.07%
275.78
275.53
0.09%
6
02/04/2013
Dose algorithm tests beyond TG-43
• MC, CC or GBBS implementation as MBDCA
involve compromises between computational
speed and sufficient accuracy
• Must be carefully benchmarked against analog
MC or experiment to ensure sufficiently accurate
dose prediction within the intended domain
M. Rivard et al. (Vision 20/20), Med. Phys. 27, 2645-2658 (2010)
Basic recommendations of TG-186
• PreviousTPSQA/QCprocessstillvalid
• MaintainsTG43doseprescriptions
• Unlesssocietalrecommendationotherwise
• ModelͲbaseddosecalculationsshouldbe
performedinparallelwithTG43
• Radiationtransportusingthemostaccuratetissue
mediumcompositionsavailable
7
02/04/2013
TGͲ186CommissionTestsRequired
• AAPM/ESTRO recommend minimum standards
to evaluate conditions similar to those of clinical
brachytherapy practice based on established TG43 formalism
• Two levels of commissioning tests that should be
performed in addition to standard TPS QC/QA
Commission level I
• A direct comparison in reference-sized
homogenous water to AAPM TG-43 consensus
data: Water sphere with proper radius
• A first step since these parameters describe the
spatial dose distribution due to the physical
source model without significant consideration of
the surrounding environment
8
02/04/2013
High-Energy Brachytherapy Sources-examples
From Rivard
Low-Energy Brachytherapy Sources- examples
Laser Welded Ends
(0.1 mm wall)
Inorganic Substrate w/Cs-131 Attached
Gold X-Ray Marker (0.25 mm diameter)
Titanium Case (0.05 mm wall)
0.8 mm
4.1 mm
mm
4.0
4.5 mm
mm
4.5
IsoRay model CS-1 Rev2
9
02/04/2013
LevelIPass/FailCriteria:
• As recommended in the 2004 AAPM TG-43U*
report: 2.0% dose tolerance should be used for
agreement with AAPM consensus TG-43
dosimetry parameters
• Deviations > 2.0% should be carefully examined.
Clinical impact understood and documented
before clinical use
• See example by Mikell & Mourtada, MedPhys,
37(9):4733Ͳ43,2010
*Med. Phys. 31 (3), 2004
PartI
Example
CaseinTGͲ186
report
Med Phys, 37(9): 4733-43, 2010
10
02/04/2013
Ir-192 VS-2000 Model in BrachyVision TPS
Note1: The published clinical TG43 data for VS2000
source assumes a 15cm long NiTi cable.
Note2: VS2000 source is modeled with 1mm of NiTi cable
in BV-Acuros
Level I:
Radial dose
function
g(r) check
For BV 8.9
Acuros
Med Phys, 37(9): 4733-43, 2010
22
11
02/04/2013
F(r,theta) reference points:
Acuros vs. published TG43
Acuros Anisotropy vs published TG43 (reference points)
35
R=0.5cm
R=1.0cm
R=3.0cm
R=5.0cm
100*(Acuros-TG43)/TG43
30
25
R
20
15
10
5
0
-5
0
20
40
60
80
100
120
angle (degrees)
140
160
180
23
BV-Acuros vs. MCNPX Monte Carlo
%Dose
Med Phys, 37(9): 4733-43, 2010
12
02/04/2013
Abs. Dose Diff: 1mm cable vs. 15cm cable
Acuros Radial Dose Function g(r)*
g(r): larger differences near source.
*BrachyVision v8.8 with GBBS (Acuros 1.3.1)
13
02/04/2013
g(r) resolved! -defined catheter length matters!
27
%Ddiff Summary Statistics
Regionaway (T>=10deg)
Mean
1Ů
Max
Min
BV-TG43 vs MCNPX(15cm cable)
0.3
0.5
7.6
-2.1
BV-Acuros vs MCNPX(1mm cable)
0.3
0.4
4.7
-3.9
MCNPX(1mm) vs MCNPX(15cm)
0
0
0.4
-0.1
BV-Acuros vs BV-TG43
0
0.7
3.1
-9.9
Regionnear (T<10deg)
Mean
1ı
Max
Min
BV-TG43 vs MCNPX(15cm cable)
-0.1
2.7
6.6
-2.4
BV-Acuros vs MCNPX(1mm cable)
1.0
0.7
4.6
-1.8
MCNPX(1mm) vs MCNPX(15cm)
5.6
7.0
25
0.1
BV-Acuros vs BV-TG43
6.6
5.6
2821
-5.4
14
02/04/2013
LevelIIͲ CommissioningofMBDCalgorithms
• Comparisons of the 3D dose distributions
calculated with the clinical MBDCA-based TPS
• Verified against virtual phantoms, applicator,
and source loading mimicking clinical
scenarios
Factor-based vs Model-based
INPUT
TG43
Source
characterization
INPUT
MBDC
Source
characterization
Tissue/applicator
information
CALCULATION
OUTPUT
Superposition of
data from source
characterization
Dw-TG43
OUTPUT
CALCULATION
Model-Based
Dose Calculation
Algorithms
Dm,m
Dw,m
From Åsa Carlsson-Tedgren
15
02/04/2013
Seed/Applicator Model Accuracy
Requirements
• Patient CT grids (>1 mm voxel) are not adequate for
accurate modeling on the spatial scale of
brachytherapy sources and applicators.
• MBDCA vendors should use analytic modeling
schemes or recursively specify meshes with 1–10
ȝm spatial resolution.
• Vendors to disclose their geometry, material
assignments, and manufacturing tolerances to both
end users and TPS vendors (if responsible for data
entry and maintenance)
TG-186 Section IV-B
IfTPSApplicatorLibraryprovided
• Willeasetheverificationtask.
• Vendormustprovidevisualizationorreportingtoolsto
endusertoverifythecorrectnessofeachincluded
applicatorandsourcemodelagainstindependentdesign
specifications.
• Inaddition,TPSvendorsmustdisclosesufficient
informationregardingthemodelorrecursivemesh
generationtoallowverificationofthespatialresolution
requirementspecifiedinrecommendation(2)inTGͲ186
SectionIVͲB
16
02/04/2013
WhichTestCasestoPick?
• It is not reasonable to validate MBDCA implementation for
all possible combinations of source-applicator-geometry
• Pick scenarios relevant to your clinic
• Creation of an official Registry should help early adopters
for efficient work
• Developers of new test cases are strongly encouraged to
share their validation geometries through the Registry.
• AAPM Working Group with international membership is
developing a few registry cases
TGͲ186SectionIV.B:Applicators
• “ItistheresponsibilityoftheendͲuserclinical
physicisttoconfirmthatMBDCAdosepredictionsare
baseduponsufficientlyaccurateandspatially
resolvedapplicatorandsourcemodels,including
correctmaterialassignments,toavoidclinically
significantdoseͲdeliveryerrorpriortoimplementing
thedosealgorithmintheclinic.”
17
02/04/2013
Example: Solid Applicator Models in BV
A large number
of Varian’s
applicators
have been
modeled and
are supplied as
part of the
BrachyVision
standard
package
18
02/04/2013
Applicator Geometry & Composition Verification
Air pocket?
BV TPS
Applicator
LibrarySolid Model
part #: AL07334001
Ti screw?
Air?
In house MC model derived
from physical verification
and vendor CAD
Brachytherapy 10 (3): S36, 2011
Results – CT/MR ovoid geometry
19
02/04/2013
Results - CT/MR ovoid dosimetry
part #: AL07334001
Results - CT/MR ovoid dosimetry
part #: AL07334001
20
02/04/2013
Results - CT/MR ovoid dosimetry
part #: AL07334001
9 VS-2000 source
9 Applicator part #:
AL07334001
IJROBP, vol 83, No 3, pp
e414-e422, 2012
21
02/04/2013
(1)
Spatial distributions of the 3 factors
contributing to differences between
GBBS and TG-43:
(1) source and boundary,
(2) applicator,
(3) Heterogeneity*
*The contrast is overridden to
(1) muscle,
(2) no override,
(2)
(3) or bone.
22
02/04/2013
Top row:
combination of all three factors: source/boundary + applicator+ heterogeneity
Contrast overridden to muscle
Contrast not overridden to bone
Contrast overridden to bone
Bottom row: Contribution of the heterogeneity only. From left to right, the
GBBS with solid applicator with contrast overridden to muscle D(Y,N,M,M),
contrast not overridden D(Y,N,N,N), or contrast
overridden to bone D(Y,N,B,B).
ICRU Rectal Point Dose Impact of GBBS relative to TG-43
23
02/04/2013
ExampleforRegistryCase
BV v8.8
Shielded GYN Cylinder
(Part # GM11004380)/ 180o shield
GeneralWorkflowforaRegistryCase
a. Download test case plan from online Registry
b. Import DICOM test case
c. Run TG-43 based tests
• Removes errors not related to MBDCA!
d. Run MBDCA based tests “Clinical Site-specific”
e. Reference CT scanner calibration data
f. Select scoring medium
24
02/04/2013
25
02/04/2013
Virtual CT Object (IDL DICOM Creator) + Solid Applicator
from TPS Library
Pre-defined ROI
TPS applicator
Pre-defined Sphere
9Import Virtual Phantom into your TPS
26
02/04/2013
TG43 Run to Check Case Setup
9Source Strength
9Sources Positions
TG43RuntoCheckCaseSetup
Ref Data Set
Site TPS
TG43 Results
9Passed Setup Check: 100% agreement
27
02/04/2013
Ready now to run the Acuros test!
Run Acuros (v. 1.3.1) with CT
28
02/04/2013
Sagittal plane: points
Point
Dose (Gy)
P1
0.25
P2
6.0
P1-sup
0.26
P2-sup
5.88
P1-inf
0.25
P2-inf
5.83
P3
1.77
9Acuros: Check Reference Points Doses
9Acuros: Check DVH
29
02/04/2013
Use BV Plan Evaluation Tools
Need more analysis tools– like gamma index analysis
LevelIIPass/FailCriteria:
• What dose accuracy tolerance is required?
• Gamma-index as a first step, similar to
IMRT.
Stock et al “Interpretation and evaluation of the Ȗ index and the Ȗ
index angle for the verification of IMRT hybrid plans,” PMB 50,
2005
• Site-specific tolerance needed
• Further research required, but use your
common sense to clinical is important
30
02/04/2013
RegistryforLevelII
AAPM/ESTRO/ABGworkinggrouptotacklethisissue
Luc Beaulieu (chair), Frank-André Siebert (vice-chair)
Facundo Ballaster, Åsa Carlsson-Tedgren, Annette
Haworth, Goeff Ibbott, Firas Mourtada, Panagiotis
Papagiannis, Mark J Rivard, Ron Sloboda, Rowan Thomson
and Frank Verhaegen.
Strategy: Registry of validated cases with reference doses
calculations.
We will try to involved the vendor as much as possible to
make the cases compatible with all TPS.
Summary&Recommendations
• Commission your brachy TPS for clinical tasks
related to your clinic
• Direct comparison to reference plan having
heterogeneous conditions is mandatory
• Specific virtual phantoms mimicking the clinical
scenarios should be independently verified
against benchmarks
31
02/04/2013
Acknowledgements
TGͲ186
LucBeaulieu,CHUdeQuebec(Chair)
ÄsaCarlssonͲTedgren,LiUniversity
JeanͲFrançoisCarrier,CHUdeMontreal
SteveDavis,McGillUniversity
FirasMourtada,ChristianaCare
MarkRivard,TuftsUniversity
RowanThomson,CarletonUniversity
FrankVerhaegen,MaastroClinic
ToddWareing,Transpireinc
JeffWilliamson,VCU
WGͲMBDCA
LucBeaulieu,CHUdeQuebec(Chair)
FrankͲAndréSiebert,UKSH(ViceͲchair)
FacundoBallaster,Valancia
ÄsaCarlssonͲTedgren,LiUniversity
AnnetteHaworth,PeterMacCallumCC
GoeffreyIbbott,MDAnderson
FirasMourtada,ChristianaCare
PanagiotisPapagiannis,Athens
MarkRivard,TuftsUniversity
RonSloboda,CrossCancerInstitute
RowanThomson,CarletonUniversity
FrankVerhaegen,MaastroClinic
32