Monte Carlo Applications in Measurement Dosimetry
Transcript of Monte Carlo Applications in Measurement Dosimetry
![Page 1: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/1.jpg)
Monte Carlo Applications in Measurement DosimetryMeasurement Dosimetry
J Seuntjens and D W O RogersJ. Seuntjens and D.W.O. RogersMcGill University & Carleton University
CANADACANADA
Clinical Dosimetry for Radiotherapy, 2009 AAPM Summer School
![Page 2: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/2.jpg)
QuestionQuestion
A th k f t i th AAPM TG 51• Are the kQ factors in the AAPM TG‐51 calculated purely by Monte Carlo techniques?
![Page 3: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/3.jpg)
Educational ObjectivesEducational Objectives
• Understand measurement dosimetryUnderstand measurement dosimetryfundamentals
• Understand the role of Monte Carlo transport• Understand the role of Monte Carlo transport in measurement dosimetry
T i h ibili i MC i• To appreciate the possibilities MC can give us in terms of making measurement dosimetrymore accurate
![Page 4: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/4.jpg)
Clinical Measurement DosimetryClinical Measurement Dosimetry
• Measurement of dose using detectors:Measurement of dose using detectors:– Ionization chambers
Diodes– Diodes
– TLD’s
R di h i Fil– Radiochromic Film
– MOSFETS
![Page 5: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/5.jpg)
Classification of dosimetersClassification of dosimeters
![Page 6: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/6.jpg)
Measurement dosimetryMeasurement dosimetry
D t t diDetector reading
LinearityDose rateEnergy
dependence
Intrinsicenergy dependenceL ffLET effects…
![Page 7: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/7.jpg)
A few known cases…A few known cases…
Ionization chamber 1 1
Fricke dosimeter 1 1*dosimeter 1 1
Water calorimeter 1 1calorimeter
*: away from region of supra‐linearity
![Page 8: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/8.jpg)
The meaning…The meaning…
“Calibrating” the detector means determining:
by explicitly measuring:
![Page 9: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/9.jpg)
is the “energy dependence” of the detectorthe detector
For absorbed dose:
For air kerma:For air kerma:
![Page 10: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/10.jpg)
How does TG51 fit into this?How does TG51 fit into this?
![Page 11: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/11.jpg)
Role of Monte Carlo calculations in dmeasurement dosimetry…
…determining fQ
![Page 12: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/12.jpg)
The correction factors…The correction factors…
![Page 13: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/13.jpg)
![Page 14: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/14.jpg)
What is a MC simulation?What is a MC simulation?
“The Monte Carlo technique for the simulation of qthe transport of electrons and photons through bulk media consists of using knowledge of the probability distributions governing the individualprobability distributions governing the individual interactions of electrons and photons in materials to simulate the random trajectories of individual
ti l O k t k f h i l titiparticles. One keeps track of physical quantities of interest for a large number of histories to provide the required information about the p qaverage quantities”
TG105 Rogers and Bielajew
![Page 15: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/15.jpg)
What is a Monte Carlo simulation?What is a Monte Carlo simulation?
Particles move in discrete steps from Particles move in discrete steps from
one interaction site to the next
(Cross sections)
Random Number GeneratorRandom Number Generator
![Page 16: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/16.jpg)
Putting a Monte Carlo simulation together:
Interaction probabilityInteraction probability
Random N mberRandom N mber
Monte Carlo probability distributionsprobability distributions
Number GeneratorNumber Generator
simulation
Dose distributions, fluence spectra, ...fluence spectra, ...
![Page 17: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/17.jpg)
Simple photon simulationSimple photon simulation
• say: paircomptontotal += cm -1
• select 2 random numbers R1, R2– uniform between 0 and 1
– whole careers devoted to doing this
– cycle length now 10>40
17/43
![Page 18: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/18.jpg)
Photon transport (cont)Photon transport (cont)
How far does photon go before interacting?
/X = ln(R1) cm/X = -ln(R1) total
is exponentially distributed [0,)
with a mean of total/1
18/43
tota
![Page 19: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/19.jpg)
Photon transport (cont)Photon transport (cont)
After going x, what interaction occurs?
t t l
compton<R2if
then a compton scatter occurs
total
p
otherwisea pair production event occurs
19/43
a pair production event occurs
![Page 20: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/20.jpg)
Sampling an interaction type: e.g. 3 MeV photon in Pb
1
28 %
1Pair production
6 % Photo‐electric effect
66 % Compton scatteringRNG
0
![Page 21: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/21.jpg)
Random sampling from probability di t ib ti R j ti th ddistributions: Rejection method
MAX
abilit
ypr
oba
El t E (M V)0
b
E1 E2
Electron Energy (MeV)a b
![Page 22: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/22.jpg)
How is simulation used?How is simulation used?
• score whatever data wanted
– average distance to interactiong
– how many of each type
energy deposited by each type– energy deposited by each type
– etc
• more useful in complex cases
22/43
p
![Page 23: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/23.jpg)
A simple MC simulationA simple MC simulation
photon2 31 electron
Compton scattering Scoring
region
3 : sample type of interaction
1 : sample particle energy, direction,
g
4 : sample direction, energy, … of new particles
2: sample distance to interaction
starting position, ...
p
![Page 24: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/24.jpg)
10 MeV photon on lead
along incidentphoton positrons red
.p
electrons bluephotons yellow
incident
incident
24
![Page 25: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/25.jpg)
10 MeV electrons on water from right
electrons bluephotons yellow
25
![Page 26: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/26.jpg)
Condensed history technique fore‐ transportCondensed history technique fore transport
• as electrons slow down, they have manyas electrons slow down, they have many interactions
• Berger’s grouping into condensed history stepsmade Monte Carlo transport of electrons feasiblemade Monte Carlo transport of electrons feasible.– individual scattering events grouped via multiple‐scattering theories l l t d i t t i t d– low‐energy‐loss events grouped into restricted stopping powers
• increases efficiency by decreasing time,T, (a lot)• modern transport mechanics algorithms are very sophisticated in order to maximize step size while maintaining accuracy (to gain speed).
26/43
g y ( g p )
![Page 27: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/27.jpg)
e‐ transport is much more complexe transport is much more complex
hard collisionscreate secondaries eg -rays / bremeg -rays / brem
ft lli i soft collisions -grouped -multiple scatter p-restricted energy loss
condensed history technique: group many individual
27
condensed h story techn que group many nd v dual interactions into steps
![Page 28: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/28.jpg)
Energy deposition near boundaryEnergy deposition near boundary
Bielajew, Rogers, Nahum (Phys Med Biol. 1985, Vol. 30, No. 5, 419‐427)
![Page 29: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/29.jpg)
The step size is parameter‐controlledThe step size is parameter controlled
• We want long steps to speed up theWe want long steps to speed up the simulations
• But … we want short steps to make theBut … we want short steps to make the calculation accurate
• Step control:Step control:– Distance to next catastrophic interaction– Fractional energy loss / stepgy / p– Absolute step length maximum– Proximity of boundaryy y
![Page 30: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/30.jpg)
Back to dosimetry…
Farmer type ionization chamberchamber
photons and
secondary electrons
David Shipley, NPL
![Page 31: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/31.jpg)
Ionization chamber formalism for photon beamsIonization chamber formalism for photon beams
Dgasphotons of energy E
gas
wallen
gasAA
LKD
)1( flwall
air
en
wallairgas AA
LgKD
)1(
![Page 32: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/32.jpg)
Ionization chamber formalism for photon beamsIonization chamber formalism for photon beams
ith l t i b t hl i iwith: average energy lost in bremsstrahlungin air
restricted stopping power ratio
ggas
L
restricted stopping power ratio
cavity gas‐to‐cavity wallwall
L
wall
i
en
mass‐energy absorption coefficient
ratio wall‐to‐airair ratio wall to air
Awallwall correction (attenuation & scattering)Aflfluence perturbation correction factor
![Page 33: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/33.jpg)
Rogers, Med Phys 20, 319 (1993)
![Page 34: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/34.jpg)
Improvements in the EGS4 system in the lead up h EGS (I K k )to the EGSnrc system (I. Kawrakow)
• new any angle multiple elastic scattering theory b d d R h f dbased on screened Rutherford c.s.
• improved electron‐step algorithm• improved electron‐step algorithm
• correct treatment of discrete interactions
• improved evaluation of energy loss
• exact boundary crossing
![Page 35: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/35.jpg)
Ionization chamber simulation at 60Co:EGS4/PRESTA artifactsEGS4/PRESTA artifacts
Artifact Aluminium20%
Carbon20%
Aluminium1%
Carbon1%20% 20% 1% 1%
electronstep
‐9.0% ‐5.0% ‐1.4% ‐0.7%p
BCA +3.4% +2.6% +1.5% +0.9%energy loss +0.3% +0.5% +0.0% +0.0%discreteinteractions
+0.7% +0.7% +0.7% +0.7%
Totals ‐4.6% ‐1.2% +0.8% +0.9%
Kawrakow, Med. Phys. 27 499 (2000)Kawrakow, Med. Phys. 27 499 (2000)
![Page 36: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/36.jpg)
Conditions for accurate MC calculations of dosimeters in radiation therapy
1. MC code must be consistent, i.e., results must be in agreement with fundamentalmust be in agreement with fundamental dosimetry theorems
2 MC code must use accurate cross2. MC code must use accurate cross sections, realistic geometry description and beam descriptionand beam description
3. Precision of results
![Page 37: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/37.jpg)
Fano theoremFano theorem
“Under conditions of equilibrium in an infinite medium the particle fluence will not be alteredmedium, the particle fluence will not be altered
by density variations from point to point”
U. Fano, 1954
![Page 38: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/38.jpg)
Fanocavity implementation in MCFanocavity implementation in MC
flwall
wallen
gas
airgas AAL
gKD
)1(
• gas and wall material are set the same, except f d i
flwallairwall
airgas g
)(
for density:
11
fl
gasAand
L
• remove the effect of photon attenuation:
11
fl
wallAand
• remove the effect of photon attenuation: unweighting of fluence– regeneration technique: Awall=1– normalization on Awall
![Page 39: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/39.jpg)
Fano cavity implementation in MC (cont’d)Fano cavity implementation in MC (cont d)
• since:wall
enKK
)1(since:
• Accuracy test:air
enairwallcoll gKK
)1(,
• Accuracy test:
llllKFanoD )( wallcollunwgas KFanoD ,, )(
Test on accuracyf l t MC t
Straightforwardof electron MC system photon Monte Carlo
![Page 40: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/40.jpg)
![Page 41: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/41.jpg)
![Page 42: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/42.jpg)
Conditions for accurate MC calculations of dosimeters in radiation therapy
1. MC code must be consistent, i.e., results must be in agreement with fundamentalmust be in agreement with fundamental dosimetry theorems
2 MC code must use accurate cross sections2. MC code must use accurate cross sections, realistic geometry description and beam descriptiondescription
3. Precision of results
![Page 43: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/43.jpg)
Effect of different cross‐section options on dose calculated in 3Coptions on dose calculated in 3C
chamber
![Page 44: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/44.jpg)
Cross section effects more important at low energiesenergies…
C552 plastic: ICRU 37 composition by weight: H: 2.4%, C: 50.2%, O: 0.5%, F: 46.5%, Si: 0.4%
mass spectrometrical analysis
B 17 4 Ti 3 0
mass spectrometrical analysis
Element Elementppm ppmBNaMgAl
17.425.09.820 0
Ti 3.0Vn 3.5Mn 3.4F 70 0Al
PS
20.0250.012.0
Fe 70.0Rb 8.8Cr <50
![Page 45: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/45.jpg)
![Page 46: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/46.jpg)
![Page 47: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/47.jpg)
Conditions for accurate MC calculations of dosimeters in radiation therapy
1. MC code must be consistent, i.e., results must be in agreement with fundamentalmust be in agreement with fundamental dosimetry theorems
2 MC code must use accurate cross sections2. MC code must use accurate cross sections, realistic geometry description and beam descriptiondescription
3. Precision of results
![Page 48: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/48.jpg)
QuestionQuestion
• What is more challenging?What is more challenging?– Doing a MC treatment planning dose calculation
Doing a MC dosimeter response calculation– Doing a MC dosimeter response calculation
– Staying awake during this presentation
![Page 49: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/49.jpg)
MC method is a statistical methodMC method is a statistical method
• Type A uncertaintiesType A uncertainties…
• Need many many many histories
d i d i h i• Need variance reduction techniques– The events that lead to the effects of interest are rare
![Page 50: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/50.jpg)
Variance reduction techniques (VRTs)Variance reduction techniques (VRTs)
• A VRT is a method which increases the• A VRT is a method which increases the efficiency for some quantity of interest by decreasing s2 for a given N while not biasing thedecreasing s2 for a given N while not biasing the result.
h f h– they often increase time per history
– VRTs may simultaneously make s2 for some other i iquantity increase
– egpathlength shrinking will improve the efficiency for dose near the surface but decrease the efficiency for dose at depth
50/43
![Page 51: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/51.jpg)
Variance reduction techniquesVariance reduction techniques
• for a review, see Sheikh‐Bagheri et al’s 2006 AAPM , gsummer school chapter
http://www.physics.carleton.ca/~drogers/pubs/papers/SB06.pdf
• examples• examples– splitting (brem splitting: UBS, DBS; in‐phantom)
– Russian roulette
– interaction forcing
– track repetition
– STOPS (simultaneous transport of particle sets)
– cross section enhancement
51/43
![Page 52: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/52.jpg)
Splitting, Roulette &particle weight
1 wi = 10 wf 10 wi = 1 wfi f i f
Split Roulettep≈
52/43from Sheikh-Bagheri’s 2006 summer school lecture
![Page 53: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/53.jpg)
Photon cross section enhancementPhoton cross section enhancementEnhance cross section in
i f i t t W i hregion of interest. Weigh subsequent particles accordingly.
Wulff, Zink, Kawrakow Med Phys 35, 1328(2008)
![Page 54: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/54.jpg)
Stopping power ratios (SPRs)Stopping power ratios (SPRs)
SPRs are central in measurement dosimetrySPRs are central in measurement dosimetry
SPRs are theoretical quantities, not always an accurate representation of f(Q)
![Page 55: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/55.jpg)
Stopping power ratios (cont’d)Stopping power ratios (cont d)
T1 T2 T3 T1 3 T4 T5
T1T
T3T4T5
T2
![Page 56: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/56.jpg)
sw,air(z) for plane parallel bremsstrahlung beams
no e contamination central a is‐ no e ‐ contamination, central axis
60Co
10 MV
20 MV
30 MV
Andreo and Brahme, Phys. Med. Biol. , 31, 839 (1986)y ( )
![Page 57: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/57.jpg)
One of the very many impacts…One of the very many impacts…
• David Burns – by choosing:David Burns by choosing:
dref= 0.6R50‐0.1 [cm]
k h S h f imakes the SPR at the reference point a very simple function of R50 only…
This is the basis of electron dosimetry in TG‐51 and TRS‐398and TRS 398
![Page 58: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/58.jpg)
Small fields
Sanchez‐Doblado et al Phys. Med. Biol 48Biol.482081‐2099 (2003)
58
![Page 59: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/59.jpg)
Narrow1.5 mm fieldRatio of dose to water to dose to air
1.70 averaged over cavity volume
Collecting electrode diameter: 1 5 mm 1.40
1.50
1.60
diameter: 1.5 mmSeparation: 1 mm
1.20
1.30
1.40Dw/D
air
stopping power ratio w/air
0 90
1.00
1.10ratio w/air
Paskalev, Seuntjens,
Off axis distance (mm)
0.80
0.90
0 2 4 6 8
Podgorsak (2002) AAPM Proc. Series 13, Med.
Phys. Publishing, Madison, Wi, 298 – 318.
Off‐axis distance (mm)
![Page 60: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/60.jpg)
Choice of ΔChoice of Δ
• Δ is not well‐defined: “the lowest energy ofΔ is not well defined: the lowest energy of electrons, which can just cross the cavity”
• Shown that use of mean chord length 4V/S• Shown that use of mean chord length 4V/S improves accuracy
![Page 61: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/61.jpg)
Wall correction factors for air kermab d dbased dosimetry
• Wall corrections for air kerma basedWall corrections for air kerma based dosimetry have been calculated since the late 70’s (Nath and Schulz 1978)70 s (Nath and Schulz 1978)
• These calculations were the basis of TG‐21 and TRS 277 protocol dosimetry walland TRS‐277 protocol dosimetry wall correction factors
Th l i i h l• The results are not sensitive to the electron transport accuracy
![Page 62: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/62.jpg)
Wall correction factors
![Page 63: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/63.jpg)
Wall correction factors (alternatively)
![Page 64: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/64.jpg)
![Page 65: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/65.jpg)
Effect of errors in transport on l l f f h bcalculations of Kwallof 3C chamber
![Page 66: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/66.jpg)
Correlated samplingCorrelated sampling
• To calculate ratios of doses or differences ofTo calculate ratios of doses or differences of doses between geometries that are very similar
• The statistical uncertainty on the ratio orThe statistical uncertainty on the ratio or difference of quantities is lower than on the absolute quantitiesabsolute quantities
• Applied by Ma and Nahum in the early 90’s in a variety of cases related to ionization chamber andvariety of cases related to ionization chamber and other dosimeter correction factors
![Page 67: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/67.jpg)
Simulated only once
D1D1
store state
![Page 68: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/68.jpg)
D2D2
retrieve state
calculate D1, D2, and D1/D2
![Page 69: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/69.jpg)
Correlated sampling efficiency gainCorrelated sampling efficiency gain
Buckley et al, 2004Med Phys 31, 3425Med Phys 31, 3425
![Page 70: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/70.jpg)
Clinical reference dosimetry – central l d h belectrode corrections – photon beams
Buckley et al, 2004Med Phys 31, 3425Med Phys 31, 3425
At Co:Dgr/DAl=0.9927±0.0001
![Page 71: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/71.jpg)
Clinical reference dosimetry – central l d l belectrode corrections – electron beams
Buckley et al, 2004Med Phys 31, 3425Med Phys 31, 3425
Protocols assume no correction for graphite wallfor graphite wall
![Page 72: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/72.jpg)
Clinical reference dosimetry – wall f l bcorrection factors – electron beams
NACP02
TG‐51 &TRS‐398TG 51 &TRS 398 assumePwall = 1Pwall
![Page 73: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/73.jpg)
Clinical reference dosimetry – wall fcorrection factors
The Almond ‐ Svensson 2‐component model
![Page 74: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/74.jpg)
Clinical reference dosimetry – wall fcorrection factors
TG‐51 and TRS‐398base their database their data on the Almond ‐Svensson2‐component model
![Page 75: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/75.jpg)
P l correction factor ‐ recallPrepl correction factor recall
![Page 76: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/76.jpg)
P lin dosimetry protocolsPreplin dosimetry protocols
Electron beams“well‐guarded” plane‐parallel chambers:Prepl = 1
cylindrical chambers:measured P at d (= P )cylindrical chambers:measured Prepl at dmax(= Pfl)
Photon beams
plane‐parallel chambers: Prepl = 1
cylindrical chambers:measured Prepl = Pgry repl gr
TRS‐398: the uncertainty in value of Prepl in photon beams “is one of the major contributions to the
76/44
final uncertainty in kQ”
![Page 77: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/77.jpg)
P l – four methods (L. Wang)Prepl four methods (L. Wang)
MethodSPR Calc of different
quantities required
HDA Layer thickness ofHDA Layer thickness of HDA slab
LDW Fluenceperturbationperturbation
FLU Fluence in cavity proportional to fluence in medium
Note: Calculation not done using correlated sampling; the correlations are weakcorrelations are weak
![Page 78: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/78.jpg)
Prepl for NACP02 chamber in electron beams and 60Co beambeams and Co beam
Calculation is done at dref for electron beams & at refdepth 5 cm for 60Co beam
SPR FLU HDA LDW
6 MeV 0.9956 0.9977 0.9976 0.9959 6 MeV (0.06%) (0.10%) (0.08%) (0.06%)
18 MeV 1.0001 (0.06%)
1.0007 (0.06%)
1.0011 (0.07%)
1.0005 (0.05%)
60Co 1.0059 (0.10%)
1.0063 (0.10%)
1.0062 (0.10%)
1.0065 (0.10%)
In all dosimetry protocols: P l= 1
78/44
In all dosimetry protocols: Prepl= 1
![Page 79: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/79.jpg)
Prepl for Farmer chamber in 60Co beamPrepl for Farmer chamber in Co beam
Cavity diameter: 6 mmCavity length: 2 cmDepth in water: 5 cm
SPR FLU HDA LDW
0 9963 0 9952 0 9969 0 9974
P value in dosimetry protocols:
0.9963 (0.08%)
0.9952 (0.08%)
0.9969 (0.09%)
0.9974 (0.07%)
Prepl value in dosimetry protocols:AAPM 0.992
IAEA 0 988
79/44
IAEA 0.988
![Page 80: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/80.jpg)
Prepl in high energy photon beamsrepl g gy p
80/44
![Page 81: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/81.jpg)
Relative dosimetry: the need for correction factors at multiple depthsat multiple depths
NACP h bNACP chamber in 6 MeV and 20 M V lMeV electrons
Buckley and Rogers, Med. Phys. 33 455 ‐ 464 (2006)
![Page 82: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/82.jpg)
Further improvements…Further improvements…Excruciatingly detailed
egs++geometry modeling egs++ libraries
D egs chamberDose calculations at multiple
egs_chamber
at multiple depths, done efficientlyy
Wulff, Zink, Kawrakow Med Phys 35, 1328(2008)
![Page 83: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/83.jpg)
Wulff, Zink, Kawrakow Med Phys 35, 1328(2008)
![Page 84: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/84.jpg)
What to do with all this information?
Reconciling measurements andcalculations in the build‐up region by adjusting EPOM
‐>simple if it works; but is chamber dependent!
( ) 33Kawrakow (2006)Med. Phys. 33, 1829
![Page 85: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/85.jpg)
![Page 86: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/86.jpg)
Inflection point variabilityMatch @ inflection point
Ververs, Siebers and Kawrakow 2008 ‐ TU‐C‐AUD B‐1
![Page 87: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/87.jpg)
Correction factors in nonstandard beamsCorrection factors in nonstandard beams
14 IMRT fields
![Page 88: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/88.jpg)
ConclusionsConclusions
• Have reviewed:– MC techniques as applied to measurement dosimetry– Attempted to give some background & examplesM C l l i l i• Monte Carlo can play an important role in measurement dosimetry
• Techniques have drastically developedTechniques have drastically developed• There is a need to design methods to make practical use of the high quality data coming out especially for relative dosimetry and dosimetry of nonstandard beams
![Page 89: Monte Carlo Applications in Measurement Dosimetry](https://reader034.fdocuments.us/reader034/viewer/2022051522/5889b2211a28abd1358b50e0/html5/thumbnails/89.jpg)
Thank YouThank You