TomoTherapy: Combining Imaging and...
Transcript of TomoTherapy: Combining Imaging and...
1
UNIVERSITY OF WISCONSINSCHOOL OF MEDICINEAND PUBLIC HEALTH
Robert JerajUW Paul P. Carbone Comprehensive Cancer Center,
Departments of Medical Physics, Human Oncology and Biomedical Engineering, University of Wisconsin, Madison, Wisconsin, USA
Jozef Stefan Institute, Ljubljana, Slovenia
JOZEF STEFAN INSTITUTE
TomoTherapy: Combining Imaging and Therapy
Radiotherapy (RT)
Conventional RT:– Jaws are used to shape the beams (rectangular around tumor)
Conformal RT (CRT):– MLCs are used to shape the beams (conformal to tumor shape)
Intensity modulated RT (IMRT):– MLCs are used to shape the beams and modulate intensity
Intensity modulated radiotherapy
Cone beam (non-rotational) IMRT– Treating cone by cone (port by port)– Using dynamic MLC– Techniques:
– Step-and-shoot (multiple static field) technique– Sliding window (dynamic MLC) technique– Several alternative techniques (e.g., IMAT,
dynamic arc therapy)
2
Intensity modulated radiotherapy
Fan beam (rotational) IMRT– Treating slice by slice– Using binary MLC– Techniques:
– Serial tomotherapy (NOMOS)– Helical tomotherapy (TomoTherapy)
Helical tomotherapy –design basis
Helical fan-beam IMRT delivery is simple, fast and effectiveCT is the most important imaging modality for radiotherapyLinac on a CT is better than a CT on a linac
– ring gantry is more stable than a C-arm gantry– CT gantry allows faster rotation– no possibility of rotational collisions– coplanar delivery is simpler
Single energy sufficientSimple binary MLC modulating the fan beamAccurate CT couch
Helical tomotherapy
3
LinacLinac
Pulse FormingPulse FormingNetwork andNetwork andModulatorModulator
DetectorDetectorBeam StopBeam StopHigh VoltageHigh VoltagePower SupplyPower Supply
ControlControlComputerComputer
MagnetronMagnetron
Data Data AcquisitionAcquisitionSystemSystem
CirculatorCirculator
Gun BoardGun Board
Helical tomotherapy components
1 Beam 5 Beams 11 Beams
17 Beams 25 Beams 51 Beams
mm mm mm
mmmmmm
More Beams = more homogeneity and conformity
Helical scan/delivery
4
Open
Closed
For low intensity a leaf is open for a short time during the angular segment
One rotation is divided into 51 segments
For high intensity a leaf is open for a long time during the angular segment – the highest intensity determines the rotation speed
Intensity modulation in helical tomotherapy
Bea
m d
irec
tion
& c
ouch
pos
ition
Leaf #
Beam Intensity
Beam delivery
Beam delivery
5
Beam delivery
Beam delivery
Prostate treatment delivery
0 to 30% 30 to 90% 90 to 100%
Dose Rate Cumulative Dose
6
Bilateral breast case
Multiple metastasis
Breast
Parotids
Kidneys
Lens
Liver
Bone
Total body irradiation (TBI) case
7
Helical tomotherapy MC model
Radiation delivery unit
Irradiated object (patient)
Detector
Static tomotherapy beam
The beam is not flat because helical tomotherapydoes not have a flattening filter - no need for IMRT
0 5 10 15 200
20
40
60
80
100
PDD
[%]
Depth [cm]
Real system MC system
-300 -200 -100 0 100 200 3000
20
40
60
80
100
Dos
e [%
]
Distance [mm]
Real system MC system
Single beam
The first step in Monte Carlo modeling is to tune the MC model to match the measurements
8
Beam energy dependence
0 5 10 15 20 25 300.0
5.0x10-17
1.0x10-16
1.5x10-16
2.0x10-16
2.5x10-16
3.0x10-16
E0 = 5.0 +/- 0.5 MeV E0 = 5.5 +/- 0.5 MeV E0 = 6.0 +/- 0.5 MeV E0 = 6.5 +/- 0.5 MeV
Dos
e [G
y/s.
p.]
Depth [cm]
⎟⎟⎠
⎞⎜⎜⎝
⎛∝ 2
030
2ln
cmE
EDe
Strong dependence of the depth dose curve with energy, but not significant %DD dependence –not very sensitive to commissioning
-20 -10 0 10 200
20
40
60
80
100
Pho
ton
fluen
ce [%
]
Distance [cm]
Helical tomotherapy Conventional linac
-4 -2 0 2 40
20
40
60
80
100
Pho
ton
fluen
ce [%
]
Distance [cm]
5cm 2cm 0.5cm
Beam profiles
The lateral beam profile has a characteristic “cone” profile - ~2-times higher in the center
The axial beam profile has a very sharp penumbra
Lateral field sensitivity
-2 -1 0 1 20
20
40
60
80
100
Dos
e [%
]
Distance [cm]
Normal incidence -0.1 mm shift -0.2 mm shift 5 degree tilt
-2 -1 0 1 20
20
40
60
80
100
Dos
e [%
]
Distance [cm]
Normal incidence -0.1 mm shift -0.2 mm shift 5 degree tilt
The beam profile is very sensitive to the linacelectron beam position – particularly for narrow fields
9
Shielding - leakage
-10 -5 0 5 101E-4
1E-3
0.01
0.1
1
Pho
ton
fluen
ce r
atio
Distance [cm]
5cm 2cm 0.5cm
-4 -2 0 2 41E-3
0.01
0.1
1
Dos
e ra
tio
Distance [cm]
5cm 2cm 0.5cm
Helical tomotherapy was designed to increase shielding in the forward direction
» 104 attenuation at 10 cm for narrow fields
Leakage compared to other systems
Ramsey et al, J App Clin Med Phys 7, 2006, 11
120.9
135.9
118.8
133.6
126.3
Total
-4 %03.0117.9Tomo(6 MV)
+8 %5.621.2109.120 MVIMRT
-6 %1.14.2113.420 MV3D CRT
+6 %016.9116.76 MVIMRT
0 %03.4122.96 MV3D CRT
ChangeFrom 6MV
3D CRT
Neutron(Reft)
PhotonOut-Field(Ramsey
and Reft)
In-Field(Aoyama)
IMRT Integral Dose
Units are in Gy-literCourtesy of Thomas R. Mackie, University of Wisconsin, Madison, WI
10
No flattening filter –radiation safety concern
0 1 2 3 4 5 60.00
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
Full target Half target No target
Pho
ton
spec
trum
[nor
mal
ised
]Energy [MeV]
0.0 0.2 0.4 0.6 0.8 1.0
0.7
0.8
0.9
1.0
1.1
1.2
Rel
ativ
e ou
tput
Target thickness [relative]
-20 -10 0 10 200
20
40
60
80
100
Pho
ton
fluen
ce [%
]
D istance [cm]
Full target Half target No target
Helical tomotherapy spectral characteristics
0 1 2 3 4 5 60.0
0.1
0.2
0.3
0.4
0.5
0.6
Helical tomotherapy Varian Clinac 2100C Varian Clinac 600C/D
Spec
trum
Energy [MeV]
0 1 2 3 4 5 6 70.000
0.005
0.010
0.015
0.020
0.025
0.030
0.035
0.040
0-5 cm 5-10 cm 10-15 cm 15-20 cm
Phot
on s
pect
rum
Energy [MeV]
Helical tomotherapy has a very comparable spectrum to other linacs (6 MV)
There is very insignificant in-field spectral change (no flattening filter)
Treatment and imaging beams
0 1 2 3 4 5 60.0
2.0x10-6
4.0x10-6
6.0x10-6
8.0x10-6
1.0x10-5
1.2x10-5
1.4x10-5
1.6x10-5
1.8x10-5
Phot
on s
pect
rum
[1/M
eV c
m2 /s
.p.]
Energy [MeV]
Treatment MVCT Imaging
Helical tomotherapy creates the imaging beam by reducing energy (from ~6 MeV to 3.5 MeV)
11
PhotonSource
Thin tungstenFoils
High pressureXenon gas
Fast electrons are generated in tungsten plates and detected in ion chambers
DQE ~ 25%
Detector system
Signal at the detectors for one linac pulse
Ion
izati
on
Sig
nal
Detector Element Number (Position)
The detector response increases from the center of the detector due to an increased cross-section of tungsten.
PhotonSource
Tungsten
Detector response across the detector
0
2
4
6
8
10
0 100 200 300 400 500 600
Wall thickness, t (μm)
Dep
osite
d en
ergy
in g
as (a
.u.)
primary signal
brass
Simulations for a fixed detector element width (w = 1.5mm)
t
ww
t
Optimization of the detector design(primary signal)
12
0
2
4
6
8
10
0 100 200 300 400 500 600
Wall thickness, t (μm)
Dep
osite
d en
ergy
in g
as (a
.u.)
primary signal
brass tungsten
t
ww
t
Optimization of the detector design(primary signal)
Simulations for a fixed detector element width (w = 1.5mm)
0
2
4
6
8
10
0 100 200 300 400 500 600
Wall thickness, t (μm)
Dep
osite
d en
ergy
in g
as (a
.u.)
total crosstalk
primary signal
brass tungsten
Optimization of the detector design(cross-talk)
MC simulations of the detector
0 100 200 300 400 500 600 7000.0
1.0x10-9
2.0x10-9
3.0x10-9
4.0x10-9
5.0x10-9
Dose
[MeV
/s.p
.]
Detector channel
Treatment beam MVCT beam
0 100 200 300 400 500 600 7000
20
40
60
80
100
120
Dose
[%]
Detector channel
Treatment beam MVCT beam
Detector signal is the lowest in the center (low septa plate cross-section), but the approximately follows the photon profile
13
Monte Carlo provides useful “immeasurable” information
0 100 200 300 400 500 600 7000
20
40
60
80
100
120 Dose Electron fluence Photon fluence
Dos
e or
flue
nce
[nor
mal
ised
]
Detector channel
Detector signal is directly proportional to the electron fluence; photon and electron fluencessimilar in the septa plates, not in the gas
Comparison to experiment
0 100 200 300 400 500 600 7000
20
40
60
80
100
120
Dose
[%]
Detector channel
Normal source Wider source (3mm) Shifted source (1mm)
0 100 200 300 400 500 600 7000
20
40
60
80
100
120 Experiment Monte Carlo model
Dete
ctor
sig
nal o
r Dos
e [n
orm
alai
sed]
Detector channel
Agreement with the experiment is not good. Why?• source description• detector geometry Sensitivity studies!• transport physics
Density Plugs +3% Contrast
-6% ContrastWater
1 cGy
UW Tomotherapy Unit UW Trilogy Unit
6 cGy
Water
Comparison of tomotherapy with CBCT
14
MVCT
Planning CT
Registered Images
Image registration
Planning CT MVCT CT
Image registration
Superposition of the planned dose distribution over the daily CT enables optimal conformity of the dose to the daily anatomy
Internal anatomy is changing
Start of treatment
7 days later
2 weeks later
3 weeks later
1 month later
Courtesy of Tim Holmes, St. Agnes Cancer Center, Baltimore, MD
15
Difference between the planned dose and actual delivered dose is significant and can, at least in principle, be compensated for with treatment adaptation
Treatment dose ≠ planned dose
Dose calculated on last days anatomy minus the original planned doseGreen = no difference in planned vs. treatment doseRed = difference in planned vs. treatment dose
Courtesy of Tim Holmes, St. Agnes Cancer Center, Baltimore, MD
Kupelian et al, Int J Rad Oncol Biol Phys 63, 2005, 1024
Tumor shrinkage
Adaptive radiotherapy3-D Imaging
OptimizedPlanning
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration
16
Adaptive radiotherapy3-D Imaging
OptimizedPlanning
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration
• Off-line planning
• On-line planning
Adaptive radiotherapy3-D Imaging
OptimizedPlanning
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration
• Patient positioning• rigid body• anatomy deformation
• Daily contour generation
Adaptive radiotherapy3-D Imaging
OptimizedPlanning
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration
• Fast plan modifications
17
Adaptive radiotherapy3-D Imaging
OptimizedPlanning
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration
• Deform patient imageto map back CTs to a common reference CTand accumulate dose
• Decide if re-optimizationis needed
Adaptive radiotherapy3-D Imaging
Re-optimizePlan
CT+ Image Fusionor Registration
TreatmentWith DeliveryVerification
DoseReconstruction
DeliveryModification
DeformableDose
Registration • Define the re-optimisationobjective function
• Create a new plan
•Define the next checkingpoint
Conclusions
Helical tomotherapy is a designed image-guided radiotherapy systemMonte Carlo simulations of the complete system include both, characterization of the treatment beam and imaging beamMonte Carlo simulations are invaluable in understanding and optimization of such a complex systemMany challenges and opportunities for Monte Carlo simulations in the future
18
Thank you for your attention