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


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


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


OptimizedPlanning



DoseReconstruction


DeformableDose

Registration

• Off-line planning

• On-line planning


OptimizedPlanning



DoseReconstruction


DeformableDose

Registration

• Patient positioning• rigid body• anatomy deformation

• Daily contour generation


OptimizedPlanning



DoseReconstruction


DeformableDose

Registration

• Fast plan modifications

17


OptimizedPlanning



DoseReconstruction


DeformableDose

Registration

• Deform patient imageto map back CTs to a common reference CTand accumulate dose

• Decide if re-optimizationis needed


Re-optimizePlan



DoseReconstruction


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

tomotherapy: combining imaging and...

Documents