simulation of ct-images: from x-ray spectrum to ct...
TRANSCRIPT
Simulation of CT-images:from X-ray spectrum to CT-numbers
Simulation of CT-images:from X-ray spectrum to CT-numbers
S. Qamhiyeh1), A.Wysocka-Rabin2), M. Ellerbrock1), C. Penssel3), M. Kachelriess3) and O. Jaekel1)
1) German Cancer Research Center (DKFZ), Medical Physics, Heidelberg, Germany2)Andrzej Soltan Institute for Nuclear Studies, Department of Accelerator Physics, Swierk, Poland3) Friedrich-Alexander-University, Institute of Medical Physics, Erlangen-Nuremberg, Germany
MCNEG / 2007NPL, Teddington 28.03.2007 [email protected]
Ion therapy vs. Photon radiotherapy
Bragg-peaks for different Carbon ion energies
Introduction (CT-calibration)Introduction (CT-calibration)
)1(1000numberCT −⋅=−wµµ
Rrel = ∆Rangewater / dmedium
Uncertainty of calibration curve 2% -3% ;
Range uncertainty 2-3 mm at 10 cm
µ(E,Z) = σ(E,Z) • ρe ( )( )⎪⎩
⎪⎨⎧
≥×±+<×±+= −
−
01012.035.4101008.041.91
4
4
HUHURrel
CT-numbers (HU)
Rel
ativ
e ra
nge
in w
ater
(Rre
l)
0.2
0.4
0.6
0.8
1.0
1.2
1.4
1.6
1.8
2.0
-1000 -500 0 500 1000 1500 2000
CT-numbers (HU)
Rel
. Ran
ge in
wat
er
TissuePlexiglasSubstitutesCalibration-1Calibration-2
Uncertainties of CT-numbers can be up to 300 HU for bone-like materials. This can lead to uncertainties in the range of ions (position of Bragg-peak)
Changes of the CT-numbers of bone-like materials affect the calibration curve and cause up to 6 mm change in the range of Carbon ions. Up to 80% of the dose is allocated to the critical organs.
What influences CT-numbers ?What influences CT-numbers ?Design parameters of the CT-scanner (filters, collimators, generation)
Voltage of the X-ray tube
Size of Field of view (FOV)
Reconstruction algorithms
Parameters of the scanned object (material of phantom, size, geometry, homogenity, position in the FOV, movement)
Artefacts, especially metal artefacts
The dependance of CT-numbers on the above parameters is only known emprically.
Aim ...Aim ...
Develop a method to calculate CT-numbers under different conditions.
Analyze the effects of measurement conditions on calculated CT-numbers.
Estimate range uncertainties due to different measurement conditions of CT-numbers.
Cylindrical phantom from PMMA (16 cm diameter) and Gammex-
RMI substitutes (Middletown, WI).
Siemens Emotion CT equipped with a DURA352 X-ray tube.
CT-numbers calculation chainCT-numbers calculation chain
Pulse-Height spectrum
Phase space files
PROJECTION(Raw-data
Format)Image (HU)
Photon transport simulation throughthe filters,collimators and phantom
in CT scanner using EGSnrc/BEAMnrc
Reconstruction
Measurements of CT spectrum
using SpectroX
X-ray-tube
simulation
Responsefunction
of CT detectorsMeasurements of
CT-numbers
Scheme of CT-scanner and phantom Scheme of CT-scanner and phantom
1) CT1-part was used to simulate X-ray source, filters and collimators. 1010
photons with a given energy spectrum. 1 week of CPU time.
2) CT2-part was used to simulate phantom and inserts. 5x108 photons. Input is PHSP from CT1. 1-3 hrs of CPU time.
CT-numbers calculation chainCT-numbers calculation chain
Pulse-Height spectrum
Phase space files
PROJECTION(Raw-data
Format)Image (HU)
Photon transport simulation throughthe filters,collimators and phantom
in CT scanner using EGSnrc/BEAMnrc
Reconstruction
Measurements of CT spectrum
using SpectroX
X-ray-tube
simulation
Responsefunction
of CT detectorsMeasurements of
CT-numbers
Projections and images
HU (Normalization ?)
Single projection Matrix – image
1
10
100
1000
1 65 129 193 257 321 385 449 513 577 641
Detector channels (672)
Flue
nce
per d
etec
tor c
hann
el
Fluence after phantom with Cortical bone insert
Initial fluence
Input CT-spectra calculated for the Emotion CT-scanner with DURA352 X-ray tube for 80, 110 and 130 kVInput CT-spectra calculated for the Emotion CT-scanner with DURA352 X-ray tube for 80, 110 and 130 kV
0.E+00
2.E-04
4.E-04
6.E-04
8.E-04
1.E-03
0 20 40 60 80 100 120 140Energy (KeV)
Cou
nts
(arb
itrar
y un
its)
80kV
110kV
130kV
0.E+00
1.E-02
2.E-02
3.E-02
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
energy (MeV)
fluen
ce/M
eV/in
ciden
t par
ticle
(cm
- 2 M
eV- 1
)
130 kV110 kV80 kV
0.00E+00
1.00E-03
2.00E-03
3.00E-03
4.00E-03
5.00E-03
6.00E-03
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
energy (MeV)
fluen
ce/M
eV/in
cide
nt p
artic
le (c
m- 2
MeV
- 1)
130 kV110 kV80 kV
Spectra in the center of FOV
Spectra at the edge of FOV
Input spectra
The effect of filters (Aluminium and Teflon) on the fluence and mean energy of the spectrum were investigated using 110 kV spectrum
The effect of filters (Aluminium and Teflon) on the fluence and mean energy of the spectrum were investigated using 110 kV spectrum
0.E+00
1.E-05
2.E-05
3.E-05
4.E-05
-50 -40 -30 -20 -10 0 10 20 30 40 50
distance from center of FOV (cm)
ener
gy fl
uenc
e/in
cide
nt p
artic
le(M
eV c
m-2
)
0 filters (Air)
1 filter (Al)
1 filter (Teflon)
2 filters (Al & Teflon)
0.0E+00
5.0E-03
1.0E-02
1.5E-02
2.0E-02
2.5E-02
0.02 0.04 0.06 0.08 0.1 0.12
energy (MeV)
fluen
ce/M
eV/in
ciden
t par
ticle
(c
m- 2
MeV
- 1)
0 filters (Air)1 filter (Al)1 filter (Teflon)2 filters (Al & Teflon)
0.0E+00
5.0E-03
1.0E-02
1.5E-02
2.0E-02
0.02 0.04 0.06 0.08 0.1 0.12energy (MeV)
fluen
ce/M
eV/in
ciden
t part
icle
(cm
- 2 M
eV- 1
)
0 filters (Air)1 filter (Al)1 filter (Teflon)2 filters (Al & Teflon)
55
60
65
70
-50 -40 -30 -20 -10 0 10 20 30 40 50Distance from center of FOV (cm)
Mea
n en
ergy
(keV
)
0 filters (Air)1 filter (Al)1 filter (Teflon)2 filters (Al and Teflon)
Simulated effectsSimulated effects
-1000
-500
0
500
1000
1500
2000
2500
70 80 90 100 110 120 130 140
Energy of the incident photons (kV)
Sim
ulat
ed C
T-nu
mbe
rs (H
U)
Air LN300LN450Adipose fatWaterIBCB2 50CB
-1000
-500
0
500
1000
1500
2000
10 15 20 25 30 35Phantom diameter (cm)
Sim
ulat
ed C
T-nu
mbe
rs (
HU
) LN300
LN450
Adipose fat
IB
CB2 50
CB
-1000
-500
0
500
1000
1500
2000
-0.2 0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4 1.6 1.8 2.0
Electron density of phantom material relative to water
Sim
ulat
ed C
T-nu
mbe
rs (H
U) LN300
LN450
Adipose fat
IB
CB2 30
CB
PMMA
Water, RW3 and MuscleAir CB
Voltage of X-ray tube Material of phantom
Diameter of phantom
Comparison of simulated and measured CT-numbersComparison of simulated and measured CT-numbers
-400
-200
0
200
400
-1000 -500 0 500 1000 1500 2000 2500Measured CT-numbers (HU)
Sim
ulat
ed -
mea
sure
d C
T-nu
mbe
rs
(HU
) Gammex 80 kV Gammex 130 kV
Effect of voltage, phantom size and phantom material on Carbon rangeEffect of voltage, phantom size and phantom material on Carbon range
0.00.20.40.60.81.01.21.41.61.82.0
-1000 -500 0 500 1000 1500 2000CT-numbers (HU)
Rel
ativ
e ra
nge
(WED
/T)
110 kV, PMMA, 16 cm
120 kV, PMMA, 16 cm
100 kV, PMMA, 16 cm
110 kV, PMMA, 30 cm
110 kV, Air, 16 cm
ConclusionsConclusions
It is POSSIBLE to calculate CT-numbers from MC simulation
The presented method was used to calculate the effects of different measurement parameters on CT-numbers and Carbon range
It is necessary to improve the effeciency of the MC simulation to use more than one projection for reconstruction (time needed foreach projection is 1-3 hrs of CPU time) to decrease the uncertainties of calculted CT-numbers
The simulation can not be used to calculate CT-numbers of irregular or off-center objects
Scheme of CT-scanner and phantom Scheme of CT-scanner and phantom
1) CT1-part was used to simulate X-ray source, filters and collimators. 1010
photons with a given energy spectrum. 1 week of CPU time.
2) CT2-part was used to simulate phantom and iserts. 5x108 photons. Input is PHSP from CT1. 1-3 hrs of CPU time.
Thank you for your attentionThank you for your attention
Questions ???Questions ???
Clinical Ion Therapy facility in Heidelberg (HIT)University Heidelberg / GSI / DKFZ1000 patients / yearp, He, C, O