Transcript Spectral measurements of a CT x-ray tube

Monte Carlo model and spectral
measurements of a computed
tomography x-ray tube
Magdalena Bazalova
Frank Verhaegen
Overview
•
•
•
•
Motivation
Diagnostic x-ray beam production
Monte Carlo (MC) simulations
Spectral measurement
–
–
–
in computed tomography (CT) geometry
experimental setup and Compton scatter
detector
• Corrections of measured spectra
• Results: Comparison of MC simulated spectra
with measured spectra
• Conclusions
Motivation
• It is very difficult to assess patient dose from CT
imaging and mostly approximative methods are
used.
• A Monte Carlo model of the x-ray tube of a CT
scanner in combination with patient CT images
will allow for an accurate estimate of CT dose.
• However, the MC model has to be accurate in
order to calculate the dose accurately.
• To developed and validate (using spectral
measurements) a MC model of a Philips PQ5000
CT simulator x-ray tube.
X-ray beam production
tungsten target
Monte Carlo model of the PQ5000
x-ray tube
• using the BEAM code and manufacturer’s
specifications of the x-ray tube
•
•
•
•
anode material: W
anode angle: 7°
tube voltage: 100-140 kVp
additional filter: 9 mm Al
Spectral measurements of
diagnostic x-rays
• Not trivial due to the high photon flux
causing pile-up in the detector of choice.
• Can be solved by using very small
collimators and/or increasing the x-ray
source to detector distance; the alignment
is then detrimental.
• Not applicable for CT geometry and our
detector, therefore, the x-ray flux has to be
reduced in another way.
Problems of spectral measurements
in CT geometry
a few meters
80 cm
• Attenuation?
• Scatter!
Experimental setup using Compton
scatter
Compton scatter- scatterer material
• interaction of a photon of energy hν with a
loosely bound orbital electron of an
absorber
Scatterer material
Solid line: Z=6
Dashed line: Z=50
Measurements of x-ray spectra:
CdTe detector
• 3×3×1 mm3 CdTe crystal (ZCd=48, ZTe=52)
• Pt and In contacts
Spectrum corrections-detector
response
2 mm
1 mm
3 mm
3 mm
Spectrum corrections-Compton
effect
Spectrum corrections-peak
broadening
Spectrum corrections - final
filtering
Comparison of MC simulations
and spectral measurements
kVp
Half value layer comparison
Conclusions
• An accurate Monte Carlo model of the PQ5000
CT simulator x-ray was developed.
• The model was validated with spectral
measurements using a CdTe detector in a
Compton scattering setup.
• A very good agreement between the MC
simulated and measured spectra for the extreme
tube voltages with and without an additional 9
mm thick aluminum filter was found.
• The MC model can be used for CT dose
calculation.
Acknowledgements
• Robin van Gils
• Wamied Abdel-Rahman
• Derek Liu
• Slobodan Devic
• Jan Seuntjens
• Joe Larkin
• Medical Physics Unit staff and students
• NSERC