> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
1
Optimisation of a Dual Head Semiconductor
Compton Camera using Geant4
L.J. Harknessa, T Beveridgec, A.J. Bostona, H.C. Bostona, R.J. Coopera, J.R. Cresswella, J.E. Gillamc,
A.N. Grinta, I. Lazarusb, P.J. Nolana, D.C. Oxleya, D.P. Scraggsa.
a
Department of Physics, University of Liverpool, Liverpool, L69 7ZE, UK
b
STFC Daresbury Laboratory, Warrington, WA4 4AD, UK
c
School of Physics and Materials Engineering, Monash University, Clayton, Victoria 3800, Australia
Abstract— Conventional gamma-camera systems utilise
mechanical collimation to provide information on the position of
an incident gamma-ray photon. Systems that use electronic
collimation utilising Compton Image reconstruction techniques
have the opportunity to offer huge improvements in detection
sensitivity. Such systems have been previously limited by the
relatively poor energy resolution of the detector material used in
the camera. The University of Liverpool Department of Physics
have been evaluating position sensitive High Purity germanium
(HPGe) detector systems as part of a Single Photon Emission
Computed Tomography (SPECT) gamma Compton Camera
system. Data has been acquired from the SmartPET detectors,
operated in Compton Camera mode. These orthogonally
segmented planar detectors are designed for the energy range of
small animal PET imaging [1]. The minimum in the energy
range of the current system is 244keV [2] due to the 20mm
thickness of the first scatter detector. This thickness of
germanium causes a large proportion of gammas with energy less
than 244keV to be completely absorbed in the detector, rather
than scatter through it. Results are presented on the outcome of
a validated Geant4 [3] simulation designed to optimise the
geometry of a new semiconductor Compton Camera system for
the energy range of medical applications.
II. PREVIOUS EXPERIMENTAL RESULTS
The two ORTEC position sensitive High Purity Germanium
segmented SmartPET detectors have been operated in a dual
head Compton Camera configuration to image a 152Eu point
source [2]. The separation between the scatter detector and
analyser detector was varied at 3cm, 5cm, 7cm, 9cm and 11cm
whilst the source was rotated from 0° to 15°, 30°, 45° and 60°
for each separation, totaling 25 acquisition positions. Simple
reconstruction codes were implemented to obtain images of
the source.
Index Terms— HPGe Planar, SPECT, Compton camera,
Geant4.
I.
INTRODUCTION
Position sensitive High Purity germanium detectors are the
sensor of choice for gamma-ray detection in the field of
nuclear physics due to their excellent intrinsic energy
resolution and the potential for excellent spatial resolution
using segmentation and pulse shape analysis techniques [1].
Members of the University of Liverpool imaging group are
currently developing a Compton Camera for medical and
security applications, which will exploit the characteristics of
high purity germanium detectors. Previous work done by the
group has shown the potential of semiconductor Compton
camera systems but show limitations to the current imaging
equipment [2]. Such limitations arise because SmartPET [3]
has been designed and developed for Positron Emission
Tomography (PET) experiments, at a higher energy range than
the Compton Camera for Single Positron Emission
Tomography (SPECT).
Fig 1: Reconstructed images of 152Eu point source for 3cm (top) and 5cm
(bottom) separation at 0° (left) and 60° (right)
Fig 2: The SmartPET detectors in Compton Camera Configuration with
source at 0° [2]
The majority of incident gammas of energy <244keV were
absorbed in the 20mm thick scatter detector, showing that the
detectors are not ideal for the energy of medical interest,
141keV. In addition, experiments have been carried out using
a 0.5mm thick Double Sided Strip Silicon Detector (Grint,
2008) as the scatter detector and a SmartPET detector as the
analyser detector. It was observed in this acquisition that
limitations arose from high noise levels on the DSSSD,
leading to a reduced number of detected events. For medical
applications this is not desirable, as dose to the patient should
be kept to a minimum through utilizing high sensitivity
instruments. It is proposed that using germanium as an
> REPLACE THIS LINE WITH YOUR PAPER IDENTIFICATION NUMBER (DOUBLE-CLICK HERE TO EDIT) <
alternative to silicon could improve on this, as noise levels
could be significantly lower than on the DSSSD.
III. GEANT4
Geant4 (Collaboration, 2007) has been used to simulate the
SmartPET detectors in Compton camera configuration, Fig. 3,
and has been successfully validated against experimental data.
2
IV. OPTIMISATION
In order to ascertain the optimal Compton camera
configuration, various parameters can be modified:




Scatterer thickness and material
Absorber thickness and material
Separation between two detectors
Angle of absorber aligned relative to scatterer
V. FURTHER STUDY
In addition to optimizing the fraction of events which depsosit
energy in a single Compton scatter event in the front detector
and single photoelectric absorption in the back detector, it is
important to examine the properties of the reconstructed
images produced from the simulated data. In particular, it will
be of interest to add the contributions of detector energy
resolution and Doppler broadening effects in addition to the
effects of finite position resolution. Properties of the image to
be examined include the resolution of a point source and the
fraction of useful events in the peak. A thorough evaluation of
the contribution of the reconstruction method to the image
resolution will also be carried out.
Fig 3: Geant 4 Simulation output of 2 SmartPet detectors
(blue), guard ring (red) and aluminium can (yellow).
Isotropic source Gamma rays are shown in green.
The simulation is currently being used to optimise the
geometry of a semiconductor Compton camera system, by
varying the detector thicknesses and geometry. The setup will
be fully optimised for 141keV but will be useful across an
energy range to be determined by the simulations. The
optimal performance will be maximum single hit scattering in
the front detector and maximum absorption in the back
detector. It can be seen in Fig.4 that the majority of the
incident 141keV radiation is absorbed in the thickness of the
SmartPET detector, and that by decreasing the thickness, a
higher proportion will scatter out of the detector.
VI. CONCLUSION
A semiconductor Compton camera is being designed and
optimised using Geant4 simulations. A proof-of-principle has
been demonstrated with the SmartPET system for energies
>344keV range. The aim is to increase sensitivity of current
SPECT systems by implementing a Compton camera system
and removing the need for mechanical collimation. The
system will be fully optimised for 141keV but will be
simulated across a wide energy range. Applications of interest
include nuclear medical imaging and security.
ACKNOWLEDGMENT
The author would like to thank J Gillam for the use of his
SmartPET Compton Camera images.
REFERENCES
[1]D. Bazzacco, Nuclear Physics A 746 (2004)
[2] H.C. Boston et al., IEEE Nuclear Science Symposium
Conference Record (2006)
[3] R.J. Cooper et al., Nuclear Instruments and Methods in
Physics Research A 579 (2007)
[4] A. Grint, IOP Nuclear Physics Conference (2008)
[5] S. Agostinelli Nuclear Instruments and Methods in Physics
Research A 506 (2003) 250–303
Fig 4: Simulated interaction position in x across the scatter
detector for incident energy 141keV