Logo Logo
Help
Contact
Switch language to German
Development and performance evaluation of detectors in a Compton camera arrangement for ion beam range monitoring in particle therapy
Development and performance evaluation of detectors in a Compton camera arrangement for ion beam range monitoring in particle therapy
The growing interest in particle beam therapy for cancer treatment is driven by the ability to provide high precision dose delivery. However, this benefit demands a high accuracy on the determination of the well-localized dose deposition (Bragg peak), which has to be located within the tumor volume. Different approaches for the beam range monitoring are worldwide being evaluated. The Compton camera is one of the proposed techniques, which aims at providing real-time, in-vivo proton (or ion) beam range monitoring by means of the detection of secondary prompt gamma rays, resulting from nuclear reactions between the particle beam and the biological tissue. The purpose of our project is to develop and commission an imaging system based on a Compton camera detector arrangement which could monitor in (ultimately) real-time the ion beam range. In the context of this thesis a Compton camera detector prototype was characterized, consolidated and commissioned with both a multi-layer and a mono-layer scatter component. The first detector arrangement belongs to the LMU Compton camera: the detector components were extensively characterized in order to determine the limitations imposed by their internal structure and the required configuration for an optimum performance. The complexity of the signal readout and processing could be reduced in view of facilitating an envisaged clinical applicability of the system. The scatter component (tracker) is formed by a stack of six highly segmented double-sided Si-strip detectors, whereas a monolithic LaBr3(Ce) scintillator (5 x 5 x 3 cm3) acts as the absorber component and is coupled to a segmented position-sensitive multi-anode photomultiplier tube (PMT). The initially applied 256-fold segmented PMT was replaced by a 64-fold segmented PMT, and similar or even superior performance was demonstrated for the latter one. The same trend of an improving spatial resolution, with an increasing energy of the incoming photon, which was observed when using the 256-fold segmented PMT, was also preserved: at 137Cs energy a value of 3.4(1) mm was obtained, while at the 1173 keV and 1332 keV 60Co photopeaks the spatial resolution reached values of 2.9(1) mm, thus below the 3 mm absorber resolution envisaged by the Compton camera design. Moreover, first tests in view of a possible replacement of the LaBr3(Ce) scintillation material with the cost-effective and radio-pure CeBr3 scintillator material were pursued and seem promising (ΔE/E ≃ 4% at 662 keV and comparable timing properties as LaBr3(Ce)). The signal processing and data readout system for the scatter component was upgraded from an ASIC-based electronics to a more flexible and higher performing electronics based on discrete components. Full compliance of the new frontend electronics with the detector signal specifications of our camera prototype was achieved: an acceptance of both signal polarities was introduced as well as a trigger capability for the scatter component, which previously did not exist. Furthermore, the upgrade of the signal processing and data acquisition was extended to the whole Compton camera setup, adapting the new frontend electronics designed initially for replacing the outdated ASIC-based modules of the scatterer also to the signal properties of the absorber scintillator and its segmented readout. This allowed for reducing the complexity of the system and finally achieve a 1 Mcps count rate capability as required in a clinical scenario: the VME-based readout modules were implemented into the new DAQ software and the data streams of scatterer and absorber were merged. The reduced granularity of the PMT signal channels combined with the use of the new signal processing and data acquisition system based on optical fibers makes the Compton camera setup less complex and more flexible. All detectors can be mounted in a newly designed Faraday cage, which includes also an active cooling, capable of reducing the dark current in the silicon detectors. The upgraded system was tested in the laboratory as well as under online conditions with particle beams at the Tandem accelerator in Garching. A validation with high energy prompt-γ rays was performed, bombarding water and PMMA targets with a 20 MeV proton beam and the same detector performance could be demonstrated also with the new signal processing system. The achievable trigger rate was increased by one order of magnitude and due to the efficient selection of Compton scattered events by triggering on the scatter component, the ratio of registered Compton events could be increased by about three orders of magnitude compared to the previous data acquisition system. The camera system was also tested by hitting a water target with a pulsed deuteron beam in order to allow for assessing the timing performance. With an envisaged improved version of the internal implantation structure of the segmented silicon scatter modules, the multi-layer Compton camera system will be ready for a full performance characterization of the imaging system's capabilities. By using the high performing LaBr3(Ce) monolithic scintillator as absorber, a Compton camera setup was also arranged with a mono-layer scattering component consisting of a pixelated 22 x 22 array of GAGG scintillator crystals. A proof of principle study was carried out using 137Cs and 60Co calibration sources: the source position reconstruction was performed with the MEGAlib software and the resulting reconstructed images from experimental data were compared to images reconstructed from simulated data. A source shift of 2 mm could be resolved by the system with sub-millimeter accuracy. A trend of improving angular resolution with the incoming photon energy reflects the detectors' (energy and spatial resolution) performance improvements with increasing energy. The system was characterized in different geometrical configurations, in order to address not only a possible prompt-gamma imaging application, but also a multi-modality detector system able to be applied also in PET- or gamma-PET-like imaging scenarios.
particle therapy, Compton camera, range verification, detector development, medical physics, semiconductors, scintillators
Liprandi, Silvia
2018
English
Universitätsbibliothek der Ludwig-Maximilians-Universität München
Liprandi, Silvia (2018): Development and performance evaluation of detectors in a Compton camera arrangement for ion beam range monitoring in particle therapy. Dissertation, LMU München: Faculty of Physics
[img] PDF
Liprandi_Silvia.pdf

72MB

Abstract

The growing interest in particle beam therapy for cancer treatment is driven by the ability to provide high precision dose delivery. However, this benefit demands a high accuracy on the determination of the well-localized dose deposition (Bragg peak), which has to be located within the tumor volume. Different approaches for the beam range monitoring are worldwide being evaluated. The Compton camera is one of the proposed techniques, which aims at providing real-time, in-vivo proton (or ion) beam range monitoring by means of the detection of secondary prompt gamma rays, resulting from nuclear reactions between the particle beam and the biological tissue. The purpose of our project is to develop and commission an imaging system based on a Compton camera detector arrangement which could monitor in (ultimately) real-time the ion beam range. In the context of this thesis a Compton camera detector prototype was characterized, consolidated and commissioned with both a multi-layer and a mono-layer scatter component. The first detector arrangement belongs to the LMU Compton camera: the detector components were extensively characterized in order to determine the limitations imposed by their internal structure and the required configuration for an optimum performance. The complexity of the signal readout and processing could be reduced in view of facilitating an envisaged clinical applicability of the system. The scatter component (tracker) is formed by a stack of six highly segmented double-sided Si-strip detectors, whereas a monolithic LaBr3(Ce) scintillator (5 x 5 x 3 cm3) acts as the absorber component and is coupled to a segmented position-sensitive multi-anode photomultiplier tube (PMT). The initially applied 256-fold segmented PMT was replaced by a 64-fold segmented PMT, and similar or even superior performance was demonstrated for the latter one. The same trend of an improving spatial resolution, with an increasing energy of the incoming photon, which was observed when using the 256-fold segmented PMT, was also preserved: at 137Cs energy a value of 3.4(1) mm was obtained, while at the 1173 keV and 1332 keV 60Co photopeaks the spatial resolution reached values of 2.9(1) mm, thus below the 3 mm absorber resolution envisaged by the Compton camera design. Moreover, first tests in view of a possible replacement of the LaBr3(Ce) scintillation material with the cost-effective and radio-pure CeBr3 scintillator material were pursued and seem promising (ΔE/E ≃ 4% at 662 keV and comparable timing properties as LaBr3(Ce)). The signal processing and data readout system for the scatter component was upgraded from an ASIC-based electronics to a more flexible and higher performing electronics based on discrete components. Full compliance of the new frontend electronics with the detector signal specifications of our camera prototype was achieved: an acceptance of both signal polarities was introduced as well as a trigger capability for the scatter component, which previously did not exist. Furthermore, the upgrade of the signal processing and data acquisition was extended to the whole Compton camera setup, adapting the new frontend electronics designed initially for replacing the outdated ASIC-based modules of the scatterer also to the signal properties of the absorber scintillator and its segmented readout. This allowed for reducing the complexity of the system and finally achieve a 1 Mcps count rate capability as required in a clinical scenario: the VME-based readout modules were implemented into the new DAQ software and the data streams of scatterer and absorber were merged. The reduced granularity of the PMT signal channels combined with the use of the new signal processing and data acquisition system based on optical fibers makes the Compton camera setup less complex and more flexible. All detectors can be mounted in a newly designed Faraday cage, which includes also an active cooling, capable of reducing the dark current in the silicon detectors. The upgraded system was tested in the laboratory as well as under online conditions with particle beams at the Tandem accelerator in Garching. A validation with high energy prompt-γ rays was performed, bombarding water and PMMA targets with a 20 MeV proton beam and the same detector performance could be demonstrated also with the new signal processing system. The achievable trigger rate was increased by one order of magnitude and due to the efficient selection of Compton scattered events by triggering on the scatter component, the ratio of registered Compton events could be increased by about three orders of magnitude compared to the previous data acquisition system. The camera system was also tested by hitting a water target with a pulsed deuteron beam in order to allow for assessing the timing performance. With an envisaged improved version of the internal implantation structure of the segmented silicon scatter modules, the multi-layer Compton camera system will be ready for a full performance characterization of the imaging system's capabilities. By using the high performing LaBr3(Ce) monolithic scintillator as absorber, a Compton camera setup was also arranged with a mono-layer scattering component consisting of a pixelated 22 x 22 array of GAGG scintillator crystals. A proof of principle study was carried out using 137Cs and 60Co calibration sources: the source position reconstruction was performed with the MEGAlib software and the resulting reconstructed images from experimental data were compared to images reconstructed from simulated data. A source shift of 2 mm could be resolved by the system with sub-millimeter accuracy. A trend of improving angular resolution with the incoming photon energy reflects the detectors' (energy and spatial resolution) performance improvements with increasing energy. The system was characterized in different geometrical configurations, in order to address not only a possible prompt-gamma imaging application, but also a multi-modality detector system able to be applied also in PET- or gamma-PET-like imaging scenarios.