- Email: [email protected]
A time-of-flight detector for thermal neutrons from radiotherapy Linacs
ARTICLE IN PRESS
Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90 www.elsevier.com/locate/nima
A time-of-flight detector for thermal neutrons from radiotherapy Linacs V. Contia,, G. Bartesaghia, D. Bologninib, V. Mascagnab, C. Perbonib, M. Prestb, S. Scazzib, A. Mozzanicac, P. Cappellettid, M. Frigeriod, S. Gelosad, A. Montid, A. Ostinellid, G. Gianninie, E. Vallazzae a
Universita` degli Studi di Milano and INFN di Milano, Italy Universita` dell’Insubria, Como and INFN di Milano, Italy c Universita` degli Studi di Brescia and INFN sezione di Pavia, Italy d Fisica Sanitaria, Ospedale S. Anna di Como, Italy e INFN, sezione di Trieste and Universita` degli Studi di Trieste, Italy b
Available online 7 August 2007
Abstract Boron Neutron Capture Therapy (BNCT) is a therapeutic technique exploiting the release of dose inside the tumour cell after a fission of a 10B nucleus following the capture of a thermal neutron. BNCT could be the treatment for extended tumors (liver, stomach, lung), radio-resistant ones (melanoma) or tumours surrounded by vital organs (brain). The application of BNCT requires a high thermal neutron flux ð45 108 n cm2 s1 Þ with the correct energy spectrum (neutron energy o10 keV), two requirements that for the moment are fulfilled only by nuclear reactors. The INFN PhoNeS (Photo Neutron Source) project is trying to produce such a neutron beam with standard radiotherapy Linacs, maximizing with a dedicated photo-neutron converter the neutrons produced by Giant Dipole Resonance by a high energy (48 MeV) photon beam. In this framework, we have developed a real-time detector to measure the thermal neutron time-of -flight to compute the flux and the energy spectrum. Given the pulsed nature of Linac beams, the detector is a single neutron counting system made of a scintillator detecting the photon emitted after the neutron capture by the hydrogen nuclei. The scintillator signal is sampled by a dedicated FPGA clock thus obtaining the exact arrival time of the neutron itself. The paper will present the detector and its electronics, the feasibility measurements with a Varian Clinac 1800/2100CD and comparison with a Monte Carlo simulation. r 2007 Published by Elsevier B.V.
1. Introduction Four years after the discovery of neutrons in 1932 by J. Chadwick, a biophysicist (G.L. Locker) introduced the concept of Boron Neutron Capture Therapy, that is the possibility of targeting a tumour with a carrier of elemental 10 B and of irradiating it with a thermal neutron beam thus producing heavy charged particles (an a particle and a nucleus of 7Li) releasing their whole energy in the cell [1]. Up to now, BNCT has been performed on a small number of patients because of biochemical problems (the specificity of the boron compounds with respect to the tumour cells) and of the neutron fluxes ð45 108 n cm2 s1 Þ in Corresponding author.
E-mail address: [email protected] (V. Conti). 0168-9002/$ - see front matter r 2007 Published by Elsevier B.V. doi:10.1016/j.nima.2007.07.145
the thermal region) that require the use of nuclear reactors. The INFN PhoNeS (Photo Neutron Source) project [2] is developing a converter þ moderator system to be applied to the standard radiotherapy Linacs to exploit the neutrons photoproduced by high energy photon beams in order to perform BNCT in hospital environments. This work describes the development in the PhoNeS framework of a real time dosimeter and spectrometer based on a standard plastic scintillator readout by a dedicated electronics (Section 2) and the first measurements performed with the Clinac 1800 and 2100 at the Radiotherapy Department of the S. Anna Hospital (Como, Italy; Section 3). The measurements are compared with the Monte Carlo simulation and with the results obtained with other non real-time neutron detectors (Section 3).
ARTICLE IN PRESS V. Conti et al. / Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90
2. The detector and its electronics Two different detectors have been assembled: a homemade plastic scintillator (Fig. 1a), readout by two P30CW05 photomultipliers by Electron Tubes used in coincidence, and a 1% in weight 10B doped BC-454 by Bicron readout by a single PM of the same type. In both cases the readout electronics (Fig. 1b) samples with a highfrequency clock (12.5 MHz), generated by an Altera Cyclone II FPGA, the scintillator discriminated signal, thus sampling the arrival time of each neutron, detecting in the first case the 2:2 MeVg emitted in the reaction H(n,g)D and in the second one also the energy released by the reaction with the boron element induced by the neutron capture. In this way the detector can work as a dosimeter since it detects all the photons that interact with the scintillator and that are the result of the neutron capture by the hydrogen atoms of the scintillator itself and of all the materials surrounding the detector (as in the patient case),
89
and as a spectrometer since from the time-of-flight spectrum it is possible to extract the energy one with the help of the simulation. The scintillator detector exploits the features of a Linac beam, that is its pulsed nature (particles are emitted in bunches with a rate of 150–300 Hz depending on the particle type and its energy) that excludes direct photons or electrons in the time interval between bunches; on the other hand neutrons in the thermal energy range (a 0.025 eV neutron has a speed of 2:2 mm=ms) arrive in the inter bunch period. Neutron doses, fluxes and spectra are usually measured with integrating detectors such as thermoluminescent ones, which feature a complicated procedure for the calibration and readout, or by bubble detectors, which are limited in the maximum dose they can handle and are not real-time at all. Plastic scintillators are low-cost and flexible detectors, historically used for fast neutrons but that can be employed also slow ones in this particular beam configuration.
Fig. 1. (a) The home-made plastic scintillator coupled with the two Electron Tubes photomultipliers operated in coincidence. (b) Photo of the readout board featuring the Altera Cyclone II FPGA which generates the sampling clock and stores the data.
a
Fig. 2. (a) Measurement setup in the S. Anna Hospital with the PhoNeS prototype positioned in front of the accelerator head. (b) Comparison of the neutron fluxes obtained by the simulation, by the activation method and by means of the real-time detector. The data have been rescaled to the first point.
ARTICLE IN PRESS 90
V. Conti et al. / Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90
Fig. 3. (a) Measured TOF spectrum of the second detector compared with the simulation one. (b) Energy spectrum corresponding to the Monte Carlo TOF presented in (a).
3. Measurements and results
4. Conclusions
The detectors have been evaluated both as dosimeters and as spectrometers with and as without the PhoNeS converter in place. All the results have been compared with a simulation developed with the MCNP4B-GN code [3]. In the case of the dose measurement, the neutron flux has been compared also with the one obtained activating an Al sample in the same experimental conditions. Fig. 2a shows the measurement setup in the Radiotherapy unit. Fig. 2b presents the results in terms of neutron fluxes for the simulation, the activation method and the real-time dosimeter as a function of the thickness of a PMMA1 absorber. The data have been rescaled to the first point. Given the fact that the scintillator acts as a dosimeter, the ‘‘flux’’ it measures gets higher with the PMMA thickness, since the detector collects also the photons generated by the neutrons stopping in the PMMA. As far as the energy spectrum is concerned, Fig. 3a shows the time-of-flight spectrum of the second detector compared with the simulation one. The fit has been performed with two exponentials to take into account two different contributions to the neutron thermalization process: the slow one, which is due to the neutrons that are already in the thermal range when they arrive on the scintillator and the faster one, which is due to the neutrons in the high energy part of the thermal spectrum and which is strictly dependent on the materials surrounding the detector. Once the real and simulated arrival time profiles agree, it is possible to obtain from the simulation a high resolution neutron beam energy spectrum (Fig. 3b). The same results have been obtained with the first prototype.
A real-time scintillator-based dosimeter and spectrometer has been developed for the neutron beam produced by a standard Linac accelerator in the framework of the INFN PhoNeS project and the PRIN05 project (MIUR, Italy). The detector behaviour has been studied with a Clinac 1800 and a Clinac 2100 and compared with the measurements of the activation of an Al sample and with a dedicated simulation. Data taking sessions are foreseen both at reactors and at different types of medical Linacs. While on one hand, the fluxes reached with the PhoNeS converter are still not enough to perform BNCT, a realtime active system such as the one described in this work will allow a fast evaluation of the performance of different converter/moderator assemblies in terms of energy spectrum and of dose. The measurement of the neutron flux only will be performed with a new prototype based on boron doped scintillating fibers readout by the same electronics or on very thin scintillators which will allow to disentangle the contribution of photons produced by neutron capture in the surrounding materials.
1
PolyMethylMetAcrylate, used for of its tissue equivalence.
References [1] J.R. Cunningam, H.E. Johns, Physics of Radiology, fourth Ed., Charles C. Thomas, Springfield, IL, 1938. [2] R. Bevilacqua, et al., Nucl. Instr. and Meth. A 572 (2007) 231. [3] J.F. Briesmeister (Ed.), MCNP-A General Monte Carlo N-Particle Transport Code, Version 4B, 1997.
Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90 www.elsevier.com/locate/nima
A time-of-flight detector for thermal neutrons from radiotherapy Linacs V. Contia,, G. Bartesaghia, D. Bologninib, V. Mascagnab, C. Perbonib, M. Prestb, S. Scazzib, A. Mozzanicac, P. Cappellettid, M. Frigeriod, S. Gelosad, A. Montid, A. Ostinellid, G. Gianninie, E. Vallazzae a
Universita` degli Studi di Milano and INFN di Milano, Italy Universita` dell’Insubria, Como and INFN di Milano, Italy c Universita` degli Studi di Brescia and INFN sezione di Pavia, Italy d Fisica Sanitaria, Ospedale S. Anna di Como, Italy e INFN, sezione di Trieste and Universita` degli Studi di Trieste, Italy b
Available online 7 August 2007
Abstract Boron Neutron Capture Therapy (BNCT) is a therapeutic technique exploiting the release of dose inside the tumour cell after a fission of a 10B nucleus following the capture of a thermal neutron. BNCT could be the treatment for extended tumors (liver, stomach, lung), radio-resistant ones (melanoma) or tumours surrounded by vital organs (brain). The application of BNCT requires a high thermal neutron flux ð45 108 n cm2 s1 Þ with the correct energy spectrum (neutron energy o10 keV), two requirements that for the moment are fulfilled only by nuclear reactors. The INFN PhoNeS (Photo Neutron Source) project is trying to produce such a neutron beam with standard radiotherapy Linacs, maximizing with a dedicated photo-neutron converter the neutrons produced by Giant Dipole Resonance by a high energy (48 MeV) photon beam. In this framework, we have developed a real-time detector to measure the thermal neutron time-of -flight to compute the flux and the energy spectrum. Given the pulsed nature of Linac beams, the detector is a single neutron counting system made of a scintillator detecting the photon emitted after the neutron capture by the hydrogen nuclei. The scintillator signal is sampled by a dedicated FPGA clock thus obtaining the exact arrival time of the neutron itself. The paper will present the detector and its electronics, the feasibility measurements with a Varian Clinac 1800/2100CD and comparison with a Monte Carlo simulation. r 2007 Published by Elsevier B.V.
1. Introduction Four years after the discovery of neutrons in 1932 by J. Chadwick, a biophysicist (G.L. Locker) introduced the concept of Boron Neutron Capture Therapy, that is the possibility of targeting a tumour with a carrier of elemental 10 B and of irradiating it with a thermal neutron beam thus producing heavy charged particles (an a particle and a nucleus of 7Li) releasing their whole energy in the cell [1]. Up to now, BNCT has been performed on a small number of patients because of biochemical problems (the specificity of the boron compounds with respect to the tumour cells) and of the neutron fluxes ð45 108 n cm2 s1 Þ in Corresponding author.
E-mail address: [email protected] (V. Conti). 0168-9002/$ - see front matter r 2007 Published by Elsevier B.V. doi:10.1016/j.nima.2007.07.145
the thermal region) that require the use of nuclear reactors. The INFN PhoNeS (Photo Neutron Source) project [2] is developing a converter þ moderator system to be applied to the standard radiotherapy Linacs to exploit the neutrons photoproduced by high energy photon beams in order to perform BNCT in hospital environments. This work describes the development in the PhoNeS framework of a real time dosimeter and spectrometer based on a standard plastic scintillator readout by a dedicated electronics (Section 2) and the first measurements performed with the Clinac 1800 and 2100 at the Radiotherapy Department of the S. Anna Hospital (Como, Italy; Section 3). The measurements are compared with the Monte Carlo simulation and with the results obtained with other non real-time neutron detectors (Section 3).
ARTICLE IN PRESS V. Conti et al. / Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90
2. The detector and its electronics Two different detectors have been assembled: a homemade plastic scintillator (Fig. 1a), readout by two P30CW05 photomultipliers by Electron Tubes used in coincidence, and a 1% in weight 10B doped BC-454 by Bicron readout by a single PM of the same type. In both cases the readout electronics (Fig. 1b) samples with a highfrequency clock (12.5 MHz), generated by an Altera Cyclone II FPGA, the scintillator discriminated signal, thus sampling the arrival time of each neutron, detecting in the first case the 2:2 MeVg emitted in the reaction H(n,g)D and in the second one also the energy released by the reaction with the boron element induced by the neutron capture. In this way the detector can work as a dosimeter since it detects all the photons that interact with the scintillator and that are the result of the neutron capture by the hydrogen atoms of the scintillator itself and of all the materials surrounding the detector (as in the patient case),
89
and as a spectrometer since from the time-of-flight spectrum it is possible to extract the energy one with the help of the simulation. The scintillator detector exploits the features of a Linac beam, that is its pulsed nature (particles are emitted in bunches with a rate of 150–300 Hz depending on the particle type and its energy) that excludes direct photons or electrons in the time interval between bunches; on the other hand neutrons in the thermal energy range (a 0.025 eV neutron has a speed of 2:2 mm=ms) arrive in the inter bunch period. Neutron doses, fluxes and spectra are usually measured with integrating detectors such as thermoluminescent ones, which feature a complicated procedure for the calibration and readout, or by bubble detectors, which are limited in the maximum dose they can handle and are not real-time at all. Plastic scintillators are low-cost and flexible detectors, historically used for fast neutrons but that can be employed also slow ones in this particular beam configuration.
Fig. 1. (a) The home-made plastic scintillator coupled with the two Electron Tubes photomultipliers operated in coincidence. (b) Photo of the readout board featuring the Altera Cyclone II FPGA which generates the sampling clock and stores the data.
a
Fig. 2. (a) Measurement setup in the S. Anna Hospital with the PhoNeS prototype positioned in front of the accelerator head. (b) Comparison of the neutron fluxes obtained by the simulation, by the activation method and by means of the real-time detector. The data have been rescaled to the first point.
ARTICLE IN PRESS 90
V. Conti et al. / Nuclear Instruments and Methods in Physics Research A 581 (2007) 88–90
Fig. 3. (a) Measured TOF spectrum of the second detector compared with the simulation one. (b) Energy spectrum corresponding to the Monte Carlo TOF presented in (a).
3. Measurements and results
4. Conclusions
The detectors have been evaluated both as dosimeters and as spectrometers with and as without the PhoNeS converter in place. All the results have been compared with a simulation developed with the MCNP4B-GN code [3]. In the case of the dose measurement, the neutron flux has been compared also with the one obtained activating an Al sample in the same experimental conditions. Fig. 2a shows the measurement setup in the Radiotherapy unit. Fig. 2b presents the results in terms of neutron fluxes for the simulation, the activation method and the real-time dosimeter as a function of the thickness of a PMMA1 absorber. The data have been rescaled to the first point. Given the fact that the scintillator acts as a dosimeter, the ‘‘flux’’ it measures gets higher with the PMMA thickness, since the detector collects also the photons generated by the neutrons stopping in the PMMA. As far as the energy spectrum is concerned, Fig. 3a shows the time-of-flight spectrum of the second detector compared with the simulation one. The fit has been performed with two exponentials to take into account two different contributions to the neutron thermalization process: the slow one, which is due to the neutrons that are already in the thermal range when they arrive on the scintillator and the faster one, which is due to the neutrons in the high energy part of the thermal spectrum and which is strictly dependent on the materials surrounding the detector. Once the real and simulated arrival time profiles agree, it is possible to obtain from the simulation a high resolution neutron beam energy spectrum (Fig. 3b). The same results have been obtained with the first prototype.
A real-time scintillator-based dosimeter and spectrometer has been developed for the neutron beam produced by a standard Linac accelerator in the framework of the INFN PhoNeS project and the PRIN05 project (MIUR, Italy). The detector behaviour has been studied with a Clinac 1800 and a Clinac 2100 and compared with the measurements of the activation of an Al sample and with a dedicated simulation. Data taking sessions are foreseen both at reactors and at different types of medical Linacs. While on one hand, the fluxes reached with the PhoNeS converter are still not enough to perform BNCT, a realtime active system such as the one described in this work will allow a fast evaluation of the performance of different converter/moderator assemblies in terms of energy spectrum and of dose. The measurement of the neutron flux only will be performed with a new prototype based on boron doped scintillating fibers readout by the same electronics or on very thin scintillators which will allow to disentangle the contribution of photons produced by neutron capture in the surrounding materials.
1
PolyMethylMetAcrylate, used for of its tissue equivalence.
References [1] J.R. Cunningam, H.E. Johns, Physics of Radiology, fourth Ed., Charles C. Thomas, Springfield, IL, 1938. [2] R. Bevilacqua, et al., Nucl. Instr. and Meth. A 572 (2007) 231. [3] J.F. Briesmeister (Ed.), MCNP-A General Monte Carlo N-Particle Transport Code, Version 4B, 1997.