WO2016110141A1 - 一种闪烁脉冲的数字化方法 - Google Patents

一种闪烁脉冲的数字化方法 Download PDF

Info

Publication number
WO2016110141A1
WO2016110141A1 PCT/CN2015/092909 CN2015092909W WO2016110141A1 WO 2016110141 A1 WO2016110141 A1 WO 2016110141A1 CN 2015092909 W CN2015092909 W CN 2015092909W WO 2016110141 A1 WO2016110141 A1 WO 2016110141A1
Authority
WO
WIPO (PCT)
Prior art keywords
pulse
energy
scintillation
information
digitizing
Prior art date
Application number
PCT/CN2015/092909
Other languages
English (en)
French (fr)
Inventor
谢庆国
张求德
龙岸文
熊章靖
Original Assignee
苏州瑞派宁科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 苏州瑞派宁科技有限公司 filed Critical 苏州瑞派宁科技有限公司
Priority to EP15876642.8A priority Critical patent/EP3244235B1/en
Priority to US15/541,220 priority patent/US9910167B2/en
Priority to JP2017535397A priority patent/JP6473509B2/ja
Publication of WO2016110141A1 publication Critical patent/WO2016110141A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16—Measuring radiation intensity
    • G01T1/20—Measuring radiation intensity with scintillation detectors
    • G01T1/208—Circuits specially adapted for scintillation detectors, e.g. for the photo-multiplier section
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16—Measuring radiation intensity
    • G01T1/17—Circuit arrangements not adapted to a particular type of detector
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01T—MEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00—Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16—Measuring radiation intensity
    • G01T1/20—Measuring radiation intensity with scintillation detectors
    • G—PHYSICS
    • G04—HOROLOGY
    • G04F—TIME-INTERVAL MEASURING
    • G04F10/00—Apparatus for measuring unknown time intervals by electric means
    • G04F10/005—Time-to-digital converters [TDC]
    • H—ELECTRICITY
    • H03—ELECTRONIC CIRCUITRY
    • H03M—CODING; DECODING; CODE CONVERSION IN GENERAL
    • H03M1/00—Analogue/digital conversion; Digital/analogue conversion
    • H03M1/12—Analogue/digital converters
    • H03M1/124—Sampling or signal conditioning arrangements specially adapted for A/D converters
    • H03M1/1245—Details of sampling arrangements or methods

Definitions

  • the invention relates to the field of ionizing radiation detection and nuclear medicine imaging, and in particular to a digitization method of a scintillation pulse.
  • the energy information of the output pulse of the ionizing radiation detector is a basic information needed in the field of ionizing radiation detection. Its uses include: distinguishing the type of radiation in the detection of ionizing radiation; and determining whether the radiation has occurred with matter in the field of nuclear medicine imaging. Scattering; determining the location of the radiation in the detector in a position sensitive optoelectronic device.
  • the amplitude of the electrical pulse signal that is usually output is linear with the energy of the radiation deposited in the detector, and the rise time and decay time of the pulse are constant. Therefore, the area enclosed by the pulse signal waveform and the time axis (i.e., the amount of total charge generated by the statistical ray in the detector) is typically used to represent the energy of the ionizing radiation source.
  • ADC Analog-to-Digital Converter
  • the integration process of the traditional energy acquisition method limits the maximum count rate of the system; at the same time, the analog integral shaping circuit is easily affected by external factors such as temperature, resulting in performance changes with environmental changes, and the parameters of the analog circuit need to be debugged according to specific applications. This makes the calibration and maintenance of the system quite difficult.
  • Using a high-speed ADC to directly digitize the electrical pulse output from the detector can solve the drawbacks of the traditional method, but at the same time bring the problem of high cost and high power consumption, and the high-speed ADC also puts higher speed on the back-end processing speed and transmission bandwidth. Request, increased backend The design difficulty of the circuit.
  • the multi-voltage threshold method simultaneously inputs the pulse signal and the voltage threshold to both ends of the comparator by setting several voltage thresholds in advance, and measures the time at which the comparator output logic pulse is inverted; these time values and corresponding voltage thresholds constitute the MVT Sampling point; using the prior knowledge of MVT sampling points and pulse signals, the pulse energy information is obtained by reconstructing the pulse by curve fitting and then using numerical integration after definite integral or resampling (specific information reference Qingguo Xie, Chien-Min Kao, Zekai Hsiau, and Chin-Tu Chen, "A New Approach for Pulse Processing in Positron Emission Tomography", IEEE TRANSACTIONS ON NUCLEAR SCIENCE, VOL. 52, NO. 4, AUGUST 2005).
  • the ADC sampling point fitting method uses a lower speed ADC to sample the pulse signal to obtain several ADC sampling points; using these sampling points and pulse prior information, the curve is used to reconstruct the pulse and then use the integral integral or resampled numerical integration method.
  • Obtain pulse energy information (for specific information, refer to Nan Zhang, Niraj Doshi, Mehmet Aykac, Ronald Grazioso, Michael Loope, Greg Givens, Lars Eriksson, Florian Bauer, John Young, and Matthias Schmand, "A Pulse Shape Restore Method for Event Localizaton in PET Scintillation Detection", Nuclear Science Symposium Conference Record, Vol. 7, 2004).
  • Empirical Bayesian estimation method Empirical Bayesian estimation method.
  • the empirical Bayesian estimation method uses the Bayesian principle and the independence hypothesis to solve the maximum likelihood solution of a given digital sample as the estimated pulse energy information (specific information referenceZhenzhou Deng , Qingguo Xie, "Empirical Bayesian energy estimation for Multi-Voltage Threshold digitizer in PET", Nuclear Science Symposium and Medical Imaging Conference, 2013 IEEE).
  • the TOT method estimates the energy expected value of the pulse as the energy information of the pulse by fitting the relationship between the TOT and the pulse energy (for details, refer to D.Nygren, "Converting vice to virtue: can time-walk be used as a measure of deposited charge In silicon detectors?”, Internal LBL note, May 1991).
  • Accurate pulse model and pulse characterization are an important condition for accurate acquisition of pulse information in systems that acquire undersampling pulse energy information based on prior information.
  • the actual pulse physical model and pulse characteristics are not only determined by the detector, but also related to the distribution parameters of the readout circuit.
  • the establishment of the accurate pulse model and the description of the accurate pulse characteristics are very difficult tasks.
  • the ideal model and the characteristic parameters deviate from the pulse model and the characteristic parameters used in the actual fitting or estimation process, the pulse energy information acquired based on the prior information will generate an error.
  • the object of the present invention is to provide a digitization method for scintillation pulses in ionizing radiation detection, which can solve the problem of inaccurate fitting or impulsive feature description in the under-sampling pulse energy information system based on prior information. Accurately bring problems that are not accurate in energy measurement.
  • the present invention provides a digitization method for a scintillation pulse, which includes the steps of:
  • S2 sampling and quantizing each pulse in the pulse database obtained by S1 to obtain complete energy information included in the pulse;
  • S3 undersampling and quantizing each pulse in the pulse database obtained by S1, and estimating or fitting energy information using pulse prior information;
  • S4 using the energy information obtained in S2 as a standard, determining a mapping relationship between the under-sampling pulse energy acquiring method based on the prior information and the energy information obtained by the S2 method;
  • the detector in the step S1 is a gas detector or a scintillator detector or a semiconductor detector.
  • the manner of acquiring the pulse database outputted by the detector under different energy ray irradiation in the step S1 is simulated generation or/and direct acquisition.
  • the method for obtaining the pulse database outputted by the detector under different energy ray irradiation in the step S1 is generated by simulation, and the specific process is as follows:
  • the method for digitizing the scintillation pulse is: firstly, the energy spectrum of the set of pulses is first, the Gaussian function is fitted to the photoluminescence peak in the energy spectrum, and the extracted energy value is located in the fitting.
  • the obtained Gaussian function has all the pulses of the full width at half maximum, assuming that the number is m, and defines S (k, i) as the kth ADC sampling point of the i th pulse, where i is an integer and 0 ⁇ i ⁇ m; k is an integer, and 0 ⁇ k ⁇ h, h is the number of ADC sampling points, and the average pulse P m is defined as a set:
  • the method of scaling the average pulse is linear scaling or non-linear scaling.
  • the method of linearly scaling the average pulse is: assuming that the ray energy corresponding to the average pulse P m obtained in step S1 is E s , and the user determines the upper limit of the energy of the ray to be tested according to actual application requirements.
  • E u the total number of scaling pulses is N, and calculate the maximum multiple of the pulse signal that needs to be amplified.
  • the set of scaling pulses P a Calculated.
  • the set of scaling pulses corresponding to the detector P a is determine.
  • the method for obtaining the pulse outputted by the detector under different energy ray irradiation is directly collected, and the specific process is as follows: firstly, several ray with different energy are used according to user requirements.
  • the source illumination detector uses the ADC to directly collect the output pulse to obtain a pulse set whose energy is distributed within a certain range.
  • the method of acquiring the pulse outputted by the detector under different energy ray irradiation in step S1 is a combination of simulation generation and direct acquisition, and the specific process is as follows: firstly, using multiple known The ray sources of the energy respectively illuminate the detectors, and after obtaining the respective average pulses respectively, the interpolation is used to simulate the pulse database generated by the ray source in a certain energy range to illuminate the detector.
  • the method for digitizing the scintillation pulse is: firstly, the method for generating the average pulse is: firstly, the energy spectrum of the pulse library is used, and the photon peak in the energy spectrum is fitted by a Gaussian function, and the energy value is taken out. All the pulses of the half-height full-width range of the obtained Gaussian function are taken out, and all the pulses whose energy values are in the full width at half maximum of the photoelectric all-energy peak are taken out, assuming that the number is m, and S (k, i) is defined as the i-th pulse.
  • the kth ADC sample point where i is an integer and 0 ⁇ i ⁇ m; k is an integer, and 0 ⁇ k ⁇ h, h is the number of ADC sample points, and the average pulse P m is defined as a set:
  • the method for digitizing the scintillation pulse in the step S2 includes: directly digitizing the pulse to obtain complete energy information included in the pulse, or indirectly sampling to represent pulse energy The physical quantity of information to obtain the complete energy information contained in the pulse.
  • the method for acquiring the complete energy information included in the pulse is realized by directly digitizing the pulse, and the specific method is: using analog-to-digital conversion
  • the scheme or multi-voltage threshold scheme critical/oversampling pulse then use the numerical integration method to obtain pulse energy information, or use the maximum voltage amplitude in the sampling point to represent its energy information.
  • the method for acquiring the complete energy information included in the pulse is realized by directly digitizing the pulse by using a peak hold circuit to maintain the peak value, and then sampling the peak point voltage amplitude.
  • the value represents the pulse energy information.
  • the method for acquiring the complete energy information included in the pulse is realized by indirect sampling, and the specific method is: transforming the pulse amplitude into a time width Wilkinson transformation method .
  • the digitization implementation method of undersampling and quantizing each pulse in the step S3 comprises: using an analog to digital converter or using a comparator with a time to digital converter.
  • the method for estimating or fitting the energy information in the step S3 is the MVT method
  • the method for calculating the pulse energy by using the MVT method is: first setting a plurality of threshold voltages, and setting the pulse sum
  • the threshold voltage is input to the input end of the comparator, and the time at which the comparator output flips the logic pulse is measured.
  • the measured time value and the corresponding threshold voltage constitute an MVT sampling point, and the MVT sampling point and the pulse model are used to reconstruct the pulse.
  • the reconstructed pulse is definitely integrated or resampled and then numerically integrated to obtain the energy information of the pulse.
  • the method for estimating or fitting the energy information in the step S3 is an ADC fitting method
  • the step of calculating the pulse energy by the ADC fitting method is: obtaining the ADC sampling by using the ADC sampling pulse signal. Point, the ADC sampling point and pulse model are used to reconstruct the pulse, and the reconstructed pulse is definitely integrated or resampled and then numerically integrated to obtain the energy information of the pulse.
  • the method of estimating or fitting the energy information in the step S3 is an empirical Bayesian estimation method
  • the specific method of calculating the pulse energy by using the empirical Bayesian estimation method is: obtaining After the pulse digitizes the sample points, the Bayesian principle and the independence hypothesis are used to solve the maximum likelihood solution of a given digital sample as the estimated pulse energy information.
  • the method for estimating or fitting the energy information in the step S3 is a Time Over Threshold (TOT) method
  • the method for calculating the pulse energy by using the TOT method is: Estimate the relationship between TOT and pulse energy The expected energy value of the pulse is used as the energy information of the pulse.
  • the energy mapping relationship in the step S5 is a coefficient lookup table or an energy mapping function.
  • the correction coefficient C is calculated by: assuming complete energy information acquisition for each pulse in the pulse library P a
  • the energy obtained by the method is E (standard, i) .
  • the energy obtained by obtaining the undersampling pulse energy based on prior information is E (statistical, i) , i is an integer, and 0 ⁇ i ⁇ N, then the correction coefficient
  • the function is derived by assuming that for each pulse in the pulse library P a
  • the energy obtained by the complete energy information acquisition method is E (standard, i)
  • the energy obtained by acquiring the undersampled pulse energy based on the prior information is E (statistical, i)
  • i is an integer, and 0 ⁇ i ⁇ N
  • a method for digitizing a scintillation pulse comprising the steps of: S1: acquiring a pulse database output by the detector under different energy ray irradiation; S2: sampling and quantizing each pulse in the pulse database obtained by S1 to obtain a pulse included Complete energy information; S3: Undersampling and quantifying each pulse in the pulse database obtained by S1, estimating or fitting energy information using pulse prior information; S4: using the energy information obtained in S2 as a standard, determining based on The mapping relationship between the under-sampling pulse energy acquisition method of the information and the energy information obtained by the S2 method; S5: correcting the energy information obtained by the under-sampling pulse energy acquisition method based on the prior information by using the energy mapping relationship obtained by S4.
  • the pulse energy information correction method in the present invention uses the pulse energy information obtained by the full energy information acquisition method as a standard to determine The mapping relationship between the under-sampling pulse energy acquisition method based on the prior information and the energy information obtained by the full energy information acquisition method, and the energy mapping relationship can be used. Correcting the energy measurement error caused by imprecise pulse model and feature description in the undersampling pulse energy acquisition method based on prior information.
  • the detector in step S1 is a gas detector or a scintillator detector or a semiconductor detector.
  • any other detector suitable for pulse energy acquisition in ionizing radiation detection is possible.
  • the gas detector utilizes the ionization effect of the ray in the gas medium to realize the conversion of the ray to the current, and has the advantages of simple preparation, stable and reliable performance, convenient use, low cost, and the like; the scintillation detector realizes the ray by using the illuminating effect of the ray in the scintillation crystal
  • the conversion of visible light has the advantages of high detection efficiency, good linearity of energy response, short response time, convenient processing, etc.
  • the semiconductor detector uses radiation to generate electron-hole pairs in the semiconductor to realize the conversion of radiation to electrical signals, with energy resolution. The rate is high and the linear range can be measured.
  • the method of obtaining the pulse database outputted by the detector under different energy ray irradiation in the step S1 is software simulation, and the specific process is as follows: (a) illuminating the detector with a ray source of known energy, and collecting a set by using the ADC Or a plurality of sets of pulse signals obtain a pulse set, and use the pulse set to generate an average pulse; (b) scale the average pulse to simulate a pulse data generated by the detector in a certain energy range.
  • the technical scheme adopts the software simulation method, and utilizes the same characteristics of the pulse characteristics of the same detector under different energy rays, and the pulse database outputted by the detector under different energy ray irradiation can be obtained only by using a single ray source, and the method is simultaneously No need to add any hardware, simple and flexible, easy to implement.
  • the method for generating an average pulse is: firstly, the energy spectrum of the set of pulses is used, and the photon peak in the energy spectrum is fitted by a Gaussian function, and the energy value is taken out in the range of the full width at half maximum of the fitted Gaussian function.
  • Pulse assuming that the number is m, defining S (k, i) as the kth ADC sampling point of the i th pulse, where i is an integer and 0 ⁇ i ⁇ m; k is an integer, and 0 ⁇ k ⁇ h, h is the number of ADC sampling points, and the average pulse P m is defined as a set:
  • the average pulse generated by the technical solution represents the standard pulse of the detector used in the experiment, greatly suppressing the impulse noise, and effectively reducing the influence of the superimposed noise on the pulse on the scaling average pulse process, which can be considered as An ideal pulse with statistical significance.
  • the accuracy of the standard waveform increases with the number m of all pulses in the full width at half maximum.
  • the method of linearly scaling the average pulse is: assuming that the ray energy corresponding to the average pulse P m obtained in step S1 is E s , the user determines the upper limit of the energy of the ray to be measured as E u and the total number of the scaled pulses according to actual application requirements. For N, calculate the maximum multiple of the pulse signal that needs to be amplified.
  • the set of scaling pulses P a Calculated.
  • the technical scheme adopts a linear scaling average pulse. In a certain energy range, it can be assumed that the detector is an ideal linear detector, and pulses outputting different energies have strict linear laws, which makes the correction method easier to implement.
  • the application requirement determines that the upper limit of the energy of the radiation to be measured is E u , the total number of the scaled pulses is N, and the maximum multiple of the pulse signal to be amplified is calculated.
  • the set of scaling pulses corresponding to the detector P a is determine.
  • the technical scheme adopts a nonlinear scaling average pulse, which can reflect the state of the real detector and reduce the influence of the nonlinearity of the detector on the final calibration result.
  • step S1 if the detector outputs the pulse under different energy ray irradiation, if the method is direct acquisition, the specific process is as follows: firstly, according to the user's requirement, a plurality of radiation sources with different energy are used to illuminate the detector, and the ADC is directly collected. The output pulse yields a collection of pulses whose energy is distributed over a range.
  • the technical scheme adopts the direct acquisition method to obtain a pulse set of a wide energy range output by the detector, which truly reflects the pulse characteristics of different energy ray sources and reduces the loss of pulse information.
  • the step S1 is to acquire the pulse outputted by the detector under different energy ray irradiation. If the method is combined with the simulation generation and the direct acquisition, the specific process is: firstly, the detector is irradiated with a plurality of ray sources of known energy, respectively. After obtaining the respective average pulses, the interpolation method is used to simulate the pulse database generated by the radiation source in a certain energy range.
  • the technical scheme adopts a combination of simulation generation and direct acquisition, and has the simplicity of the simulation acquisition method and the authenticity and integrity of the pulse information acquisition by the direct acquisition method.
  • the method for generating an average pulse is: firstly, the energy spectrum of the pulse bank is obtained, and the photon peak in the energy spectrum is fitted by a Gaussian function, and the energy value is taken out in the range of the full width at half maximum of the Gauss function obtained by the fitting.
  • the method for acquiring the complete energy information included in the pulse is realized by direct digital digitization of the pulse by using an analog-to-digital conversion scheme or a multi-voltage threshold scheme critical/oversampling pulse, and then using a numerical integration method to obtain pulse energy. Information, or use the maximum voltage amplitude in the sample point to represent its energy information.
  • the technical solution directly digitizes the pulse by using an analog-to-digital conversion scheme such as a high sampling rate ADC to obtain a pulse “complete data set”, which records all the information of the pulse signal; obtaining the pulse energy information in this way can ensure the acquisition.
  • the energy information is accurate enough.
  • the method for acquiring the complete energy information included in the pulse is realized by directly digitizing the pulse.
  • the specific method is: first, the peak hold circuit is used to maintain the peak value, and then the peak point voltage amplitude is sampled to represent the pulse energy information.
  • the direct digitization method of the technical solution can form a pulse signal by using a peak hold circuit to form a slow signal, and the pulse voltage amplitude can be sampled by using a low sampling rate analog-to-digital conversion device. The energy information obtained in this manner is also very accurate.
  • the method for acquiring the complete energy information included in the pulse is implemented by indirect sampling, and the specific method is: transforming the pulse amplitude into a time width Wilkinson transform method.
  • This technical scheme uses the indirect sampling method of Wilkinson conversion to exert its good differential nonlinearity.
  • the improved Wilkinson can realize high-speed and high-precision analog-to-digital conversion to ensure the complete energy information obtained. Accuracy.
  • the digitized implementation method of undersampling and quantizing each pulse in the step S3 comprises: using an analog to digital converter or using a comparator with a time to digital converter.
  • analog-to-digital converters is the most commonly used digital implementation method, especially when using advanced large-scale integrated circuit technology, the digitization process of analog signals can be realized by using a small chip; and the comparator is represented by the MVT method.
  • the digitalization of the time-to-digital converter it is very easy to obtain the sampling information of the fast-changing signal like the scintillation pulse, which has great advantages in cost and power consumption compared with the ADC which is sampled at equal intervals.
  • the method for estimating or fitting the energy information in the step S3 is the MVT method
  • the method for calculating the pulse energy by using the MVT method is: first setting a plurality of threshold voltages, and inputting the pulse and the threshold voltage to the input of the comparator respectively. End, and measure the time when the comparator output flips the logic pulse, the measured time value and the corresponding threshold voltage constitute the MVT sampling point, and the MVT sampling point and the pulse model are used to reconstruct the pulse, and the reconstructed pulse is definitely integrated or heavily After sampling, the numerical integration is performed to obtain the energy information of the pulse.
  • the technical scheme conveniently and flexibly records the time information corresponding to different voltages, and then uses the characteristic model of the pulse to restore the pulse energy characteristics, which can effectively overcome the sampling of the time interval interval in the face of fast and steep signal sampling. Defects with insufficient sampling rate.
  • the method for estimating or fitting the energy information in the step S3 is an ADC fitting method
  • the step of calculating the pulse energy by the ADC fitting method is: obtaining the ADC sampling point by using the ADC sampling pulse signal, using the ADC sampling point and the pulse model
  • the reconstruction pulse is fitted, and the reconstructed pulse is definitely integrated or resampled, and then numerically integrated to obtain the energy information of the pulse.
  • the technical solution utilizes a few
  • the sampling point and pulse feature model restore the pulse energy characteristics, and it is possible to make the hardware of the system simpler and cheaper by using the rapidly developing DSP technology.
  • the method for estimating or fitting the energy information in the step S3 is the empirical Bayesian estimation method, and the specific method of calculating the pulse energy by using the empirical Bayesian estimation method is: after obtaining the pulse digitized sample point, using the Bayeux
  • the principle and independence hypothesis solves the maximum likelihood solution for a given digital sample as the estimated pulse energy information.
  • the relationship between the system parameters and the pulse information is quantitatively established by effectively extracting and accurately characterizing the statistical characteristics of the pulse, and the accuracy of the calculated pulse energy is maximized by using a statistical model having a fluctuation relationship.
  • the method for estimating or fitting the energy information in the step S3 is the TOT method, and the method for calculating the pulse energy by using the TOT method is: estimating the energy expected value of the pulse as the energy of the pulse by fitting the relationship between the TOT and the pulse energy. information.
  • the technical solution does not require an additional shaping filter circuit or a peak detection and hold circuit, and the analog front end electronics can be very simple and easy to integrate in multiple channels.
  • the energy mapping relationship in the step S5 is a coefficient lookup table, and the correction coefficient C is calculated by: assuming that for each pulse in the pulse library P a , the energy obtained by the complete energy information acquisition method is E (standard, i ) , the energy obtained by acquiring the undersampling pulse energy based on the prior information is E (statistical, i) , i is an integer, and 0 ⁇ i ⁇ N, the correction coefficient
  • FIG. 1 is a flow chart of a method for correcting under-sampling pulse energy information based on prior information in ionizing radiation detection according to the present invention
  • Figure 3 is the average pulse obtained by the NaI/PMT detector under Cs-137 gamma source illumination
  • Figure 5 is a cross-sectional view of the NaI/PMT detector before and after energy spectrum correction by Cs-137 source illumination using the present invention.
  • the invention discloses a digitization method of a scintillation pulse in ionizing radiation detection, which can solve the problem that the energy measurement is inaccurate due to the inaccuracy of the fitting model and the pulse feature description in the under-sampling pulse energy information system based on the prior information. problem.
  • the digitization method of the scintillation pulse in the present invention comprises the steps of:
  • S2 sampling and quantizing each pulse in the pulse database obtained by S1 to obtain complete energy information included in the pulse;
  • S3 undersampling and quantizing each pulse in the pulse database obtained by S1, and estimating or fitting energy information using pulse prior information;
  • S4 using the energy information obtained in S2 as a standard, determining a mapping relationship between the under-sampling pulse energy acquiring method based on the prior information and the energy information obtained by the S2 method;
  • the pulse energy information correction method of the present invention uses the pulse energy information obtained by the full energy information acquisition method as a standard to determine the undersampled pulse based on the prior information.
  • the mapping relationship between the energy acquisition method and the energy information obtained by the full energy information acquisition method, and the energy mapping error is used to correct the energy measurement error caused by the impulsiveness of the pulse model and the feature description in the undersampling pulse energy acquisition method based on the prior information.
  • the detector in the step S1 may be any type of detector suitable for pulse energy acquisition in ionizing radiation detection, and specifically includes a gas detector, a scintillator detector and a semiconductor detector. In addition to these suitable types of detectors, any other detector suitable for pulse energy acquisition in ionizing radiation detection is possible.
  • the gas detector utilizes the ionization effect of the ray in the gas medium to realize the conversion of the ray to the current, and has the advantages of simple preparation, stable and reliable performance, convenient use, low cost, and the like; the scintillation detector realizes the ray by using the illuminating effect of the ray in the scintillation crystal
  • the conversion of visible light has the advantages of high detection efficiency, good linearity of energy response, short response time, convenient processing, etc.
  • the semiconductor detector uses radiation to generate electron-hole pairs in the semiconductor to realize the conversion of radiation to electrical signals, with energy resolution. The rate is high and the linear range can be measured.
  • the step S1 obtains the pulse signal outputted by the detector under different energy ray irradiation, and can be a pulse database for simulating the output of the detector under different energy ray irradiation, or can directly collect the detector under multiple illumination sources.
  • the resulting pulse database can also be a combination of the two methods.
  • step S1 if the detector outputs the pulse signal under different energy ray irradiation, if the simulation is generated, the specific process is as follows:
  • the method of scaling the average pulse is linear scaling or non-linear scaling.
  • the pulse database output by the detector under different energy ray irradiation can be obtained by using only a single ray source, and the method does not need to be increased. Any hardware, simple and flexible, easy to implement.
  • the method of generating the average pulse is as follows: firstly, the energy spectrum of the set of pulses is used, and the Gaussian function is fitted to the photo-peak in the energy spectrum, and the energy value is taken out in the full-width range of the Gaussian function obtained by the fitting.
  • the average linear scaling pulse method Suppose ray energy pulse P m corresponding to an average obtained from step S1 E s, the user determines the upper limit of radiation energy E u is measured based on application requirements, the scaling The total number of pulses is N, and the maximum multiple that the pulse signal needs to be amplified is calculated. The set of scaling pulses P a Calculated. Linearly averaging pulses are used. In a certain energy range, the detector can be assumed to be an ideal linear detector. Pulses with different energy outputs have strict linear laws, which makes the calibration method easier to implement.
  • the user determines the upper limit of the energy of the radiation to be measured as E u according to the actual application requirement, and the total number of the scaling pulses is N, and calculates the maximum multiple of the pulse signal to be amplified.
  • the set of scaling pulses corresponding to the detector P a is determine. Using a nonlinear scaling average pulse, it can reflect the state of the real detector and reduce the influence of the nonlinearity of the detector on the final calibration result.
  • the specific process is as follows: firstly, according to the user's requirement, a plurality of radiation sources with different known ray energies are used to illuminate the detector, and the ADC is directly used. The output pulses are collected to obtain a set of pulses whose energy is distributed over a certain range. Using the direct acquisition method, the pulse set of the wide energy range of the detector output is obtained, which truly reflects the pulse characteristics of different energy ray sources and reduces the loss of pulse information.
  • the step S1 is to obtain the combination of the simulation output and the direct acquisition when the detector outputs the pulse of the different energy ray irradiation, and the specific process is: firstly, using a plurality of ionizing radiation sources of known energy to respectively illuminate the detector, After obtaining the respective average pulses, the interpolation method is used to simulate the pulse database generated by the radiation source in a certain energy range. Using simulation to generate and directly adopt The combination of the combination of the simplicity of the simulation acquisition method and the direct acquisition method for the authenticity and integrity of the pulse information acquisition.
  • the method for generating an average pulse is: firstly, the energy spectrum of the set of pulse banks is obtained, and the photon peak in the energy spectrum is fitted by a Gaussian function, and the extracted energy value is located in the range of the full width at half maximum of the Gauss function obtained by the fitting.
  • the method for acquiring the complete energy information included in the pulse after digitizing the pulse in the step S2 may be: directly digitizing the pulse to obtain the complete energy information included in the pulse, or may be: Indirect sampling can represent the physical quantity of the pulse energy information to obtain the full energy information contained in the pulse.
  • the complete energy information acquisition method is implemented by directly digitizing the pulse to obtain the complete energy information included in the pulse, and the step may be: first using an analog-to-digital conversion scheme or a multi-voltage threshold scheme critical/oversampling pulse. Then use the numerical integration method to obtain the pulse energy information, or use the maximum voltage amplitude in the sampling point to represent its energy information.
  • the digital pulse is directly digitized by an analog-to-digital conversion scheme such as a high sampling rate ADC to obtain a pulse "complete data set", which records all the information of the pulse signal; obtaining pulse energy information in this way ensures that the acquired energy information is sufficient accurate.
  • the complete energy information acquisition method is implemented by directly digitizing the pulse to obtain the complete energy information included in the pulse, and the step may also be: first using the peak hold circuit to maintain the peak value, and then sampling the peak point voltage amplitude. To represent pulse energy information.
  • the pulse signal is formed into a slow signal by the peak hold circuit, and the pulse voltage amplitude can be sampled by using a low sampling rate analog-to-digital conversion device. The energy information obtained in this manner is also very accurate.
  • the full energy information acquisition method in the step S2 is implemented by indirect sampling to obtain the complete energy information included in the pulse, and may be a Wilkinson transform method that converts the pulse amplitude into a time width.
  • the improved Wilkinson can realize high-speed and high-precision analog-to-digital conversion to ensure the accuracy of the complete energy information obtained.
  • the undersampling and quantization process for each pulse in the step S3 is not limited to the digitalization process using an analog-to-digital converter, and the comparator can be digitalized by the time-to-digital converter.
  • the use of analog-to-digital converters is the most commonly used digital implementation, especially with the use of advanced Scale-sized integrated circuit technology can realize the digitization process of analog signals by using a small chip; while the MVT method is used to represent the digitization of the comparator and the time-to-digital converter, it can be easily changed as fast as a scintillation pulse.
  • the sampling information of the signal has a great advantage in cost and power consumption compared with the ADC which is sampled at equal intervals.
  • the method for calculating the energy information by using the statistical estimation method or the curve fitting method in the step S3 may be the MVT method, and the method for calculating the pulse energy by using the MVT method is: first setting a plurality of threshold voltages, respectively, respectively, the pulse and the threshold voltage Input to the input of the comparator, and measure the time when the comparator output flips the logic pulse.
  • the measured time value and the corresponding threshold voltage constitute the MVT sampling point, and the reconstructed pulse is reconstructed by using the MVT sampling point and the pulse model.
  • the pulse is definitely integrated or resampled and then numerically integrated to obtain the energy information of the pulse.
  • the method for calculating the energy information by using the statistical estimation method or the curve fitting method in the step S3 may be an ADC fitting method, and the step of calculating the pulse energy by the ADC fitting method is: obtaining the ADC sampling point by using the ADC sampling pulse signal, and utilizing The ADC sampling point and the pulse model are fitted to reconstruct the pulse, and the reconstructed pulse is definitely integrated or resampled and then numerically integrated to obtain the energy information of the pulse.
  • the method for calculating the energy information by using the statistical estimation method or the curve fitting method in the step S3 may be an empirical Bayesian estimation method, and the specific method of calculating the pulse energy by using the empirical Bayesian estimation method is: obtaining the pulse digital sample After the point, the Bayesian principle and the independence hypothesis are used to solve the maximum likelihood solution of a given digital sample as the estimated pulse energy information.
  • the relationship between the system parameters and the pulse information is quantitatively established, and the accuracy of the calculated pulse energy is maximized by using the statistical model with the fluctuation relationship.
  • the method for calculating the energy information by using the statistical estimation method or the curve fitting method in the step S3 may be a TOT method, and the method for calculating the pulse energy by using the TOT method is: estimating the energy of the pulse by fitting the relationship between the TOT and the pulse energy. The expected value is used as the energy information of the pulse.
  • the technical solution No additional shaping filters or peak detection and hold circuits are required, and analog front-end electronics can be very simple and easy to integrate in multiple channels.
  • the energy mapping relationship in the step S5 may be a coefficient lookup table or an energy mapping function.
  • the pulse database output by the detectors obtained by different methods under different energy ray irradiation is collectively referred to as a wide energy range pulse set P a .
  • the correction coefficient C calculation method may It is assumed that for each pulse in the pulse library P a , the energy obtained by the complete energy information acquisition method is E (standard, i) , and the energy obtained by the under-sampling pulse energy acquisition method based on the prior information is E (statistical, i ) , i is an integer, and 0 ⁇ i ⁇ N, then the correction coefficient By establishing a coefficient lookup table, the correction coefficient corresponding to the energy to be corrected can be quickly and accurately determined.
  • the algorithm is simple, can be implemented on most hardware platforms, and consumes less resources.
  • the function is derived by: assuming that for each pulse in the pulse library P a , the energy obtained by the complete energy information acquisition method is E ( Standard,i) , based on the prior information, the energy obtained by the undersampling pulse energy acquisition method is E (statistical, i) , i is an integer, and 0 ⁇ i ⁇ N, then the complete energy information acquisition method and the priori can be obtained.
  • E Standard,i
  • the energy obtained by the undersampling pulse energy acquisition method is E (statistical, i) , i is an integer, and 0 ⁇ i ⁇ N, then the complete energy information acquisition method and the priori can be obtained.
  • the technical solutions in the embodiments of the present invention are described in detail below with reference to the accompanying drawings in the embodiments of the present invention.
  • the complete energy information acquisition method is a traditional ADC solution.
  • the undersampling pulse energy acquisition method based on prior information is the MVT method and the ADC fitting method.
  • the undersampling pulse energy acquisition method based on the prior information is not limited to the two methods, and the present invention is only for the two methods. The method is explained as a specific embodiment.
  • Embodiment 1 The detector used is a scintillator detector, specifically a sodium iodide (NaI) crystal coupled photomultiplier CR105 (NaI/PMT CR105 detector), and the method for acquiring undersampling pulse energy based on prior information For the MVT method.
  • a scintillator detector specifically a sodium iodide (NaI) crystal coupled photomultiplier CR105 (NaI/PMT CR105 detector)
  • the ray energy value in the full width at half maximum is [632 keV, 692 keV]; all the scintillation pulses within this interval are taken out, assuming that the number is m, and S (k, i) is defined as the ith (i is an integer, and 0 ⁇ i ⁇ m) the kth of the pulse (k is an integer, and 0 ⁇ k ⁇ h, h is the number of ADC sampling points) ADC sampling points, according to The average pulse P m is obtained as shown in FIG.
  • the voltage threshold selected in the MVT method is determined (the setting method can refer to the patent WO2012142778A1).
  • the energy value E (MVT, i) is calculated using the MVT method, i is an integer, and 0 ⁇ i ⁇ 10000, wherein the scintillation pulse model in the MVT method adopts:
  • ⁇ is the decay time constant of the scintillation crystal
  • V p is the amplitude of the scintillation pulse
  • t 0 is the event arrival time
  • t p is the pulse peak time
  • each scintillation pulse ADC sampling point in the scaling pulse set is directly accumulated to obtain the energy value E (ADC, i) in the conventional ADC scheme, where i is an integer and 0 ⁇ i ⁇ 10000.
  • E (ADC, i) the energy value in the conventional ADC scheme, where i is an integer and 0 ⁇ i ⁇ 10000.
  • the circle of Figure 4 is the energy calculated using a conventional ADC scheme for a scintillation pulse in the [0 keV, 3 MeV] energy range.
  • the MVT method to obtain the scintillation pulse energy can be realized by the following process:
  • TDC time-to-digital converter
  • FPGA Field Programmable Gate Array
  • the dichotomy method After fitting the equation, use the dichotomy method to find the intersection point P(Px, Py) of the two curves; After the flicker pulse, the recovered scintillation pulse is resampled at the same sampling rate of 10 GSPS in step (1), and the voltage amplitude of all the sampling points is accumulated to obtain the scintillation pulse energy in the MVT method.
  • the triangle in Fig. 4 is the energy calculated before the correction of the scintillation pulse in the [0keV, 3MeV] energy range using the MVT method.
  • Fig. 5 is a front view before and after correction of the energy spectrum obtained by irradiating a CI-137 source with a NaI/PMT detector.
  • Embodiment 2 The detector used is a NaI/PMT detector, and the under-sampling pulse energy acquisition method based on prior information is an ADC fitting method.
  • the ray energy value in the full width at half maximum is [632 keV, 692 keV]; all the scintillation pulses in the interval [632 keV, 692 keV] are taken, assuming that the number is m, and S (k, i) is defined as the i (i is an integer) And 0 ⁇ i ⁇ m) the kth (k is an integer, and 0 ⁇ k ⁇ h, h is the number of ADC sampling points) ADC sampling points, according to The average pulse P m is obtained .
  • the average pulse is scaled to obtain a database of scintillation pulses generated by the radiation detector corresponding to the range of [0 keV, 3 MeV].
  • the scintillation pulses are required to calculate the amplification
  • the energy value E (ADC-fit, i) is calculated for each scintillation pulse in the scaled pulse set using the ADC fitting method, i is an integer, and 0 ⁇ i ⁇ 10000 .
  • the voltage amplitude of each scintillation pulse ADC sampling point in the scaling pulse set is directly accumulated to obtain the energy value E (ADC, i) in the conventional ADC scheme, where i is an integer and 0 ⁇ i ⁇ 10000.
  • the ADC fitting method to obtain the scintillation pulse energy can be realized by the following process:
  • the ADC2 sampling point is obtained by resampling the scintillation pulse signal with the ADC different from the step (1) (referred to as ADC2), and the data set of one scintillation pulse sampling includes N sample points (t i , v i ), and the method is blinking.
  • Pulse model see, for example, Nan Zhang, Niraj Doshi, Mehmet Aykac, Ronald Grazioso, Michael Loope, Greg Givens, Lars Eriksson, Florian Bauer, John Young, and Matthias Schmand, "A Pulse Shape Restore Method for Event Localizaton in PET Scintillation Detection", Nuclear Science Symposium Conference Record, Vol.7, 2004) can be:
  • a 1 , m 0 , ⁇ 1 , ⁇ 0 are parameters to be determined, ⁇ 0 is the time constant of crystal flicker, and ⁇ 1 is the time constant of the photoelectric conversion device and front-end electronics.

Landscapes

  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • High Energy & Nuclear Physics (AREA)
  • Molecular Biology (AREA)
  • Spectroscopy & Molecular Physics (AREA)
  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Measurement Of Radiation (AREA)

Abstract

一种闪烁脉冲的数字化方法，其包括步骤：获取探测器在不同能量射线照射下输出的脉冲数据库（S1）；对步骤（S1）的脉冲数据库中的每一个脉冲采样并量化以获取到脉冲所包含的完全能量信息（S2）；对步骤（S1）所得脉冲数据库中的每一个脉冲进行欠采样并量化，使用脉冲先验信息估计或者拟合出能量信息（S3）；用步骤（S2）所得能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法和步骤（S2）所述方法得到的能量信息的映射关系（S4）；利用该能量映射关系校正基于先验信息的欠采样脉冲能量获取方法得到的能量信息（S5）。该方案能够校正基于先验信息的欠采样脉冲能量获取方法中由于脉冲模型不精确导致的能量计算误差。

Description

一种闪烁脉冲的数字化方法
本申请要求于2015年1月5日提交中国专利局、申请号为201510003057.1、发明名称为“一种闪烁脉冲的数字化方法”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本发明涉及电离辐射探测和核医学成像领域，具体涉及一种闪烁脉冲的数字化方法。
背景技术
电离辐射探测器输出脉冲的能量信息是电离辐射探测领域需要得到的一个基本信息，其用途包括：在电离辐射探测中用于区分射线种类；在核医学成像领域用于判断射线是否与物质发生了散射作用；在位置敏感型光电器件中确定射线在探测器中的沉积位置。在电离辐射探测器中，对相同探测器，通常其输出的电脉冲信号幅值与射线沉积在探测器内的能量成线性关系，且脉冲的上升时间和衰减时间都为常数。故通常使用脉冲信号波形与时间轴围成的面积(即统计射线在探测器中产生的总电荷的数量)代表电离辐射源的能量。
传统电离辐射探测器中的脉冲能量获取方法有两种：一是通过使用电荷积分电路收集探测器输出脉冲携带的电荷，然后使用低速的模拟数字转换器(Analog-to-Digital Converter，以下简称ADC)采样积分电路中电容器存储的最大电荷量来代表射线沉积在探测器中的能量值。二是将从探测器输出的脉冲进行积分整形成相对较慢速的信号，然后使用低速ADC采样，对采样点做数值积分得到脉冲的能量。传统能量获取方法的积分过程限制了系统的最大计数率；同时模拟的积分整形电路很容易受温度等外界因素的影响，导致性能随环境变化而变化，而且模拟电路的参数需要根据特定应用进行调试，使得系统的校正和维护工作变得相当困难。使用高速ADC直接数字化探测器输出的电脉冲信号可以解决传统方法的缺陷，但同时带来了成本提升和功耗高的问题，而且高速ADC也对后端处理速度和传输带宽提出了更高的要求，增加了后端处 理电路的设计难度。
为了解决高计数能量获取的瓶颈问题，有学者提出了基于先验信息的欠采样脉冲能量信息获取方法。该方法的核心是需要利用探测器输出电脉冲的先验信息，包括脉冲的物理模型、脉冲特征等，只需获取少量的几个采样点，就可以利用统计学原理求解给定样本点的最大似然解或者通过曲线拟合重建脉冲信号，并由此获得脉冲的能量信息。几种典型的基于先验信息获取欠采样脉冲能量信息的方法包括：(1)多电压阈值(Multi-voltage threshold,以下简称MVT)方法。多电压阈值方法通过预先设定好几个电压阈值，将脉冲信号和电压阈值同时输入到比较器的两端，并测量比较器输出逻辑脉冲翻转的时间；这些时间值和对应的电压阈值组成了MVT采样点；使用MVT采样点和脉冲信号的先验知识，通过曲线拟合重建脉冲之后采用定积分或重采样后数值积分的方式获得脉冲能量信息(具体信息参考文献Qingguo Xie,Chien-Min Kao,Zekai Hsiau,and Chin-Tu Chen，“A New Approach for Pulse Processing in Positron Emission Tomography”,IEEE TRANSACTIONS ON NUCLEAR SCIENCE,VOL.52,NO.4,AUGUST 2005)。(2)ADC采样点拟合方法。ADC采样点拟合方法使用较低速ADC采样脉冲信号，得到若干个ADC采样点；使用这些采样点和脉冲先验信息，利用曲线拟合重建脉冲之后采用定积分或重采样后数值积分的方式获得脉冲能量信息(具体信息参考文献Nan Zhang,Niraj Doshi,Mehmet Aykac,Ronald Grazioso,Michael Loope,Greg Givens,Lars Eriksson,Florian Bauer,John Young,and Matthias Schmand,“A Pulse Shape Restore Method for Event Localizaton in PET Scintillation Detection”,Nuclear Science Symposium Conference Record,Vol.7,2004)。(3)经验贝叶斯估计法。经验贝叶斯估计法在获得脉冲数字化样本点后，利用贝叶斯原理和独立性假设求解给定数字样本的最大似然解，将其作为被估计的脉冲能量信息(具体信息参考文献Zhenzhou Deng,Qingguo Xie,“Empirical Bayesian energy estimation for Multi-Voltage Threshold digitizer in PET”,Nuclear Science Symposium and Medical Imaging Conference,2013IEEE)。(4)TOT方法。TOT方法通过拟合TOT与脉冲能量的关系，估计脉冲的能量预期值作为脉冲的能量信息(具体信息参考文献D.Nygren,“Converting vice to virtue:can time-walk be used as a measure of deposited charge in silicon detectors？”,Internal LBL note,May 1991)。
在基于先验信息获取欠采样脉冲能量信息的系统中，精确的脉冲模型和脉冲特征描述是准确获取脉冲信息的一个重要条件。但实际的脉冲物理模型和脉冲特征不仅由探测器所决定，还与其读出电路的分布参数相关，精确脉冲模型的建立和准确脉冲特性的描述都是件非常困难的工作。当理想模型和特征参数与实际拟合或估计过程中采用的脉冲模型和特征参数有偏差时，基于先验信息获取到的脉冲能量信息就会产生误差。
因此，有必要提出一种电离辐射探测中闪烁脉冲的数字化方法，以克服上述缺陷。
发明内容
有鉴于此，本发明的目的在于提供一种电离辐射探测中闪烁脉冲的数字化方法，通过该方法可以解决基于先验信息获取欠采样脉冲能量信息系统中由于拟合模型不精确或脉冲特征描述不准确带来能量测量不准的问题。
为实现本发明上述目的，本发明提供：一种闪烁脉冲的数字化方法，其包括步骤：
S1：获取探测器在不同能量射线照射下输出的脉冲数据库；
S2：对S1所得脉冲数据库中的每一个脉冲采样并量化以获取脉冲所包含的完全能量信息；
S3：对S1所得脉冲数据库中的每一个脉冲进行欠采样并量化，使用脉冲先验信息估计或者拟合出能量信息；
S4：用S2中所得能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法和S2方法得到的能量信息的映射关系；
S5：利用S4所得能量映射关系校正基于先验信息的欠采样脉冲能量获取方法得到的能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S1中的探测器为气体探测器或闪烁体探测器或半导体探测器。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S1中获取探测器在不同能量射线照射下输出的脉冲数据库的方式为仿真产生或/和直接采集。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S1获取探测器在不同能量射线照射下输出的脉冲数据库的方式若为仿真产生，其具体过程为：
(a)使用已知能量的射线源照射探测器，利用ADC采集一组或若干组脉冲信号得到脉冲集合，利用该脉冲集合生成平均脉冲；
(b)将该平均脉冲进行缩放，模拟得到一定能量范围内的射线照射该探测器产生的脉冲数据库。
上述的闪烁脉冲的数字化方法，优选地，所述生成平均脉冲的方法为：首先做出这组脉冲的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000001
上述的闪烁脉冲的数字化方法，优选地，所述缩放平均脉冲的方法为线性缩放或非线性缩放。
上述的闪烁脉冲的数字化方法，优选地，线性缩放平均脉冲的方法为：假 设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000002
缩放脉冲集合Pa由
Figure PCTCN2015092909-appb-000003
计算得到。
上述的闪烁脉冲的数字化方法，优选地，非线性缩放平均脉冲的方法为：首先确定探测器的非线性能量响应曲线y＝f(x)，假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000004
该探测器对应的缩放脉冲集合Pa则由
Figure PCTCN2015092909-appb-000005
确定。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S1获取探测器在不同能量射线照射下输出脉冲的方式若为直接采集，其具体过程为：首先根据用户需求使用若干个具有不同能量的射线源照射探测器，利用ADC直接采集输出脉冲得到能量分布在一定范围内的脉冲集合。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S1获取探测器在不同能量射线照射下输出的脉冲的方式若为仿真产生和直接采集的结合，其具体过程为：首先使用多个已知能量的射线源分别照射探测器，在分别得到各自的平均脉冲后，使用插值法模拟得到一定能量范围内的射线源照射该探测器产生的脉冲数据库。
上述的闪烁脉冲的数字化方法，优选地，所述生成平均脉冲的方法为：首先做出这组脉冲库的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，取出脉冲库中能量值位于光电全能峰半高全宽的范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000006
上述的闪烁脉冲的数字化方法，优选地，所述步骤S2中获取脉冲所包含的完全能量信息的方法包括：对脉冲直接数字化以获取脉冲所包含的完全能量信息，或者经过间接采样可代表脉冲能量信息的物理量来获取脉冲所包含的完全能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述获取脉冲所包含的完全能量信息的方法如果通过脉冲直接数字化的方式实现，具体方法为：使用模数转换 方案或者多电压阈值方案临界/过采样脉冲，然后使用数值积分方法获取脉冲能量信息，或使用采样点中最大电压幅值代表其能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述获取脉冲所包含的完全能量信息的方法如果通过脉冲直接数字化的方式实现，具体方法为：先使用峰值保持电路保持峰值，然后采样峰值点电压幅值来代表脉冲能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述获取脉冲所包含的完全能量信息的方法如果通过间接采样的方式实现，具体方法为：将脉冲幅值变换成时间宽度的威尔金逊变换法。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S3中对每一个脉冲进行欠采样并量化的数字化实现方法包括：使用模拟数字转换器实现或者使用比较器配合时间数字转换器实现。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S3中估计或者拟合出能量信息的方法为MVT方法，利用MVT方法计算脉冲能量的方式为：首先设定多个阈值电压，将脉冲和阈值电压分别输入到比较器的输入端，并测量比较器输出翻转逻辑脉冲的时间，测量得到的时间值和对应的阈值电压组成MVT采样点，利用MVT采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S3中估计或者拟合出能量信息的方法为ADC拟合方法，ADC拟合方法计算脉冲能量的步骤为：使用ADC采样脉冲信号得到ADC采样点，利用ADC采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S3中估计或者拟合出能量信息的方法为经验贝叶斯估计法，利用经验贝叶斯估计法计算脉冲能量的具体方式为：在获得脉冲数字化样本点后，使用贝叶斯原理和独立性假设求解给定数字样本的最大似然解，将其作为被估计的脉冲能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S3中估计或者拟合出能量信息的方法为过阈值时间(Time Over Threshold，以下简称TOT)方法，利用TOT方法计算脉冲能量的方式为：通过拟合TOT与脉冲能量的关系，估 计脉冲的能量预期值作为脉冲的能量信息。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S5中的能量映射关系是系数查找表或者能量映射函数。
上述的闪烁脉冲的数字化方法，优选地，若所述步骤S5中的能量映射关系为系数查找表，则校正系数C计算方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则校正系数
Figure PCTCN2015092909-appb-000007
上述的闪烁脉冲的数字化方法，优选地，所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝C*Estatistical。
上述的闪烁脉冲的数字化方法，优选地，若所述步骤S4中的能量映射关系若为映射函数y＝g(x)，该函数的推导方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠 采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则可以得到完全能量信息获取方法和基于先验信息获取欠采样脉冲能量获取方法的若干个系数映射点：
Figure PCTCN2015092909-appb-000008
利用这些系数映射点，经过曲线拟合得到映射函数y＝g(x)。
上述的闪烁脉冲的数字化方法，优选地，所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝g(Estatistical)。
以上技术方案相对于现有技术具有如下优点：
1、一种闪烁脉冲的数字化方法，其包括步骤：S1：获取探测器在不同能量射线照射下输出的脉冲数据库；S2：对S1所得脉冲数据库中的每一个脉冲采样并量化以获取脉冲所包含的完全能量信息；S3：对S1所得脉冲数据库中的每一个脉冲进行欠采样并量化，使用脉冲先验信息估计或者拟合出能量信息；S4：用S2中所得能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法和S2方法得到的能量信息的映射关系；S5：利用S4所得能量映射关系校正基于先验信息的欠采样脉冲能量获取方法得到的能量信息。从上述技术方案可以看出，因为脉冲模型与特征描述不会影响到完全能量信息获取方法的精度，本发明中的脉冲能量信息校正方法使用完全能量信息获取方法得到的脉冲能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法与完全能量信息获取方法所得能量信息之间的映射关系，使用该能量映射关系可以 校正基于先验信息的欠采样脉冲能量获取方法中由于脉冲模型与特征描述不精确导致的能量测量误差。
2、所述步骤S1中的探测器为气体探测器或闪烁体探测器或半导体探测器。除了这几种比较适合的探测器种类外，其他适用于电离辐射探测中脉冲能量获取的任何一种探测器都是可以的。气体探测器利用射线在气体介质中的电离效应实现射线到电流的转换，具有制备简单、性能稳定可靠、使用方便、成本低廉等优点；闪烁探测器利用射线在闪烁晶体中的发光效应实现射线到可见光子的转换，具有探测效率高、能量响应线性度好、响应时间短、加工方便等优点；半导体探测器利用射线在半导体中产生电子空穴对来实现射线到电信号的转换，具有能量分辨率高、可测量线性范围宽的优点。
3、所述步骤S1获取探测器在不同能量射线照射下输出的脉冲数据库的方式若为软件仿真，其具体过程为：(a)使用已知能量的射线源照射探测器，利用ADC采集一组或若干组脉冲信号得到脉冲集合，利用该脉冲集合生成平均脉冲；(b)将该平均脉冲进行缩放，模拟得到一定能量范围内的射线照射该探测器产生的脉冲数据库。本技术方案通过软件仿真的方式，利用同一个探测器在不同能量射线作用下脉冲特征相同的性质，仅使用单一射线源即可获取探测器在不同能量射线照射下输出的脉冲数据库，同时此方法不需要增加任何硬件，简单灵活，易于实现。
4、所述生成平均脉冲的方法为：首先做出这组脉冲的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样 点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000009
本技术方案所产生的平均脉冲，代表了当次实验所用探测器的标准脉冲，极大地抑制了脉冲噪声，有效地减小了脉冲上叠加的噪声对缩放平均脉冲过程的影响，可以被认为是具有统计意义的理想脉冲。同时，标准波形的精度会随半高全宽范围内所有脉冲的数目m增加而提高。
5、线性缩放平均脉冲的方法为：假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000010
缩放脉冲集合Pa由
Figure PCTCN2015092909-appb-000011
计算得到。本技术方案采用线性缩放平均脉冲，在一定能量范围内，可以假设探测器是理想的线性探测器，输出不同能量的脉冲具有严格的线性规律，使得本校正方法实现起来更为简便。
6、非线性缩放平均脉冲的方法为：首先确定探测器的非线性能量响应曲线y＝f(x)，假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000012
该探测器对应的缩放脉冲集合Pa则由
Figure PCTCN2015092909-appb-000013
确定。本技术方案采用非线性缩放平均脉冲，能够反映真实探测器的状态，减少探测器的非线性对最终校正结果的影响。
7、所述步骤S1获取探测器在不同能量射线照射下输出脉冲的方式若为直接采集，其具体过程为：首先根据用户需求使用若干个具有不同能量的放射源照射探测器，利用ADC直接采集输出脉冲得到能量分布在一定范围内的脉冲集合。本技术方案采用直接采集的方式，得到探测器输出的宽能量范围脉冲集合，真实反映不同能量射线源的脉冲特性，减少了脉冲信息损失。
8、所述步骤S1获取探测器在不同能量射线照射下输出的脉冲的方式若为仿真产生和直接采集的结合，其具体过程为：首先使用多个已知能量的射线源分别照射探测器，在分别得到各自的平均脉冲后，使用插值法模拟得到一定能量范围内的射线源照射该探测器产生的脉冲数据库。本技术方案采用仿真产生和直接采集结合的方式，兼具了仿真获取方式的简便性与直接采集方式对脉冲信息获取的真实性、完整性。
9、所述生成平均脉冲的方法为：首先做出这组脉冲库的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000014
本技术方案所产生的平均脉冲，代表了当次实验所用探测器的标准脉冲，极大地抑制了脉冲噪声，有效地减小了脉冲上叠加的噪声对缩放平均脉冲过程的影响，可以被认为是具有统计意义的理想脉冲。同时，标准波形的精度会随半高全宽范围内所有脉冲的数目m增加而提高。
10、所述获取脉冲所包含的完全能量信息的方法通过脉冲直接数字化的方式实现，具体方法为：使用模数转换方案或者多电压阈值方案临界/过采样脉冲，然后使用数值积分方法获取脉冲能量信息，或使用采样点中最大电压幅值代表其能量信息。本技术方案利用高采样率ADC等模数转换方案直接数字化脉冲，以获取脉冲“完全数据集”，此数字信号记录了脉冲信号的所有信息；用这种方式获取脉冲能量信息，能保证获取的能量信息足够准确。
11、所述获取脉冲所包含的完全能量信息的方法通过脉冲直接数字化的方式实现，具体方法为：先使用峰值保持电路保持峰值，然后采样峰值点电压幅值来代表脉冲能量信息。本技术方案直接数字化的方法，通过峰值保持电路将脉冲信号整形成慢速信号，就可以使用低采样率模数转换器件采样脉冲电压幅值，此方式获取的能量信息也非常准确。
12、所述获取脉冲所包含的完全能量信息的方法如果通过间接采样的方式实现，具体方法为：将脉冲幅值变换成时间宽度的威尔金逊变换法。本技术方案使用威尔金逊转换的间接采样的方式，发挥其良好的微分非线性性质，利用改进型的威尔金逊可以实现高速高精度的模拟数字转换，以保证获得的完全能量信息的精确度。
13、所述步骤S3中对每一个脉冲进行欠采样并量化的数字化实现方法包括：使用模拟数字转换器实现或者使用比较器配合时间数字转换器实现。使用模拟数字转换器是最常用的数字化实现方法，特别是利用先进的大规模集成电路技术时，使用一块体积很小的芯片即可实现模拟信号的数字化过程；而以MVT方法为代表使用比较器配合时间数字转换器实现数字化的方式，可以非常容易获取像闪烁脉冲一样快速变化信号的采样信息，与等时间间隔采样的ADC相比，在成本和功耗上具有很大优势。
14、所述步骤S3中估计或者拟合出能量信息的方法为MVT方法，利用MVT方法计算脉冲能量的方式为：首先设定多个阈值电压，将脉冲和阈值电压分别输入到比较器的输入端，并测量比较器输出翻转逻辑脉冲的时间，测量得到的时间值和对应的阈值电压组成MVT采样点，利用MVT采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。本技术方案通过设计多个阈值电压，方便灵活地记录下不同电压所对应的时间信息，再利用脉冲的特征模型还原脉冲能量特性，可以有效克服等时间间隔采样在面临快速、陡峭信号采样时的采样率不够高的缺陷。
15、所述步骤S3中估计或者拟合出能量信息的方法为ADC拟合方法，ADC拟合方法计算脉冲能量的步骤为：使用ADC采样脉冲信号得到ADC采样点，利用ADC采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。本技术方案通过利用少数几个 采样点和脉冲特征模型还原脉冲能量特性，有可能利用快速发展的DSP技术使系统的硬件更加简单、成本更加低廉。
16、所述步骤S3中估计或者拟合出能量信息的方法为经验贝叶斯估计法，利用经验贝叶斯估计法计算脉冲能量的具体方式为：在获得脉冲数字化样本点后，使用贝叶斯原理和独立性假设求解给定数字样本的最大似然解，将其作为被估计的脉冲能量信息。本技术方案，通过有效提取和准确刻画脉冲的统计学特征，定量的建立起系统参数与脉冲信息之间的关系，利用具有涨落关系的统计学模型使得计算脉冲能量的精确度最大化。
17、所述步骤S3中估计或者拟合出能量信息的方法为TOT方法，利用TOT方法计算脉冲能量的方式为：通过拟合TOT与脉冲能量的关系，估计脉冲的能量预期值作为脉冲的能量信息。本技术方案不需要额外整形滤波电路或者峰值检测和保持电路，模拟前端电子学可以非常简单，易于多通道集成。
18、所述步骤S5中的能量映射关系为系数查找表，则校正系数C计算方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则校正系数
Figure PCTCN2015092909-appb-000015
所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝C*Estatistical。本技术方案通过建立系数查找表，可以快速准确地确定待校正能量所对应的校正系数，算法 简单，可以在绝大部分硬件平台上实现，而且占用资源少。
19、所述步骤S4中的能量映射关系若为映射函数y＝g(x)，该函数的推导方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则可以得到完全能量信息获取方法和基于先验信息获取欠采样脉冲能量获取方法的若干个系数映射点：
Figure PCTCN2015092909-appb-000016
利用这些系数映射点，使用曲线拟合得到映射函数y＝g(x)，所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝g(Estatistical)。本技术方案通过曲线拟合的方式，减小单个系数映射点上叠加的计算误差，使得校正过程更加稳健。
附图说明
图1为本发明涉及到的电离辐射探测中基于先验信息获取欠采样脉冲能量信息校正方法的流程图；
图2为Cs-137伽马射线源照射NaI/PMT探测器使用完全能量信息获取方法得到的能量谱及其光电峰的高斯拟合(将光电峰校正到662keV)；
图3为NaI/PMT探测器在Cs-137伽马射源照射下得到的平均脉冲；
图4为NaI/PMT探测器运用本发明校正前和校正后的能量曲线对比图；
图5为运用本发明对NaI/PMT探测器在Cs-137射源照射下能谱校正前后对照图。
具体实施方式
本发明公开了一种电离辐射探测中闪烁脉冲的数字化方法，通过该方法可以解决基于先验信息获取欠采样脉冲能量信息系统中由于拟合模型与脉冲特征描述不精确带来能量测量不准的问题。
如图1所示，本发明中的闪烁脉冲的数字化方法，包括步骤：
S1：获取探测器在不同能量射线照射下输出的脉冲数据库；
S2：对S1所得脉冲数据库中的每一个脉冲采样并量化以获取脉冲所包含的完全能量信息；
S3：对S1所得脉冲数据库中的每一个脉冲进行欠采样并量化，使用脉冲先验信息估计或者拟合出能量信息；
S4：用S2中所得能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法和S2方法得到的能量信息的映射关系；
S5：利用S4所得能量映射关系校正基于先验信息的欠采样脉冲能量获取方法得到的能量信息。
因为脉冲模型与特征描述不会影响到完全能量信息获取方法的精度，本发明中的脉冲能量信息校正方法使用完全能量信息获取方法得到的脉冲能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法与完全能量信息获取方法所得能量信息之间的映射关系，使用该能量映射关系校正基于先验信息的欠采样脉冲能量获取方法中由于脉冲模型与特征描述不精确导致的能量测量误差。
所述步骤S1中探测器可以为适用于电离辐射探测中脉冲能量获取的任何一种探测器，具体包括气体探测器、闪烁体探测器和半导体探测器。除了这几种比较适合的探测器种类外，其他适用于电离辐射探测中脉冲能量获取的任何一种探测器都是可以的。气体探测器利用射线在气体介质中的电离效应实现射线到电流的转换，具有制备简单、性能稳定可靠、使用方便、成本低廉等优点；闪烁探测器利用射线在闪烁晶体中的发光效应实现射线到可见光子的转换，具有探测效率高、能量响应线性度好、响应时间短、加工方便等优点；半导体探测器利用射线在半导体中产生电子空穴对来实现射线到电信号的转换，具有能量分辨率高、可测量线性范围宽的优点。
所述步骤S1获取探测器在不同能量射线照射下输出的脉冲信号的方式可以为仿真产生探测器在不同能量射线照射下输出的脉冲数据库，也可以为直接采集在多个射源照射下探测器产生的脉冲数据库，也可以是两种方法的结合。
所述步骤S1获取探测器在不同能量射线照射下输出脉冲信号的方式若为仿真产生，其具体过程为：
(a)使用已知能量的射线源照射探测器，利用ADC采集一组或若干组脉冲信号得到脉冲集合，利用该脉冲集合生成平均脉冲；
(b)将该平均脉冲进行缩放，模拟得到一定能量范围内的射线照射该探测器产生的脉冲数据库。所述缩放平均脉冲的方法为线性缩放或非线性缩放。
通过软件仿真的方式，利用同一个探测器在不同能量射线作用下脉冲特征相同的性质，仅使用单一射线源即可获取探测器在不同能量射线照射下输出的脉冲数据库，同时此方法不需要增加任何硬件，简单灵活，易于实现。
上述(a)中，生成平均脉冲的方法为：首先做出这组脉冲的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000017
本方法产生的平均脉冲，代表了当次实验所用探测器的标准脉冲，极大地抑制了脉冲噪声，有效地减小了脉冲上叠加的噪声对缩放平均脉冲过程的影响，可以 被认为是具有统计意义理想脉冲，同时标准波形的精度会随半高全宽范围内所有脉冲的数目m增加而提高。
上述(b)中，线性缩放平均脉冲的方法为：假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000018
缩放脉冲集合Pa由
Figure PCTCN2015092909-appb-000019
计算得到。采用线性缩放平均脉冲，在一定能量范围内，可以假设探测器是理想的线性探测器，输出不同能量的脉冲具有严格的线性规律，使得本校正方法实现起来更为简便。
上述(b)中，非线性缩放平均脉冲的方法为：首先确定探测器的非线性能量响应曲线y＝f(x)，假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目 为N，计算出脉冲信号需要放大的最大倍数
Figure PCTCN2015092909-appb-000020
该探测器对应的缩放脉冲集合Pa则由
Figure PCTCN2015092909-appb-000021
确定。采用非线性缩放平均脉冲，能够反映真实探测器的状态，减少探测器的非线性对最终校正结果的影响。
所述步骤S1获取探测器在不同能量射线照射下输出脉冲的方式若为直接采集，其具体过程为：首先根据用户需求使用若干个具有不同已知射线能量的放射源照射探测器，利用ADC直接采集输出脉冲得到能量分布在一定范围内的脉冲集合。采用直接采集的方式，得到探测器输出的宽能量范围脉冲集合，真实反映不同能量射线源的脉冲特性，减少了脉冲信息损失。
所述步骤S1获取探测器在不同能量射线照射下输出的脉冲的方式若为仿真产生和直接采集的结合，其具体过程为：首先使用多个已知能量的电离辐射源分别照射探测器，在分别得到各自的平均脉冲后，使用插值法模拟得到一定能量范围内的辐射源照射该探测器产生的脉冲数据库。采用仿真产生和直接采 集结合的方式，兼具了仿真获取方式的简便性与直接采集方式对脉冲信息获取的真实性、完整性。
其中，所述生成平均脉冲的方法为：首先做出这组脉冲库的能量谱，对能量谱中的光电峰使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，取出脉冲库中能量值位于光电全能峰半高全宽的范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
Figure PCTCN2015092909-appb-000022
本方法产生的平均脉冲，代表了当次实验所用探测器的标准脉冲，极大地抑制了脉冲噪声，有效地减小了脉冲上叠加的噪声对缩放平均脉冲过程的影响，可以被认为是具 有统计意义的理想脉冲。同时，标准波形的精度会随半高全宽范围内所有脉冲的数目m增加而提高。
所述步骤S2中数字化脉冲后获取到脉冲所包含的完全能量信息的方法(以下简称，完全能量信息获取方法)可以是：脉冲直接数字化以获取脉冲所包含的完全能量信息，也可以是：经过间接采样可代表脉冲能量信息的物理量来获取脉冲所包含的完全能量信息。
所述步骤S2中完全能量信息获取方法若通过脉冲直接数字化以获取脉冲所包含的完全能量信息的方式实现，则其步骤可以为：先使用模数转换方案或者多电压阈值方案临界/过采样脉冲，然后使用数值积分方法获取脉冲能量信息，或使用采样点中最大电压幅值代表其能量信息。利用高采样率ADC等模数转换方案直接数字化脉冲，以获取脉冲“完全数据集”，此数字信号记录了脉冲信号的所有信息；用这种方式获取脉冲能量信息，能保证获取的能量信息足够准确。
所述步骤S2中完全能量信息获取方法若通过脉冲直接数字化以获取脉冲所包含的完全能量信息的方式实现，则其步骤也可以为：先使用峰值保持电路保持峰值，然后采样峰值点电压幅值来代表脉冲能量信息。采用直接数字化的方法，通过峰值保持电路将脉冲信号整形成慢速信号，就可以使用低采样率模数转换器件采样脉冲电压幅值，此方式获取的能量信息也非常准确。
所述步骤S2中完全能量信息获取方法若通过间接采样以获取脉冲所包含的完全能量信息的方式实现，可以为将脉冲幅值变换成时间宽度的威尔金逊变换法。使用威尔金逊转换的间接采样的方式，发挥其良好的微分非线性性质，利用改进型的威尔金逊可以实现高速高精度的模拟数字转换，以保证获得的完全能量信息的精确度。
所述步骤S3中对每一个脉冲进行欠采样并量化过程，不仅限于使用模拟数字转换器实现数字化过程，还可以是比较器配合时间数字转换器实现数字化过程。使用模拟数字转换器是最常用的数字化实现方法，特别是利用先进的大 规模集成电路技术，使用一块体积很小的芯片即可实现模拟信号的数字化过程；而以MVT方法为代表使用比较器配合时间数字转换器实现数字化的方式，可以非常容易获取像闪烁脉冲一样快速变化的信号的采样信息，与等时间间隔采样的ADC相比，在成本和功耗上具有很大优势。
所述步骤S3中使用统计学估计法或者曲线拟合法计算出能量信息的方法，可以为MVT方法，利用MVT方法计算脉冲能量的方式为：首先设定多个阈值电压，将脉冲和阈值电压分别输入到比较器的输入端，并测量比较器输出翻转逻辑脉冲的时间，测量得到的时间值和对应的阈值电压组成MVT采样点，利用MVT采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。通过设计多个阈值电压，方便灵活地记录下不同电压所对应的时间信息，再利用脉冲的特征模型还原脉冲能量特性，可以有效克服等时间间隔采样在面临快速、陡峭信号采样时的采样率不够高的缺陷。
所述步骤S3中使用统计学估计法或曲线拟合法计算出能量信息的方法，可以为ADC拟合方法，ADC拟合方法计算脉冲能量的步骤为：使用ADC采样脉冲信号得到ADC采样点，利用ADC采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。通过利用少数几个采样点和脉冲特征模型还原脉冲能量特性，有可能利用快速发展的DSP技术使系统的硬件更加简单、成本更加低廉。
所述步骤S3中使用统计学估计法或曲线拟合法计算出能量信息的方法，可以为经验贝叶斯估计法，利用经验贝叶斯估计法计算脉冲能量的具体方式为：在获得脉冲数字化样本点后，使用贝叶斯原理和独立性假设求解给定数字样本的最大似然解，将其作为被估计的脉冲能量信息。通过有效提取和准确刻画脉冲的统计学特征，定量的建立起系统参数与脉冲信息之间的关系，利用具有涨落关系的统计学模型使得计算脉冲能量的精确度最大化。
所述步骤S3中使用统计学估计法或者曲线拟合法计算出能量信息的方法，可以为TOT方法，利用TOT方法计算脉冲能量的方式为：通过拟合TOT与脉冲能量的关系，估计脉冲的能量预期值作为脉冲的能量信息。本技术方案 不需要额外整形滤波电路或者峰值检测和保持电路，模拟前端电子学可以非常简单，易于多通道集成。
所述步骤S5中的能量映射关系可以是系数查找表，也可以是能量映射函数。
约定通过不同方法获取的探测器在不同能量射线照射下输出的脉冲数据库统称为宽能量范围脉冲集Pa，所述步骤S5中的能量映射关系若为系数查找表，则校正系数C计算方法可以为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则校正系数
Figure PCTCN2015092909-appb-000023
通过建立系数查找表，可以快速准确地确定待校正能量所对应的校正系数，算法简单，可以在绝大部分硬件平台上实现，而且占用资源少。
所述步骤S4中的能量映射关系若为映射函数y＝g(x)，该函数的推导方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量 为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则可以得到完全能量信息获取方法和基于先验信息获取欠采样脉冲能量获取方法的若干个系数映射点：
Figure PCTCN2015092909-appb-000024
利用这些系数映射点，使用曲线拟合得到映射函数y＝g(x)。通过曲线拟合的方式，减小单个系数映射点上叠加的计算误差，使得校正过程更加稳健。
所述步骤S5中基于先验信息的欠采样脉冲能量获取方法得到的能量ECStatistical的计算公式为：ECStatistical＝C*Estatistical或者ECStatistical＝g(Estatistical)。
下面将结合本发明实施例中的附图，对本发明实施例中的技术方案进行详细地描述。实施例在获取探测器在不同能量射线照射下输出的脉冲集的过程中选取三种方法中最简单的仿真产生脉冲数据集的方式说明技术方案，实施例中完全能量信息获取方法为传统ADC方案，基于先验信息的欠采样脉冲能量获取方法为MVT方法和ADC拟合方法，当然基于先验信息的欠采样脉冲能量获取方法并不局限于该两种方法，本发明仅仅是将这两种方法作为具体的实施例对本发明进行解释。
实施例一：采用的探测器为闪烁体探测器，具体为碘化钠(NaI)晶体耦合光电倍增管CR105(NaI/PMT CR105探测器)，所述基于先验信息的欠采样脉冲能量获取方法为MVT方法。
(1)使用Cs-137射源照射NaI/PMT探测器，将探测器输出的闪烁脉冲直接输入示波器中，在采样率为10GSPS的工作状态下采集5000个闪烁脉冲，并统计得到如图2虚线所示的能量谱(将峰值能量校正到662keV)。使用高斯曲线拟合其光电峰，如图2实线，得到其能量分辨率为8.6％；通过能量分辨率和峰值能量值计算得到半高全宽值约为30keV，因此得到当峰值能量为662keV时对应半高全宽范围内的射线能量值为[632keV,692keV]；取出这个区间之内的所有闪烁脉冲，假设其数目为m，定义S(k,i)为其中第i(i为整数，且0<i≤m)个脉冲的第k(k为整数，且0<k≤h,h为ADC采样点的数目)个ADC采样点，按照
Figure PCTCN2015092909-appb-000025
得到平均脉冲Pm，如图3所示。
(2)将该平均脉冲进行比例缩放，模拟得到对应[0keV,3MeV]范围内 的射线照射探测器产生的闪烁脉冲数据库。假设由步骤(1)得到的平均脉冲Pm对应的射线能量Es＝662keV，待测射线的能量上限Eu＝3MeV,则计算出闪烁脉冲需要放大的最大倍数
Figure PCTCN2015092909-appb-000026
选择缩放脉冲的总数目N＝10000，缩放脉冲集Pa由
Figure PCTCN2015092909-appb-000027
计算得到。
(3)根据步骤(1)得到的平均脉冲，确定MVT方法中选取的电压阈值(设置方法可参考专利WO2012142778A1)。对缩放脉冲集中的每一个闪烁脉冲使用MVT方法计算得到能量值E(MVT,i),i为整数，且0<i≤10000，其中MVT方法中的闪烁脉冲模型采用：
Figure PCTCN2015092909-appb-000028
   (式1)
其中，τ为闪烁晶体的衰减时间常数，Vp为闪烁脉冲的幅值，t0为事件到达时间，tp为脉冲峰值时间(具体参见Qingguo Xie,Chien-Min Kao,Zekai Hsiau,and Chin-Tu Chen，“A New Approach for Pulse Processing in Positron Emission Tomography”,IEEE TRANSACTIONS ON NUCLEAR SCIENCE,VOL.52,NO.4,AUGUST 2005.)。同时，将缩放脉冲集中的每一个闪烁脉冲ADC采样点的电压幅值直接累加得到传统ADC方案中的能量值E(ADC,i)，其中i为整数，且0<i≤10000。图4的圆圈为使用传统ADC方案对[0keV,3MeV]能量范围内的闪烁脉冲计算得到的能量。
MVT方法获取闪烁脉冲能量具体可以通过以下过程实现：
①预先设定8个电压阈值为[20 38 56 80 250 500 1000 1500]，将探测器输出的闪烁脉冲和设定的电压阈值输入到8个比较器中，将比较器的输出逻辑电 平跳转信号对应的时间信息使用时间-数字转换器(Time-to-digital converter,TDC)转换成数字信号，其中TDC由可编程门阵列(Field Programmable Gate Array,FPGA)实现，假设8个比较器输出逻辑电平的正跳变和负跳变对应触发的16个时间值为TrigTime＝{Tr1,Tr2,Tr3,Tr4,Tr5,Tr6,Tr7,Tr8,Td8,Td7,Td6,Td5,Td4,Td3,Td2,Td1}，其中Tr1,Tr2,Tr3,Tr4,Tr5,Tr6,Tr7,Tr8表示8个阈值触发的上升沿采样时间，Td8,Td7,Td6,Td5,Td4,Td3,Td2,Td1表示对应的下降沿采样时间；这样{(Tr1,20),(Tr2,38),…,(Tr8,1500),(Td8,1500),…,(Td1,20)}组成MVT方法的16组采样点。
②利用上一步得到的16组采样点恢复出原始闪烁脉冲。首先建立闪烁脉冲模型，本例中使用一个直线的上升沿加上一个指数的下降沿(具体表达式见式1)拟合闪烁脉冲。设上升沿函数方程表示为yr＝k*xr+b，下降沿函数方程表示为yd＝ea*xd+c，拟合方程后利用二分法求出两曲线交点P(Px,Py)；重建闪烁脉冲后，以第(1)步中相同的采样率10GSPS对恢复出的闪烁脉冲进行重采样，并累加所有采样点的电压幅值得到MVT方法中的闪烁脉冲能量。图4中的三角为使用MVT方法对[0keV,3MeV]能量范围内的闪烁脉冲校正前计算得到的能量。
(4)计算校正系数
Figure PCTCN2015092909-appb-000029
建立校正系数查找表，或者通过拟合校正系数建立E(MVT,i)与E(ADC,i)的函数映射关系。图4中的十字形为使用MVT方法对[0keV,3MeV]能量范围内的闪烁脉冲校正后得到的能量。
(5)通过步骤(4)得到的查找表或者函数方程生成校正系数C，并计算校正后的能量ECMVT＝C*EMVT。图5为运用本发明对Cs-137射源照射NaI/PMT探测器得到的能谱校正前后对照图。
实施例二：采用的探测器为NaI/PMT探测器，所述基于先验信息的欠采样脉冲能量获取方法为ADC拟合方法。
(1)使用Cs-137射源照射NaI/PMT探测器，将探测器输出的闪烁脉冲直接输入示波器中，在采样率为10GSPS的工作状态下采集5000个闪烁脉冲，并统计得到能量谱(将峰值能量校正到662keV)。使用高斯曲线拟合其光电峰，如图2实线，得到其能量分辨率为8.6％；通过能量分辨率和峰值能量值计算得到半高全宽值约为30keV，因此得到当峰值能量为662keV时对应半高全宽范围内的射线能量值为[632keV,692keV]；取能量为[632keV,692keV] 区间的所有闪烁脉冲，假设其数目为m，定义S(k,i)为其中第i(i为整数，且0<i≤m)个脉冲的第k(k为整数，且0<k≤h,h为ADC采样点的数目)个ADC采样点，按照
Figure PCTCN2015092909-appb-000030
得到平均脉冲Pm。
(2)将该平均脉冲进行比例缩放，模拟得到对应[0keV,3MeV]范围内的射线照射探测器产生的闪烁脉冲数据库。假设由步骤(1)得到的平均脉冲Pm对应的射线能量Es＝662keV，待测射线的能量上限Eu＝3MeV,则计算出闪烁脉冲需要放大的最大倍数
Figure PCTCN2015092909-appb-000031
根据实际需求确定缩放脉冲的总数目N＝10000，缩放脉冲集Pa由
Figure PCTCN2015092909-appb-000032
计算得到。
(3)根据步骤(1)得到的平均脉冲，对缩放脉冲集中的每一个闪烁脉冲使用ADC拟合方法计算得到能量值E(ADC-fit,i),i为整数，且0<i≤10000。同时，将缩放脉冲集中的每一个闪烁脉冲ADC采样点的电压幅值直接累加得到传统ADC方案中的能量值E(ADC,i)，其中i为整数，且0<i≤10000。
ADC拟合方法获取闪烁脉冲能量具体可以通过以下过程实现：
使用与步骤(1)中不同的ADC(记作ADC2)重新采样闪烁脉冲信号得到ADC2采样点，设一个闪烁脉冲采样的数据集包括N个样本点(ti,vi)，该方法中闪烁脉冲模型(具体可参见文献Nan Zhang,Niraj Doshi,Mehmet Aykac,Ronald Grazioso,Michael Loope,Greg Givens,Lars Eriksson,Florian Bauer,John  Young,and Matthias Schmand,“A Pulse Shape Restore Method for Event Localizaton in PET Scintillation Detection”,Nuclear Science Symposium Conference Record,Vol.7,2004)可采用：
Figure PCTCN2015092909-appb-000033
   (式2)
其中，A1、m0、τ1、τ0是需要待确定的参数，τ0为晶体闪烁的时间常数，τ1为光电转换器件和前端电子学的时间常数。使用ADC2采样点曲线拟合确定未知参数后，重建闪烁脉冲，并使用定积分或数值积分技术获得闪烁脉冲能量值。
(4)计算校正系数
Figure PCTCN2015092909-appb-000034
建立校正系数查找表，或者通过拟合校正系数建立E(ADC-fit,i)与E(ADC,i)的函数映射关系。
通过步骤(4)得到的查找表或者函数方程得到校正系数C，并计算校正后的能量ECADC-fit＝C*EADC-fit。

Claims (25)

  1. 一种闪烁脉冲的数字化方法，其特征在于：包括步骤：
    S1：获取探测器在不同能量射线照射下输出的脉冲数据库；
    S2：对S1所得脉冲数据库中的每一个脉冲采样并量化以获取脉冲所包含的完全能量信息；
    S3：对S1所得脉冲数据库中的每一个脉冲进行欠采样并量化，使用脉冲先验信息估计或者拟合出能量信息；
    S4：用S2中所得能量信息作为标准，确定基于先验信息的欠采样脉冲能量获取方法和S2方法得到的能量信息的映射关系；
    S5：利用S4所得能量映射关系校正基于先验信息的欠采样脉冲能量获取方法得到的能量信息。
  2. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S1中的探测器为气体探测器或闪烁体探测器或半导体探测器。
  3. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S1中获取探测器在不同能量射线照射下输出的脉冲数据库的方式为仿真产生或/和直接采集。
  4. 根据权利要求3所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S1获取探测器在不同能量射线照射下输出的脉冲数据库的方式若为仿真产生，其具体过程为：
    (a)使用已知能量的射线源照射探测器，利用ADC采集一组或若干组脉冲信号得到脉冲集合，利用该脉冲集合生成平均脉冲；
    (b)将该平均脉冲进行缩放，模拟得到一定能量范围内的射线照射该探测器产生的脉冲数据库。
  5. 根据权利要求4所述的闪烁脉冲的数字化方法，其特征在于：所述生成平均脉冲的方法为：首先做出这组脉冲的能量谱，对能量谱中的光电峰使用 高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
    Figure PCTCN2015092909-appb-100001
  6. 根据权利要求4所述的闪烁脉冲的数字化方法，其特征在于：所述缩放平均脉冲的方法为线性缩放或非线性缩放。
  7. 根据权利要求6所述的闪烁脉冲的数字化方法，其特征在于：线性缩放平均脉冲的方法为：假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
    Figure PCTCN2015092909-appb-100002
    缩放脉冲集合Pa由
    Figure PCTCN2015092909-appb-100003
    计算得到。
  8. 根据权利要求6所述的闪烁脉冲的数字化方法，其特征在于：非线性缩放平均脉冲的方法为：首先确定探测器的非线性能量响应曲线y＝f(x)，假设由步骤S1中得到的平均脉冲Pm对应的射线能量为Es，用户根据实际应用需求确定待测射线的能量上限为Eu，缩放脉冲的总数目为N，计算出脉冲信号需要放大的最大倍数
    Figure PCTCN2015092909-appb-100004
    该探测器对应的缩放脉冲集合Pa则由
    Figure PCTCN2015092909-appb-100005
    确定。
  9. 根据权利要求3所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S1获取探测器在不同能量射线照射下输出脉冲的方式若为直接采集，其具体过程为：首先根据用户需求使用若干个具有不同能量的射线源照射探测器，利用ADC直接采集输出脉冲得到能量分布在一定范围内的脉冲集合。
  10. 根据权利要求3所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S1获取探测器在不同能量射线照射下输出的脉冲的方式若为仿真产生和直接采集的结合，其具体过程为：首先使用多个已知能量的射线源分别照射探测器，在分别得到各自的平均脉冲后，使用插值法模拟得到一定能量范围内的射线照射该探测器产生的脉冲数据库。
  11. 根据权利要求10所述的闪烁脉冲的数字化方法，其特征在于：所述生成平均脉冲的方法为：首先做出这组脉冲库的能量谱，对能量谱中的光电峰 使用高斯函数拟合，并取出能量值位于拟合得到的高斯函数半高全宽范围的所有脉冲，假设其数目为m，定义S(k,i)为其中第i个脉冲的第k个ADC采样点，其中i为整数，且0<i≤m；k为整数，且0<k≤h，h为ADC采样点的数目，平均脉冲Pm定义为集合：
    Figure PCTCN2015092909-appb-100006
  12. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S2中获取脉冲所包含的完全能量信息的方法包括：对脉冲直接数字化以获取脉冲所包含的完全能量信息，或者经过间接采样可代表脉冲能量信息的物理量来获取脉冲所包含的完全能量信息。
  13. 根据权利要求12所述的闪烁脉冲的数字化方法，其特征在于：所述获取脉冲所包含的完全能量信息的方法如果通过脉冲直接数字化的方式实现，具体方法为：使用模数转换方案或者多电压阈值方案临界/过采样脉冲，然后使用数值积分方法获取脉冲能量信息，或使用采样点中最大电压幅值代表其能量信息。
  14. 根据权利要求12所述的闪烁脉冲的数字化方法，其特征在于：所述获取脉冲所包含的完全能量信息的方法如果通过脉冲直接数字化的方式实现，具体方法为：先使用峰值保持电路保持峰值，然后采样峰值点电压幅值来代表脉冲能量信息。
  15. 根据权利要求12所述的闪烁脉冲的数字化方法，其特征在于：所述获取脉冲所包含的完全能量信息的方法如果通过间接采样的方式实现，具体方法为：将脉冲幅值变换成时间宽度的威尔金逊变换法。
  16. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S3中对每一个脉冲进行欠采样并量化的数字化实现方法包括：使用模拟数字转换器实现或者使用比较器配合时间数字转换器实现。
  17. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S3中估计或者拟合出能量信息的方法为MVT方法，利用MVT方法计算脉冲能量的方式为：首先设定多个阈值电压，将脉冲和阈值电压分别输入到比较器的输入端，并测量比较器输出翻转逻辑脉冲的时间，测量得到的时间值和对应的阈值电压组成MVT采样点，利用MVT采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。
  18. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S3中估计或者拟合出能量信息的方法为ADC拟合方法，ADC拟合方法计算脉冲能量的步骤为：使用ADC采样脉冲信号得到ADC采样点，利用ADC采样点和脉冲模型拟合重建脉冲，对重建得到的脉冲做定积分或重采样后做数值积分得到脉冲的能量信息。
  19. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S3中估计或者拟合出能量信息的方法为经验贝叶斯估计法，利用经验贝叶斯估计法计算脉冲能量的具体方式为：在获得脉冲数字化样本点后，利用贝叶斯原理和独立性假设求解给定数字样本的最大似然解，将其作为被估计的脉冲能量信息。
  20. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S3中估计或者拟合出能量信息的方法为TOT方法，利用TOT方法计算脉冲能量的方式为：通过拟合TOT与脉冲能量的关系，估计脉冲的能量预期值作为脉冲的能量信息。
  21. 根据权利要求1所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S5中的能量映射关系是系数查找表或者能量映射函数。
  22. 根据权利要求21所述的闪烁脉冲的数字化方法，其特征在于：若所述步骤S5中的能量映射关系为系数查找表，则校正系数C计算方法为：假设 对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则校正系数
    Figure PCTCN2015092909-appb-100007
  23. 根据权利要求22所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝C*Estatistical。
  24. 根据权利要求21所述的闪烁脉冲的数字化方法，其特征在于：若所述步骤S4中的能量映射关系若为映射函数y＝g(x)，该函数的推导方法为：假设对于脉冲库Pa中的每一个脉冲，完全能量信息获取方法得到的能量为E(standard,i)，基于先验信息获取欠采样脉冲能量获取方法得到的能量为E(statistical,i)，i为整数，且0<i≤N，则可以得到完全能量信息获取方法和基于先验信息获取 欠采样脉冲能量获取方法的若干个系数映射点：
    Figure PCTCN2015092909-appb-100008
    利用这些系数映射点，经过曲线拟合得到映射函数y＝g(x)。
  25. 根据权利要求24所述的闪烁脉冲的数字化方法，其特征在于：所述步骤S5中校正后的能量ECStatistical的计算公式为：ECStatistical＝g(Estatistical)。
PCT/CN2015/092909 2015-01-05 2015-10-27 一种闪烁脉冲的数字化方法 WO2016110141A1 (zh)

Priority Applications (3)

Application Number Priority Date Filing Date Title
EP15876642.8A EP3244235B1 (en) 2015-01-05 2015-10-27 Method for digitalizing scintillation pulse
US15/541,220 US9910167B2 (en) 2015-01-05 2015-10-27 Method for digitalizing scintillation pulse
JP2017535397A JP6473509B2 (ja) 2015-01-05 2015-10-27 シンチレーションパルスデジタル化方法

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510003057.1A CN105824817B (zh) 2015-01-05 2015-01-05 一种闪烁脉冲的数字化方法
CN201510003057.1 2015-01-05

Publications (1)

Publication Number Publication Date
WO2016110141A1 true WO2016110141A1 (zh) 2016-07-14

Family

ID=56355484

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2015/092909 WO2016110141A1 (zh) 2015-01-05 2015-10-27 一种闪烁脉冲的数字化方法

Country Status (6)

Country Link
US (1) US9910167B2 (zh)
EP (1) EP3244235B1 (zh)
JP (1) JP6473509B2 (zh)
CN (1) CN105824817B (zh)
HU (1) HUE049124T2 (zh)
WO (1) WO2016110141A1 (zh)

Families Citing this family (15)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106443760B (zh) * 2016-09-14 2019-02-12 湖北锐世数字医学影像科技有限公司 一种闪烁脉冲简化模型、重建及能量获取的方法
CN107024711B (zh) * 2017-04-17 2019-02-26 苏州瑞派宁科技有限公司 一种闪烁脉冲数字化信号的拟合方法
CN107462917A (zh) * 2017-09-19 2017-12-12 南京瑞派宁信息科技有限公司 一种电离辐射探测装置及方法
CN108646285B (zh) * 2018-06-12 2020-05-22 高超 伽马射线能谱测量方法、装置及系统
CN109350098B (zh) * 2018-08-27 2021-02-26 苏州瑞派宁科技有限公司 信号的拟合方式的确定方法、重建方法和装置
CN109274369B (zh) * 2018-08-29 2022-05-10 苏州瑞迈斯医疗科技有限公司 一种闪烁脉冲数字化的方法及装置
CN109507716B (zh) * 2018-11-16 2020-04-21 江苏赛诺格兰医疗科技有限公司 一种获取闪烁体探测器能量信息的方法
CN110226943B (zh) * 2019-07-05 2023-08-15 上海联影医疗科技股份有限公司 光子到达探测器的参数计算方法、装置和计算机设备
CN110694186B (zh) * 2019-09-20 2020-11-17 华中科技大学 基于硬件高斯拟合粒子束束斑位置的计算机可读存储介质
CN110598348B (zh) * 2019-09-23 2022-08-02 四川大学 一种基于Matlab的核信号产生及处理的仿真方法
CN111142148B (zh) * 2020-01-10 2022-05-20 南昌大学 一种位置敏感型闪烁探测器的倒装sql方法
CN111582023B (zh) * 2020-03-27 2024-06-14 中国科学院高能物理研究所 脉冲信号能量值的确定方法、系统、设备及存储介质
CN111624644A (zh) * 2020-06-11 2020-09-04 北京卫星环境工程研究所 一种三维位置灵敏探测器及其能量校正方法
CN111929720B (zh) * 2020-07-13 2023-04-25 中广核核电运营有限公司 中子探测器性能检测方法、装置、系统和计算机设备
CN116821097B (zh) * 2023-07-05 2024-01-26 中国人民解放军92728部队 全脉冲数据处理方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101561507A (zh) * 2009-05-21 2009-10-21 清华大学 射线探测器的射线能量检测方法
CN102486402A (zh) * 2010-12-01 2012-06-06 中国科学院西安光学精密机械研究所 一种测量脉冲激光能量的方法及系统
CN102843139A (zh) * 2012-09-20 2012-12-26 苏州瑞派宁科技有限公司 一种闪烁脉冲数字化的方法及装置
CN103969675A (zh) * 2013-02-05 2014-08-06 苏州瑞派宁科技有限公司 数字化闪烁脉冲的基线校正方法及系统
US20140318268A1 (en) * 2011-12-02 2014-10-30 Kamstrup A/S Ultrasonic flow meter with digitally under-sampled flow measurements

Family Cites Families (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB1216548A (en) * 1968-02-15 1970-12-23 Mullard Ltd Improvements in or relating to infra-red thermal imaging systems
JP2000131440A (ja) * 1998-10-20 2000-05-12 Toshiba Corp 放射線検出処理システム
DE10045929A1 (de) * 2000-09-15 2002-04-11 Siemens Ag Verfahren zum Betrieb einer Röntgendiagnostikeinrichtung mit Sofortbilddarstellung
US6664543B2 (en) * 2002-01-28 2003-12-16 Cti Pet Systems, Inc. Continuous sampling and digital integration for PET scintillation
CN100585687C (zh) * 2007-06-21 2010-01-27 北京巨数数字技术开发有限公司 脉冲信号产生方法及使用该方法的控制电路和显示系统
EP2232444B1 (en) * 2007-12-20 2011-10-12 Wisconsin Alumni Research Foundation Method for dynamic prior image constrained image reconstruction
CN102262238B (zh) * 2011-04-19 2014-07-23 苏州瑞派宁科技有限公司 一种提取闪烁脉冲信息的方法及装置
US8309933B1 (en) * 2011-06-03 2012-11-13 Kabushiki Kaisha Toshiba Count rate adaptive filter for medical imaging systems
WO2013171644A1 (en) * 2012-05-14 2013-11-21 Koninklijke Philips N.V. Determination of a fractional flow reserve (ffr) value for a stenosis of a vessel

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101561507A (zh) * 2009-05-21 2009-10-21 清华大学 射线探测器的射线能量检测方法
CN102486402A (zh) * 2010-12-01 2012-06-06 中国科学院西安光学精密机械研究所 一种测量脉冲激光能量的方法及系统
US20140318268A1 (en) * 2011-12-02 2014-10-30 Kamstrup A/S Ultrasonic flow meter with digitally under-sampled flow measurements
CN102843139A (zh) * 2012-09-20 2012-12-26 苏州瑞派宁科技有限公司 一种闪烁脉冲数字化的方法及装置
CN103969675A (zh) * 2013-02-05 2014-08-06 苏州瑞派宁科技有限公司 数字化闪烁脉冲的基线校正方法及系统

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3244235A4 *

Also Published As

Publication number Publication date
JP6473509B2 (ja) 2019-02-20
EP3244235B1 (en) 2020-03-04
JP2018506711A (ja) 2018-03-08
US20170357012A1 (en) 2017-12-14
EP3244235A4 (en) 2018-09-12
CN105824817A (zh) 2016-08-03
HUE049124T2 (hu) 2020-09-28
US9910167B2 (en) 2018-03-06
CN105824817B (zh) 2019-05-21
EP3244235A1 (en) 2017-11-15

Similar Documents

Publication Publication Date Title
WO2016110141A1 (zh) 一种闪烁脉冲的数字化方法
CN103969675B (zh) 数字化闪烁脉冲的基线校正方法及系统
JP6657398B2 (ja) パルスパイルアップイベントのリアルタイム処理の方法及びシステム
JP2014516411A (ja) シンチレーションパルス情報取得のための方法及び装置
Xi et al. Fpga-only mvt digitizer for tof pet
CN109581461B (zh) 核脉冲能量测量方法及系统
WO2014002022A2 (en) Spectral photon counting detector
EP3073292B1 (en) Method and system for recovering scintillation pulse information
CN104656115B (zh) 一种时间标记组合的方法与系统
Wang et al. Advantages of digitally sampling scintillation pulses in pileup processing in PET
CN106706127B (zh) 基于SiPM的多光子探测方法
CN202177701U (zh) 一种提取闪烁脉冲信息的装置
Boorboor et al. Development of a novel approach for precise pulse height extraction using Lagrange interpolation
Faisal et al. A data processing system for real-time pulse processing and timing enhancement for nuclear particle detection systems
CN111624645B (zh) 双端读出晶体探测器及其作用深度刻度表快速生成方法
CN113729751B (zh) 一种示踪剂活度的检测方法、设备及pet扫描仪
CN115420226B (zh) 基于脉冲宽度的射线作用位置定位装置及定位方法
CN111789623A (zh) 一种pet信号的全带压缩感知方法
CN111948695A (zh) 一种pet信号的事件属性计算方法
CN115420226A (zh) 基于脉冲宽度的射线作用位置定位装置及定位方法
Qiu et al. Mathematical Considerations in Energy Spectrum Recovery for Digital Energy Spectrometers
Ianakiev et al. Neutron and gamma discrimination by compensation of long components in liquid scintillators
Liu et al. Enhanced MVT digitizer in PET
Garbino et al. A pulse modeling tool for PET scanners
CN111736207A (zh) 一种pet时间标定方法

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 15876642

Country of ref document: EP

Kind code of ref document: A1

ENP Entry into the national phase

Ref document number: 2017535397

Country of ref document: JP

Kind code of ref document: A

WWE Wipo information: entry into national phase

Ref document number: 15541220

Country of ref document: US

NENP Non-entry into the national phase

Ref country code: DE

REEP Request for entry into the european phase

Ref document number: 2015876642

Country of ref document: EP