WO2014121548A1 - 数字化闪烁脉冲的基线校正方法及系统 - Google Patents

数字化闪烁脉冲的基线校正方法及系统 Download PDF

Info

Publication number
WO2014121548A1
WO2014121548A1 PCT/CN2013/073126 CN2013073126W WO2014121548A1 WO 2014121548 A1 WO2014121548 A1 WO 2014121548A1 CN 2013073126 W CN2013073126 W CN 2013073126W WO 2014121548 A1 WO2014121548 A1 WO 2014121548A1
Authority
WO
WIPO (PCT)
Prior art keywords
baseline
scintillation
pulse
scintillation pulse
module
Prior art date
Application number
PCT/CN2013/073126
Other languages
English (en)
French (fr)
Inventor
谢庆国
陈源宝
朱俊
吴中毅
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
Publication of WO2014121548A1 publication Critical patent/WO2014121548A1/zh

Links

Classifications

    • GPHYSICS
    • G01MEASURING; TESTING
    • G01TMEASUREMENT OF NUCLEAR OR X-RADIATION
    • G01T1/00Measuring X-radiation, gamma radiation, corpuscular radiation, or cosmic radiation
    • G01T1/16Measuring radiation intensity
    • G01T1/20Measuring radiation intensity with scintillation detectors
    • G01T1/208Circuits specially adapted for scintillation detectors, e.g. for the photo-multiplier section

Definitions

  • the scintillation pulses collected and processed by the data acquisition system are usually converted into visible light by scintillation crystals, such as X-rays, gamma rays, etc., and then by optoelectronic devices. It is converted into an electrical pulse signal that can be measured.
  • a typical scintillation pulse waveform in the prior art is shown in FIG.
  • the scintillation pulse is often superimposed on an unstable baseline due to factors such as detector leakage current, flicker pulse tailing, and noise interference.
  • PET Positron Emission Tomography
  • baseline correction is often closely related to the architecture of the front-end electronics circuitry where the data acquisition system is located.
  • One of the most widely used baseline correction methods in traditional analog or analog-digital hybrid circuits is the baseline restorer.
  • Robinson first proposed a baseline restorer for unipolar signals [LB Robinson, "Reduction of baseline shift in pulse-amplitude measurements”, Rev. Sci. Instrum., Vol. 32, p. 1057, 1961]. Subsequently Chase [RL Chase and LR Poulo, "A high precision DC restorer", IEEE Trans. Nucl. Sci., vol. NS-14, no. 6, pp. 83-88, Dec. 1967] , Fairstein [E. Fairstein, "Gated baseline Restorer with adjustable asymmetry", IEEE Trans. Nucl.
  • the methods for digitizing the scintillation pulse are mainly: Equal-interval sampling method based on Analog-to-Digital Converter (ADC) and Multi-Voltage Threshold (Multi-Voltage Threshold) Time axial sampling method called MVT).
  • ADC Analog-to-Digital Converter
  • MVT Multi-Voltage Threshold
  • the former uses a fast ADC to sample the scintillation pulse signal and the baseline at equal intervals, and obtain pulse signal and baseline voltage amplitude sample points, wherein the pulse signal amplitude can be used for pulse information extraction, and the baseline sample points can be averaged or statistically analyzed for baseline correction.
  • the MVT-based time-axis sampling method is a known threshold reference voltage to obtain the time information of the sampling point, and then extracts the event information of the scintillation pulse through a digital signal processing algorithm [Qingguo Xie, Chien- Min Kao, Zekai Hsiau et al""A new approach for pulse processing in positron emission tomography", IEEE Trans. Nucl. Sci., vol. 52, no.
  • MVT sampling method can effectively break through Shannon sampling The limitation of the theorem, accurate and efficient acquisition of event information of scintillation pulses [Qingguo Xie, Chien-Min Kao, Xi Wang et al, "Potentials of digitally sampling scintillation pulses in timing determination in pet", Vol. 56, no. 5, Oct. 2009].
  • MVT sampling method it will also suffer from the interference of baseline drift to the acquisition of scintillation pulse information, and the existing baseline correction methods can not effectively solve the baseline drift problem of MVT.
  • an object of the present invention is to provide a baseline correction method and system for digitizing scintillation pulses for solving high-energy physical particle detection and front-end electronics baseline drift in medical imaging equipment.
  • the method performs multi-threshold voltage sampling on the scintillation pulse, utilizes the prior knowledge of the scintillation pulse, and uses the innovative digital signal processing algorithm to effectively and accurately predict the baseline drift caused by the detector leakage current, pulse tailing and noise interference. And fast adaptive correction, while improving the signal-to-noise ratio of the scintillation pulse data measurement results and the stability of the data acquisition system.
  • the present invention provides the following technical solutions:
  • a baseline correction method for digitized scintillation pulses the steps of the baseline correction method are as follows:
  • step (3) Subtracting the baseline flicker mean obtained in step (2) from a number of reconstructed scintillation pulse waveforms to complete the baseline correction.
  • a plurality of threshold pulse sampling methods are used to perform time axial sampling on a plurality of scintillation pulse signals, and the specific steps of the multi-threshold voltage sampling method are :
  • a plurality of threshold reference voltages are preset, and the specific moments when the voltage amplitudes of the respective scintillation pulses reach the respective threshold reference voltages are recorded, and each of the threshold reference voltages and the corresponding trigger time constitutes a time-threshold sampling point.
  • the shape characteristic of the scintillation pulse waveform in the step (2) is obtained by constructing a pulse model according to the type of the scintillation crystal and the photoelectric conversion device coupled, and the pulse is constructed.
  • the model is expressed as a set of function expressions that characterize the shape of the scintillation pulse.
  • the identifying the model parameter in the step (2) refers to performing function approximation on the time-threshold sampling point of the selected scintillation pulse waveform according to the shape characteristic of the scintillation pulse waveform.
  • a mathematical model is characterized, and parameter values characterizing the characteristics of a single scintillation pulse are extracted therefrom.
  • the scintillation pulse baseline offset in the step (2) is obtained by finding an amount describing the baseline level from the extracted parameter values of the scintillation pulse characteristic, and obtaining a scintillation pulse.
  • the specific steps of the baseline offset include: adding a feature quantity capable of indicating a baseline drift in a model parameter of the reconstructed scintillation pulse waveform, and using a function approximation method to obtain a single pulse by using a plurality of digitized sampling points
  • the parameter value, and the parameter value representing the baseline drift characteristic is used as the baseline drift of the scintillation pulse.
  • the method for analyzing the baseline offset of each scintillation pulse obtained in the step (2) is a mean calculation or a statistical analysis.
  • a baseline correction system for digitizing scintillation pulses comprising:
  • the digital sampling module is configured to perform time axial sampling on the scintillation pulse signal by using a multi-threshold voltage sampling method, and obtain a digitized sampling point of the scintillation pulse waveform corresponding to the scintillation pulse signal;
  • a baseline offset calculation module is configured to reconstruct a scintillation pulse waveform and identify a model parameter according to a shape characteristic of the scintillation pulse waveform and the selected digitized sampling point, obtain a baseline drift amount of the scintillation pulse, and then analyze the obtained baseline drift amount, Obtain a baseline drift mean;
  • a baseline correction module is used to perform baseline correction on the scintillation pulse to restore the raw data information of the scintillation pulse.
  • the digitized sampling module includes a threshold voltage setting module, a threshold discriminator module, and a time stamping module, where
  • the threshold voltage setting module is configured to set a threshold reference voltage, and send the threshold reference voltage to the threshold discriminator module and the time stamping module;
  • the threshold discriminator module is configured to compare the magnitude relationship between the flicker pulse threshold voltage and the threshold reference voltage, and generate a logic pulse when the flicker pulse voltage crosses the threshold reference voltage, and send the generated logic pulse to the time stamping module for time Standard
  • the time stamping module is configured to time mark the logic pulse output by the threshold discriminator module, and combine the resulting time stamp with its corresponding threshold reference voltage to form a time-threshold sampling point and transmit it to the baseline offset calculation module.
  • the baseline offset calculation module comprises an event stack rejection module, a pulse reconstruction module and a baseline offset calculation module, wherein the event stack rejection module is used for identifying and rejecting the flicker a stacking event in a pulse;
  • a pulse reconstruction module is configured to reconstruct the scintillation pulse waveform, identify a model parameter, and transmit the reconstruction parameter value to a baseline offset calculation module;
  • the baseline offset calculation module calculates a baseline offset of the scintillation pulse according to the reconstruction parameters acquired by the pulse reconstruction module, and then performs a statistical histogram analysis on the baseline offset over a period of time to obtain an average baseline offset of the scintillation pulse and Transfer to the baseline correction module.
  • the method of analyzing the baseline offset of each of the obtained scintillation pulses is a mean calculation or a statistical analysis.
  • the baseline correction method of the digital scintillation pulse performs multi-threshold voltage sampling on the scintillation pulse, utilizes the prior knowledge of the scintillation pulse, and uses an innovative digital signal processing algorithm to The baseline drift caused by leakage current, pulse tailing and noise interference is effectively, accurately and quickly adaptively corrected, and the signal-to-noise ratio of the scintillation pulse data measurement result and the stability of the data acquisition system are improved.
  • 1 is a schematic diagram of a scintillation pulse signal that is common in the prior art
  • 2 is a flow chart of a method for correcting a baseline of a digital scintillation pulse according to the present invention
  • FIG. 3 is a result of analyzing a baseline offset value by using a statistical distribution method in a digital scintillation pulse baseline correction method according to the present invention
  • Fig. 4 is a view showing the effect of correcting the forward drift of the scintillation pulse baseline by the digital scintillation pulse baseline correction method of the present invention.
  • FIG. 5 is a schematic diagram of sampling and pulse shape fitting of a flicker pulse output by a LYSO/PMT detector using a 4-threshold MVT sampling method in a digital scintillation pulse baseline correction method according to the present invention
  • FIG. 6 is a diagram of a digital scintillation pulse baseline correction method according to the present invention
  • FIG. 7 is a system structural diagram of a digital scintillation pulse baseline correction system according to the present invention.
  • Fig. 8 is a view showing the effect of energy resolution in the present invention, wherein 8 (a) is an effect diagram of the energy resolution obtained without using the present invention, and 8 (b) is an effect diagram of the energy resolution obtained by using the present invention.
  • the present invention discloses a baseline correction method and system for digitizing scintillation pulses for solving the problem of baseline drift of front-end electronics systems in the field of high-energy physical particle detection and medical imaging equipment.
  • the method performs multi-threshold voltage sampling on the scintillation pulse, utilizes the prior knowledge of the scintillation pulse, and uses the innovative digital signal processing algorithm to effectively and accurately predict the baseline drift caused by the detector leakage current, pulse tailing and noise interference.
  • fast adaptive correction while improving the signal-to-noise ratio of the scintillation pulse data measurement results and the stability of the data acquisition system.
  • the baseline correction method for the digitized scintillation pulse disclosed by the present invention includes the following steps:
  • step (3) Subtracting the baseline flicker mean obtained in step (2) from a number of reconstructed scintillation pulse waveforms to complete the baseline correction.
  • the scintillation pulse in the step (1) is specifically a pulse signal having the same shape and different size determined by the scintillation crystal, the light guide and the photoelectric conversion device.
  • the specific steps of the multi-threshold voltage sampling method in the step (1) are: presetting a plurality of threshold reference voltages, and recording specific moments of voltage amplitudes of the respective scintillation pulses reaching respective threshold reference voltages, each threshold reference voltage and corresponding The trigger time constitutes a time-threshold sample point.
  • the shape characteristic of the scintillation pulse waveform in the step (2) is obtained by constructing a pulse model according to the type of the scintillation crystal and the photoelectric conversion device coupled, and the pulse model is formed in the form of a set of scintillation pulse shapes. Function expression.
  • the identifying the model parameter in the step (2) refers to performing a function approximation to obtain a characteristic mathematical model by approximating the time-threshold sampling point of the selected scintillation pulse waveform according to the shape characteristic of the scintillation pulse waveform, and extracting and characterizing the single scintillation pulse characteristic therefrom The parameter value.
  • the scintillation pulse baseline offset in the step (2) is obtained by finding an amount describing the baseline level from the extracted parameter values of the scintillation pulse characteristic
  • the specific steps of obtaining the scintillation pulse baseline offset include: reconstructing The model parameters of the flicker pulse waveform are added with a feature quantity capable of indicating the baseline drift.
  • a function approximation method is used to obtain a parameter value capable of characterizing a single pulse, and a parameter value indicating a baseline drift characteristic is obtained. The amount of baseline drift as a scintillation pulse.
  • FIG. 3 is a result of analyzing a baseline offset value by using a statistical analysis method in the digital scintillation pulse baseline correction method of the present invention.
  • FIG. 4 is a diagram showing the effect of correcting the baseline drift of the scintillation pulse by the digital scintillation pulse baseline correction method of the present invention.
  • the solid line 11 is the flashing output of the LYSO/PMT detector with an oscilloscope.
  • the dotted line 12 is a waveform obtained by fitting the scintillation pulse model under the baseline correction method proposed by the present invention, and it can be seen that the waveform after fitting has a significant baseline offset; 13 is to use the baseline correction method proposed by the present invention, according to the waveform obtained by the scintillation pulse model fitting, it can be seen that there is no drift at the baseline at this time.
  • the coupling of the various types of forms may have different pulse shape representations.
  • the following is a detailed explanation of the pulse shape characterization form.
  • S101 acquiring a corresponding scintillation pulse shape characteristic model according to the coupled scintillation crystal and the photoelectric conversion device category;
  • the mathematical model of the scintillation pulse shape can be considered to be composed of a rising edge of a rising line and a falling edge of an exponential decay, regardless of the influence of noise. Shown as follows:
  • an ideal scintillation pulse can be derived from five model eigenvalues Z we , LineB r , ExpK f , ExpB f ⁇ BaseL ⁇ , the start time, peak time, peak amplitude, afterglow constant, and baseline value of the flashing eternal signal
  • the information can be calculated from the five model eigenvalues. The formula is as follows:
  • V p LineK r xt p + LineB r ;
  • Baseline value ⁇ ⁇ ⁇ BaseL.
  • S102 selecting digitized sampling points corresponding to each single scintillation pulse, sequentially reconstructing the pulse waveform according to the shape model of the scintillation pulse in step S101, and identifying the model parameters, thereby estimating the baseline drift amount of each single scintillation pulse;
  • V(t) LineK r xt + LineB r
  • the parameter is the slope of the rising edge line and ⁇ >0
  • the parameter Z « e is the intercept of the rising edge line, which can be any value; it is the time value obtained by the MVT sampling method when the threshold reference voltage is V (0).
  • the method can also reconstruct the rising edge of the scintillation pulse according to other scintillation pulse models.
  • V (t) Qxp(-ExpK f xt + ExpB f ) + BaseL
  • the parameter Exp ⁇ is the decay time constant and Exp X), the parameter Ex/ ⁇ can be any value; the parameter ZteeL is the baseline parameter of the scintillation pulse, which can be any value; the time obtained by the MVT sampling method when the threshold reference voltage is value.
  • the method can also reconstruct the falling edge of the flicker pulse according to other scintillation pulse models.
  • FIG. 5 is a schematic diagram of sampling and pulse shape fitting of a flicker pulse output by a LYSO/PMT detector by using a 4-threshold MVT sampling method in a digital scintillation pulse baseline correction method.
  • the solid line 21 represents a waveform reconstructed by the oscilloscope after sampling the output stroboscopic pulse of the LYSO/PMT detector;
  • the dot 22 is a sampling point obtained by sampling the scintillation pulse at the set reference threshold voltage by the MVT method;
  • 23 is a waveform obtained by fitting a linear rising edge and an exponential falling edge along the scintillation pulse model according to the present invention. It can be seen from the figure that the waveform obtained by the scintillation pulse model can better approximate the waveform reconstructed by the oscilloscope sampling, and also confirms that the coupling of different scintillation crystals and photoelectric conversion device categories needs to obtain the corresponding scintillation pulse. model.
  • V(t) A r x xp(-ExpK r xt) + B r t s ⁇ t ⁇
  • parameter A is the amplitude coefficient of the rising edge of the exponent
  • E ⁇ is the time constant of the rising edge of the exponent
  • the parameter can be any value
  • the parameter ⁇ is the amplitude coefficient of the exponential falling edge
  • the parameter Ep ⁇ is the time constant of the exponential falling edge and Ep ⁇ X)
  • the parameter is the baseline parameter of the scintillation pulse, which can be any value
  • s is the starting time of the scintillation pulse
  • is the peak time of the scintillation pulse.
  • an ideal scintillation pulse can be described by six model eigenvalues A, ExpK r , B r , A f , ExpK f and « ⁇ , the start time, peak time, peak amplitude and baseline value of the scintillation pulse signal.
  • the information can be calculated from the eigenvalues of the six models. The formula is as follows:
  • V p A r X Qxp(-ExpK r xt p ) + B r ;
  • S202 selecting digitized sampling points corresponding to each single scintillation pulse, sequentially reconstructing the pulse waveform according to the shape model of the scintillation pulse in step 201, and identifying the model parameters, thereby estimating the baseline drift amount of each single scintillation pulse;
  • V(t) A r x exp(-ExpK r xt) + B r
  • parameter A is the amplitude coefficient of the rising edge of the exponent
  • E ⁇ is the time constant of the exponential rising edge JLE w K r >0
  • the parameter can It is an arbitrary value; it is the time value obtained by the MVT sampling method when the threshold reference voltage is v (0). This method can also reconstruct the rising edge of the scintillation pulse according to other scintillation pulse models.
  • V(t) A f XQxp(-ExpK f xt) + BaseL
  • parameter A is the magnitude coefficient of the exponential falling edge and parameter Ep ⁇ is the time of the exponential falling edge Constant and E p X);
  • the parameter ZteeL is the baseline parameter of the scintillation pulse, which can be any value; it is the time value obtained by the MVT sampling method when the threshold reference voltage is V (this method can also be based on other scintillation pulse models, The falling edge of the scintillation pulse is reconstructed.
  • FIG. 6 is a schematic diagram of digital sampling and pulse shape fitting of the flicker pulse outputted by the LYSO/SiPM detector by using the 4-threshold MVT sampling method.
  • the solid line 31 is a waveform reconstructed by sampling the scintillation pulse outputted by the LYSO/SiPM detector with an oscilloscope;
  • the dot 32 is a scintillation pulse outputted by the LYSO/SiPM detector at the set reference threshold voltage by the MVT method.
  • the sampled points are sampled;
  • the dashed line 33 is a waveform obtained by fitting the sampling points by the exponential rising edge and the exponential falling edge according to the present invention.
  • the waveform obtained by the scintillation pulse model can better approximate the waveform reconstructed by the oscilloscope sampling, and also confirms that the coupling of different scintillation crystals and photoelectric conversion device categories needs to obtain the corresponding scintillation pulse. model.
  • the baseline correction system for digital scintillation pulses disclosed in the present invention includes: a digitization sampling module 100, configured to perform time axial sampling on a scintillation pulse signal by using a multi-threshold voltage sampling method, and obtain a corresponding scintillation pulse signal. a digitized sampling point of a scintillation pulse waveform;
  • the baseline offset calculation module 200 is configured to reconstruct a scintillation pulse waveform and identify a model parameter according to a shape characteristic of the scintillation pulse waveform and the selected digitized sampling point, obtain a baseline drift amount of the scintillation pulse, and then analyze the obtained baseline drift amount. , obtaining a baseline drift mean;
  • the baseline correction module 300 is configured to perform baseline correction on the scintillation pulse and restore original data information of the scintillation pulse.
  • the digitized sampling module 100 is divided into three sub-modules, namely a threshold voltage setting module 110, a threshold discriminator module 120, and a time stamping module 130.
  • the threshold voltage setting module 110 is configured to set the threshold reference voltage and send the threshold reference voltage to the threshold discriminator module 120 and the time stamping module 130.
  • the threshold discriminator module 120 is configured to compare the flicker pulse threshold voltage with the threshold reference voltage The size relationship between the two, and a logic pulse is generated when the flicker pulse voltage crosses the threshold reference voltage, and the generated logic pulse is sent to the time stamping module 130 for time marking.
  • the time stamping module 130 is configured to time stamp the logic pulse output by the threshold discriminator module 120, and form the resulting time stamp with its corresponding threshold reference voltage to form a time-threshold sampling point and transmit it to the baseline offset calculating module 200.
  • the baseline offset calculation module 200 includes an event stack culling module 210, a pulse reconstruction module 220, and a baseline offset calculation module 230.
  • the event stack culling module 210 is configured to identify and reject stacking events in the scintillation pulse.
  • the pulse reconstruction module 220 is configured to reconstruct the scintillation pulse waveform, identify model parameters, and transmit the reconstruction parameter values to the baseline offset calculation module 230.
  • the baseline offset calculation module 230 calculates a baseline offset of the scintillation pulse according to the reconstruction parameter acquired by the pulse reconstruction module 220, and then performs a statistical histogram analysis on the baseline offset within a period of time to obtain an average baseline of the scintillation pulse.
  • the offset is passed to the baseline correction module 300.
  • the method of analyzing the baseline offset of each of the obtained scintillation pulses is a mean calculation or a statistical analysis.
  • the method and system for correcting the baseline of the digitized scintillation pulse of the present invention are further verified by a specific embodiment data, wherein several parameters are involved, and these parameters need to be adjusted for specific processing data to achieve good performance.
  • the application instance processes the parameters of the data.
  • step (1) six threshold reference voltages are set in the multi-threshold reference voltage sampling method, and the specific voltage amplitudes are 1.5mV, 21.5mV, 41.5 mV, 61.5 mV, 81.5, 101.5, respectively;
  • the imported digital pulse is a single LYSO (Strontium silicate scintillation crystal)
  • the crystal strip and the Hammatsu R9800 PMT photomultiplier tube) are coupled to the scintillation pulse collected.
  • the sampling rate is 20 GSps, the number of pulse samples is 3000, and the sample points of each pulse waveform are 4000.
  • the energy spectrum is in the range of 511 keV, the average pulse rise time is about Ins, and the detector decay time constant is about 47ns.
  • the scintillation pulse in the method example 1 in step (2) is made up of LYSO crystal strips with Hamamatsu
  • the R9800 PMT coupling produces a scintillation pulse model consisting of a linear rising edge and an exponential decay falling edge;
  • the scintillation pulse in the second example of the method in step (2) is made up of LYSO crystal strips and SensL Array4
  • the SiPM coupling produces a scintillation pulse model consisting of an exponential rising edge and an exponential decay falling edge.
  • the baseline drift of 3000 scintillation pulses output by the LYSO/PMT detector is statistically analyzed.
  • the baseline offset is -10mv;
  • FIG. 8 shows the correction of the energy resolution by the method of the present invention.
  • (a) The energy resolution obtained when the baseline correction is not performed under the MVT sampling method is 22.3%, and (b) is MVT.
  • the energy resolution obtained by baseline correction under the sampling method was 18.1%.

Landscapes

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

Abstract

一种数字化闪烁脉冲的基线校正方法及系统。该方法首先使用多阈值电压采样法,对闪烁脉冲信号进行时间轴向采样,获取闪烁脉冲波形的时间-阈值采样点;然后选取这些时间-阈值采样点,利用闪烁脉冲的形状特性来重建脉冲波形及辨识模型参数,据此估计闪烁脉冲的基线漂移均值;最后根据该基线漂移均值对闪烁脉冲波形进行基线校正。本发明还公开了一种数字化闪烁脉冲的基线校正系统,包括数字采样模块、基线偏移计算模块和基线校正模块。本发明能够有效、准确地甄别闪烁脉冲的基线漂移,实现对基线的快速实时自适应校正,提高闪烁脉冲数据测量结果的信噪比和数据获得系统的稳定性。

Description

数字化闪烁脉沖的基线校正方法及系统
本申请要求于 2013 年 2 月 5 日提交中国专利局、 申请号为 201310045576.5、 发明名称为 "数字化闪烁脉沖的基线校正方法及系统 "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 本发明属于高能物理探测器和信号处理领域,具体涉及一种数字化闪烁脉 沖的基线校正方法及系统, 可应用于高能粒子探测及医疗影像设备。 背景技术
在大多数的高能粒子探测领域以及医疗影像设备中,数据获得系统采集和 处理的闪烁脉沖通常是由闪烁晶体将高能粒子(如: X射线、 γ射线等)转换 成可见光, 然后再由光电器件转换成可以进行测量的电脉沖信号,现有技术中 典型的闪烁脉沖波形如图 1所示。在对闪烁脉沖的信息提取中, 由于受到探测 器漏电流、 闪烁脉沖拖尾、 噪声干扰等因素的影响, 闪烁脉沖往往是叠加在一 个不稳定的基线上。这样不仅会造成闪烁脉沖到达时间和脉沖高度测量的不精 确, 降低原始数据的信噪比, 同时也会对晶体分割、 随机事件和散射事件的剔 除产生间接影响, 尤其是对正电子发射断层成像仪 (Positron Emission Tomography, 以下筒称 PET ) 重建图像的分辨率、 对比度以及信噪比造成较 大影响。 因此, 为了减少基线漂移对辐射探测系统稳定性和性能指标的影响, 有必要在提取闪烁脉沖信息之前进行基线校正。
基线校正的实现方式往往与数据获得系统所在前端电子学电路的架构紧密 相关。 传统的基于模拟电路或者模拟-数字混合电路结构中, 应用最为广泛的 一种基线校正方法是基线恢复器。 Robinson最早提出了一种针对单极信号的基 线恢复器 [L. B. Robinson, "Reduction of baseline shift in pulse-amplitude measurements", Rev. Sci. Instrum., Vol. 32, p. 1057, 1961]。 随后 Chase [R. L. Chase and L. R. Poulo, "A high precision DC restorer", IEEE Trans. Nucl. Sci., vol. NS-14, no. 6, pp. 83-88, Dec. 1967] , Fairstein [E. Fairstein, "Gated baseline restorer with adjustable asymmetry", IEEE Trans. Nucl. Sci., vol. NS-22, no. 1, pp. 463-466, Feb. 1975]以及 Kuwata [M. Kuwata, H. Maeda, and K. Husimi, "New baseline restorer based on feedforward differential compensation", IEEE Trans. Nucl. Sci., vol. 41, no. 4, pp. 1236-1239, Aug. 1994] 等研究组相继提出了各种改 进型的基线恢复器,以提高基线校正的效率。另夕卜, Geronimo [G. De. Geronimo, P. O'Connor, and J. Grosholz, "A CMOS baseline holder (BLH) for readout ASICs", IEEE Trans. Nucl. Sci., vol. 47, no. 3, Jim. 2000] 也提出了一种可供选 择的基于 CMOS工艺的基线保持器。 尽管这些电路在具体实现细节和功能特 性上各不相同,但是它们都是依赖于模拟技术而发展起来的, 这些电路的设计 优化往往是针对某一特定的探测器结构而进行的,一旦设计完成就无法根据应 用需求而改变, 灵活性、 扩展性和升级性受到了极大的制约。
随着各种通用数字化设备和电子器件的普及以及数字信号处理算法的广泛 应用, 数字化技术正被越来越多的引入到对闪烁脉沖的信息提取中。 目前, 对 闪烁脉沖进行数字化处理的方法主要有: 基于模拟 -数字转换器 ( Analog-to-Digital Converter, 以下筒称 ADC )的等间隔采样法和基于多阈值 电压 ( Multi- Voltage Threshold, 以下筒称 MVT ) 的时间轴向采样法。 前者利 用快速 ADC对闪烁脉沖信号及基线进行等间隔采样, 获取脉沖信号及基线的 电压幅度样本点, 其中脉沖信号幅度可用于脉沖信息提取,基线样本点可进行 均值处理或者统计分析以进行基线校正 [Hongdi Li, Chao Wang, Hossain Baghaei et al, "A new statistics-based online baseline restorer for a high count rate fully digital system", IEEE Trans. Nucl. Sci., vol. 57, no. 2, Apr. 2010]。 与基于 ADC的等间隔采样不同, 基于 MVT的时间轴向采样法是已知阈值参考电压, 来获取采样点的时间信息,然后通过数字信号处理算法提取闪烁脉沖的事件信 息 [Qingguo Xie, Chien-Min Kao, Zekai Hsiau et al" "A new approach for pulse processing in positron emission tomography", IEEE Trans. Nucl. Sci., vol. 52, no. 4, Aug. 2005]。 MVT采样方法能够有效突破香农采样定理的限制,精确有效地获 取闪烁脉沖的事件信息 [Qingguo Xie, Chien-Min Kao, Xi Wang et al, "Potentials of digitally sampling scintillation pulses in timing determination in pet", vol. 56, no. 5, Oct. 2009]。 对于 MVT采样方法, 其同样会受到基线漂移对闪烁 脉沖信息获取的干扰问题, 而目前现有的基线校正方法均无法有效地解决 MVT的基线漂移问题。
因此,针对上述技术问题,有必要提出一种新的数字化闪烁脉沖的基线校 正方法及系统, 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数字化闪烁脉沖的基线校正方法及 系统, 用于解决高能物理粒子探测及医疗影像设备中前端电子学基线漂移问 题。 该方法通过对闪烁脉沖进行多阈值电压采样, 利用闪烁脉沖的先验知识, 运用创新的数字信号处理算法,对由探测器漏电流、脉沖拖尾及噪声干扰等引 起的基线漂移进行有效、 准确和快速地自适应校正, 同时提高闪烁脉沖数据测 量结果的信噪比和数据获得系统的稳定性。
为实现上述目的, 本发明提供如下技术方案:
一种数字化闪烁脉沖的基线校正方法, 所述基线校正方法步骤如下:
( 1 )对若干闪烁脉沖信号进行时间轴向采样, 获取若干闪烁脉沖信号对 应的闪烁脉沖波形的数字化采样点;
( 2 )选取各个闪烁脉沖波形的至少两个数字化采样点, 根据闪烁脉沖波 形的形状特性及选取的数字化采样点,重建出该些闪烁脉沖波形并辨识重建出 的闪烁脉沖波形的模型参数,据此获得各个闪烁脉沖波形的基线漂移量, 然后 对所得的基线漂移量进行分析, 获得基线漂移均值;
( 3 )将若干重建后的闪烁脉沖波形减去步骤( 2 )中获得的基线漂移均值, 完成基线校正。
优选的, 在上述数字化闪烁脉沖的基线校正方法中, 所述步骤(1 ) 中, 采用多阈值电压采样法对若干闪烁脉沖信号进行时间轴向采样,所述多阈值电 压采样法的具体步骤为: 预先设置若干阈值参考电压,记录各个闪烁脉沖的电 压幅度达到各个阈值参考电压的具体时刻,每个阈值参考电压和对应的触发时 间即组成一个时间-阈值采样点。 优选的, 在上述数字化闪烁脉沖的基线校正方法中, 所述步骤(2 ) 中闪 烁脉沖波形的形状特性根据所耦合的闪烁晶体和光电转换器件的类别进行建 脉沖模型获得,所建的该脉沖模型的表现形式为一组可以刻画闪烁脉沖形状的 函数表达式。
优选的, 在上述数字化闪烁脉沖的基线校正方法中, 所述步骤(2 ) 中辨 识模型参数是指根据闪烁脉沖波形的形状特性,对所选取闪烁脉沖波形的时间 -阈值采样点进行函数逼近得到特征数学模型, 并从中提取刻画单个闪烁脉沖 特性的参数值。
优选的, 在上述数字化闪烁脉沖的基线校正方法中, 所述步骤(2 ) 中闪 烁脉沖基线偏移量是从提取的刻画闪烁脉沖特性的参数值中找到描述基线水 平的量得到, 获得闪烁脉沖基线偏移量的具体步骤包括: 在重建出的闪烁脉沖 波形的模型参数中加入能够表示基线漂移的特征量, 通过多个数字化的采样 点, 利用函数逼近的方法, 求得能够刻画单个脉沖的参数值, 并将表示基线漂 移特征的参数值作为闪烁脉沖的基线漂移量。
优选的, 在上述数字化闪烁脉沖的基线校正方法中, 所述步骤(2 ) 中对 所得到的各个闪烁脉沖的基线偏移量进行分析的方法为均值计算或统计分析。
一种数字化闪烁脉沖的基线校正系统, 其包括:
数字化采样模块,用于使用多阈值电压采样方法对闪烁脉沖信号进行时间 轴向采样, 获取闪烁脉沖信号对应的闪烁脉沖波形的数字化采样点;
基线偏移计算模块,用于根据闪烁脉沖波形的形状特性及选取的数字化采 样点, 重建出闪烁脉沖波形并辨识模型参数, 获得闪烁脉沖的基线漂移量, 然 后对所得的基线漂移量进行分析, 获得基线漂移均值;
基线校正模块, 用于对闪烁脉沖进行基线校正,还原闪烁脉沖的原始数据 信息。
优选的,在上述数字化闪烁脉沖的基线校正系统中, 所述数字化采样模块 包括阈值电压设置模块、 阈值甄别器模块和时间标记模块, 其中,
阈值电压设置模块用于设定阈值参考电压,并将阈值参考电压送到阈值甄 别器模块和时间标记模块; 阈值甄别器模块用于比较闪烁脉沖阈值电压与阈值参考电压之间的大小 关系, 并在闪烁脉沖电压穿过阈值参考电压时产生逻辑脉沖, 并将产生的逻辑 脉沖送到时间标记模块进行时间打标;
时间标记模块用于对阈值甄别器模块输出的逻辑脉沖进行时间标记,并将 所得的时间戳与其相应的阈值参考电压组成时间-阈值采样点并传送到基线偏 移计算模块。
优选的,在上述数字化闪烁脉沖的基线校正系统中, 所述基线偏移计算模 块包括事件堆积剔除模块、 脉沖重建模块和基线偏移量计算模块, 其中, 事件堆积剔除模块用于鉴别及剔除闪烁脉沖中的堆积事件;
脉沖重建模块用于重建所述闪烁脉沖波形, 辨识模型参数, 并将重建参数 值传送到基线偏移量计算模块;
基线偏移量计算模块根据脉沖重建模块获取的重建参数计算闪烁脉沖的 基线偏移量, 然后对一段时间范围内的基线偏移量进行统计直方图分析, 获得 闪烁脉沖的平均基线偏移量并传送到基线校正模块。
优选的,在上述数字化闪烁脉沖的基线校正系统中,对所得到的各个闪烁 脉沖的基线偏移量进行分析的方法为均值计算或统计分析。
从上述技术方案可以看出,本发明实施例的数字化闪烁脉沖的基线校正方 法通过对闪烁脉沖进行多阈值电压采样, 利用闪烁脉沖的先验知识,运用创新 的数字信号处理算法,对由探测器漏电流、脉沖拖尾及噪声干扰等引起的基线 漂移进行有效、 准确和快速地自适应校正, 同时提高闪烁脉沖数据测量结果的 信噪比和数据获得系统的稳定性。
附图说明 为了更清楚地说明本发明实施例或现有技术中的技术方案,下面将对实施 例或现有技术描述中所需要使用的附图作筒单地介绍,显而易见地, 下面描述 中的有关本发明的附图仅仅是本发明的一些实施例,对于本领域普通技术人员 来讲, 在不付出创造性劳动的前提下, 还可以根据这些附图获得其他的附图。
图 1为现有技术中常见的闪烁脉沖信号的示意图; 图 2为本发明数字化闪烁脉沖基线校正方法的流程图;
图 3 为本发明数字化闪烁脉沖基线校正方法中采用统计分布方法对基线 偏移值进行分析的结果图;
图 4 为本发明数字化闪烁脉沖基线校正方法对闪烁脉沖基线正向漂移进 行校正的效果图。
图 5为本发明数字化闪烁脉沖基线校正方法中采用 4阈值 MVT采样方法 对由 LYSO/PMT探测器输出闪烁脉沖进行采样及脉沖形状拟合的示意图; 图 6为本发明数字化闪烁脉沖基线校正方法中采用 4阈值 MVT采样方法 对由 LYSO/SiPM探测器输出闪烁脉沖进行采样及脉沖形状拟合的示意图; 图 7为本发明数字化闪烁脉沖基线校正系统的系统结构图;
图 8为本发明对能量分辨率的效果图, 其中, 8 ( a )为未使用本发明获得 的能量分辨率的效果图, 8 ( b ) 为使用本发明获得的能量分辨率的效果图。 具体实施方式 本发明公开了一种数字化闪烁脉沖的基线校正方法及系统,用于解决高能 物理粒子探测领域及医疗影像设备中前端电子学系统基线漂移问题。该方法通 过对闪烁脉沖进行多阈值电压采样, 利用闪烁脉沖的先验知识,运用创新的数 字信号处理算法,对由探测器漏电流、脉沖拖尾及噪声干扰等引起的基线漂移 进行有效、 准确和快速地自适应校正, 同时提高闪烁脉沖数据测量结果的信噪 比和数据获得系统的稳定性。
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, 显然, 所描述的实施例仅仅是本发明一部分实施例, 而不是全部的 实施例。基于本发明中的实施例, 本领域普通技术人员在没有做出创造性劳动 的前提下所获得的所有其他实施例, 都属于本发明保护的范围。
如图 2 所示, 本发明公开的数字化闪烁脉沖的基线校正方法包括如下步 骤:
( 1 )使用多阈值电压采样法, 对若干闪烁脉沖信号进行时间轴向采样, 获取若干闪烁脉沖信号对应的闪烁脉沖波形的数字化采样点; ( 2 )选取各个闪烁脉沖波形的至少两个数字化采样点, 根据闪烁脉沖波 形的形状特性及选取的数字化采样点,重建出该些闪烁脉沖波形并辨识重建出 的闪烁脉沖波形的模型参数,据此获得各个闪烁脉沖波形的基线漂移量, 然后 对所得的基线漂移量进行分析, 获得基线漂移均值;
( 3 )将若干重建后的闪烁脉沖波形减去步骤( 2 )中获得的基线漂移均值, 完成基线校正。
其中, 所述步骤(1 ) 中闪烁脉沖具体为一类由闪烁晶体、 光导和光电转 换器件决定的, 形状相同, 大小不同的脉沖信号。
其中, 所述步骤(1 ) 中多阈值电压采样方法的具体步骤为: 预先设置若 干阈值参考电压,记录各个闪烁脉沖的电压幅度达到各个阈值参考电压的具体 时刻, 每个阈值参考电压和对应的触发时间即组成一个时间-阈值采样点。
其中, 所述步骤(2 ) 中闪烁脉沖波形的形状特性根据所耦合的闪烁晶体 和光电转换器件的类别进行建脉沖模型获得,所建的该脉沖模型的表现形式为 一组可以刻画闪烁脉沖形状的函数表达式。
其中, 所述步骤(2 )中辨识模型参数是指根据闪烁脉沖波形的形状特性, 对所选取闪烁脉沖波形的时间 -阈值采样点进行函数逼近得到特征数学模型, 并从中提取刻画单个闪烁脉沖特性的参数值。
其中, 所述步骤(2 ) 中闪烁脉沖基线偏移量是从提取的刻画闪烁脉沖特 性的参数值中找到描述基线水平的量得到,获得闪烁脉沖基线偏移量的具体步 骤包括:在重建出的闪烁脉沖波形的模型参数中加入能够表示基线漂移的特征 量, 通过多个数字化的采样点, 利用函数逼近的方法, 求得能够刻画单个脉沖 的参数值, 并将表示基线漂移特征的参数值作为闪烁脉沖的基线漂移量。
其中, 为了减少噪声、脉沖形状拟合误差以及事件堆积等影响因素对基线 漂移量计算过程的干扰, 在进行基线校正之前有必要对多个基线值进行分析, 常见的方法包括: 均值计算、 统计分析等。 如图 3所示, 图 3为本发明数字化 闪烁脉沖基线校正方法中采用统计分析方法对基线偏移值进行分析的结果图。
图 4 为本发明数字化闪烁脉沖基线校正方法对闪烁脉沖基线正向漂移进 行校正的效果图。 其中, 实线 11为用示波器对 LYSO/PMT探测器输出的闪烁 脉沖进行采样重建得到的波形; 虚线 12为在未采用本发明提出的基线校正方 法下,根据闪烁脉沖模型拟合得到的波形, 可以看出拟合后的波形有明显的基 线偏移情况; 虚线 13为采用本发明提出的基线校正方法下, 根据闪烁脉沖模 型拟合得到的波形, 可以看出此时基线已无漂移情况。
由于闪烁晶体和光电转换器件的特性不同,其多种类别形式的耦合会有不 同的脉沖形状表征形式。 下面就具体的方法进行脉沖形状表征形式的阐释。
方法例一:
S101 :根据所耦合的闪烁晶体和光电转换器件类别,获取相应的闪烁脉沖 形状特性模型;
对于 LYSO闪烁晶体与线性光电倍增管 PMT进行耦合的情况下, 在不考 虑噪声影响下,闪烁脉沖形状的数学模型可考虑成由直线上升的上升沿和指数 衰减的下降沿组成, 其表达式如下所示:
BaseL t〈ts
LineKr x t + LineBr ts < t <
Qxp(-ExpKf xt + ExpBf ) + BaseL t≥t 其中, 参数 为上升沿直线的斜率且 ϋ >0, 参数 Z «e 是上升沿直 线的截距,可以为任意数值;参数 Exp ^为衰减时间常数且 Exp ^X),参数 ExpBf 可以为任意数值; 参数 ZteeL为闪烁脉沖的基线参量, 可以为任意数值; s为闪 烁脉沖的起始时间; ^为闪烁脉沖的峰值时间。 因此, 一个理想的闪烁脉沖可 以由五个模型特征值 Z we 、 LineBr、 ExpK f 、 ExpBf ^ BaseL ϋ , 闪烁月永沖 信号的开始时间、峰值时间、 峰值幅值、 余辉常数以及基线值等信息均可以由 这五个模型特征值计算获得, 公式如下:
( a ) 闪烁脉沖起始时间 ts 可以由如下公式求得:
_ BaseL― LineBr ·
5 LineKr
( b ) 闪烁脉沖峰值时间 ίρ
可以由如下公式求得: LineK r xtp+ LineBr = Qxp(-ExpKf xtp+ ExpBf ) + BaseL;
(c) 闪烁脉沖峰值幅值 Vp
可以由如下公式求得:
Vp = LineK r xtp + LineBr
(d)余辉常熟 r
可以由如下公式求得:
Figure imgf000011_0001
(e)基线值 Βναι = BaseL。 S102: 选取各个单次闪烁脉沖相应的数字化采样点, 根据步骤 S101中闪 烁脉沖的形状模型,依次重建脉沖波形并辨识模型参数,据此估计各个单次闪 烁脉沖的基线漂移量;
(a)对闪烁脉沖的直线上升沿产生的时间-阈值采样点, 按照如下公式进 行拟合:
V(t) = LineK r x t + LineBr
其中, 参数 为上升沿直线的斜率且 ϋ >0, 参数 Z «e 是上升沿直 线的截距, 可以为任意数值; 为阈值参考电压为 V(0时采用 MVT采样方法得 到的时间值。本方法亦可根据其他闪烁脉沖模型,对闪烁脉沖上升沿进行重建。
(b)对闪烁脉沖的指数衰减的下降沿产生的时间-阈值采样点, 按照如下 公式进行拟合:
V (t) = Qxp(-ExpKf x t + ExpBf ) + BaseL
其中,参数 Exp^为衰减时间常数且 Exp X),参数 Ex/^可以为任意数值; 参数 ZteeL为闪烁脉沖的基线参量,可以为任意数值; 为阈值参考电压为 时 采用 MVT采样方法得到的时间值。 本方法亦可根据其他闪烁脉沖模型, 对闪 烁脉沖下降沿进行重建。
(C)从重建的脉沖波形数学表达式中提取五个模型特征量, 据此获取闪 烁脉沖信号的开始时间、 峰值时间、 峰值幅值、 余辉常数以及基线值等信息; 具体请参照图 5所示, 图 5为数字化闪烁脉沖基线校正方法中采用 4阈值 MVT采样方法对由 LYSO/PMT探测器输出闪烁脉沖进行采样及脉沖形状拟合 的示意图。 其中, 实线 21表示由示波器对 LYSO/PMT探测器输出闪烁脉沖进 行采样后重建得到的波形; 圓点 22为用 MVT方法在设置的参考阈值电压下 对闪烁脉沖进行采样得到的采样点; 虚线 23为根据本发明给出的直线上升沿 和指数下降沿闪烁脉沖模型进行拟合得到的波形。从图中可以看出,根据闪烁 脉沖模型拟合得到的波形能够较好地逼近由示波器采样重建得到的波形,同时 也印证了不同的闪烁晶体和光电转换器件类别的耦合需要获取相应的闪烁脉 沖模型。
S103: 对所得的基线漂移量数据集进行分析, 以获取基线漂移均值。 方法例二:
S201: 根据所耦合的闪烁晶体和光电转换器件类别, 获取相应的闪烁脉沖 形状特性模型; 对于 LYSO闪烁晶体与非线性硅光电倍增器 SiPM进行耦合的情况下, 闪 烁脉沖形状的数学模型可近似由快速的指数上升沿和衰减的指数下降沿组成, 其表达式如下所示:
BaseL t〈ts
V(t) = Ar x xp(-ExpKr xt) + Br ts <t <
Af XQxp(-ExpKf xt) + BaseL t≥t 其中, 参数 A为指数上升沿的幅值系数, E^^为指数上升沿的时间常数 且 EWKr>0, 参数 可以为任意数值; 参数 ^为指数下降沿的幅值系数, 参数 Ep^为指数下降沿的时间常数且 Ep^X), 参数 为闪烁脉沖的基线参 量, 可以为任意数值; s为闪烁脉沖的起始时间; ^为闪烁脉沖的峰值时间。 因此, 一个理想的闪烁脉沖可以由六个模型特征值 A、 ExpKr、 Br, Af , ExpKf 和 «^来描述, 闪烁脉沖信号的开始时间、 峰值时间、 峰值幅值及基线值等 信息均可以由这六个模型特征值计算获得, 公式如下:
(a) 闪烁脉沖起始时间 s 可以由如下公式求得:
Figure imgf000013_0001
(b ) 闪烁脉沖峰值时间 ^
可以由如下公式求得:
·χ
Figure imgf000013_0002
Qxp(-ExpKf xtp) + BaseL;
( c ) 闪烁脉沖峰值幅值 ^
可以由如下公式求得:
Vp =ArX Qxp(-ExpKr xtp) + Br
( d)基线值 Bval = BaseL。
S202: 选取各个单次闪烁脉沖相应的数字化采样点,根据步骤 201中闪烁 脉沖的形状模型,依次重建脉沖波形并辨识模型参数,据此估计各个单次闪烁 脉沖的基线漂移量;
(a)对闪烁脉沖的指数上升沿产生的时间-阈值采样点, 按照如下公式进 行拟合:
V(t) = Arx exp(-ExpKr xt) + Br 其中, 参数 A为指数上升沿的幅值系数, E^^为指数上升沿的时间常数 JLEwKr>0, 参数 可以为任意数值; 为阈值参考电压为 v(0时采用 MVT采 样方法得到的时间值。本方法亦可根据其他闪烁脉沖模型,对闪烁脉沖上升沿 进行重建。
(b )对闪烁脉沖的指数衰减的下降沿产生的时间-阈值采样点, 按照如下 公式进行拟合:
V(t) = Af XQxp(-ExpKf xt) + BaseL 其中, 参数 A,为指数下降沿的幅值系数, 参数 Ep^为指数下降沿的时间 常数且 E p X); 参数 ZteeL为闪烁脉沖的基线参量, 可以为任意数值; 为阈 值参考电压为 V(0时采用 MVT采样方法得到的时间值。 本方法亦可根据其他 闪烁脉沖模型, 对闪烁脉沖下降沿进行重建。
( C )从重建的脉沖波形数学表达式中提取六个模型特征量, 据此获取闪 烁脉沖信号的开始时间、 峰值时间、 峰值幅值以及基线值等信息;
具体请参照图 6所示, 图 6 为本发明采用 4 阈值的 MVT采样方法对 LYSO/SiPM 探测器输出的闪烁脉沖进行数字化采样及脉沖形状拟合的示意 图。其中, 实线 31为用示波器对 LYSO/SiPM探测器输出的闪烁脉沖进行采样 后重建的波形;圓点 32为用 MVT方法在设置的参考阈值电压下对 LYSO/SiPM 探测器输出的闪烁脉沖进行采样得到的采样点; 虚线 33为根据本发明给出的 指数上升沿和指数下降沿闪烁脉沖模型对采样点进行拟合得到的波形。从图中 可以看出,根据闪烁脉沖模型拟合得到的波形能够较好地逼近由示波器采样重 建得到的波形,同时也印证了不同的闪烁晶体和光电转换器件类别的耦合需要 获取相应的闪烁脉沖模型。
S203: 对所得的基线漂移量数据集进行分析, 以获取基线漂移均值。
如图 7所示, 本发明公开的数字化闪烁脉沖的基线校正系统, 其包括: 数字化采样模块 100, 用于使用多阈值电压采样方法对闪烁脉沖信号进行 时间轴向采样, 获取闪烁脉沖信号对应的闪烁脉沖波形的数字化采样点;
基线偏移计算模块 200, 用于根据闪烁脉沖波形的形状特性及选取的数字 化采样点,重建出闪烁脉沖波形并辨识模型参数,获得闪烁脉沖的基线漂移量, 然后对所得的基线漂移量进行分析, 获得基线漂移均值;
基线校正模块 300, 用于对闪烁脉沖进行基线校正, 还原闪烁脉沖的原始 数据信息。
上述数字化闪烁脉沖的基线校正系统中,数字化采样模块 100分为 3个子 模块,分别为阈值电压设置模块 110、阈值甄别器模块 120和时间标记模块 130。
其中, 阈值电压设置模块 110用于设定阈值参考电压, 并将阈值参考电压 送到阈值甄别器模块 120和时间标记模块 130。
其中,阈值甄别器模块 120用于比较闪烁脉沖阈值电压与阈值参考电压之 间的大小关系, 并在闪烁脉沖电压穿过阈值参考电压时产生逻辑脉沖, 并将产 生的逻辑脉沖送到时间标记模块 130进行时间打标。
其中,时间标记模块 130用于对阈值甄别器模块 120输出的逻辑脉沖进行 时间标记, 并将所得的时间戳与其相应的阈值参考电压组成时间 -阈值采样点 并传送到基线偏移计算模块 200。
上述数字化闪烁脉沖的基线校正系统中,基线偏移计算模块 200包括事件 堆积剔除模块 210、 脉沖重建模块 220和基线偏移量计算模块 230。
其中, 事件堆积剔除模块 210用于鉴别及剔除闪烁脉沖中的堆积事件。 其中, 脉沖重建模块 220用于重建所述闪烁脉沖波形, 辨识模型参数, 并 将重建参数值传送到基线偏移量计算模块 230。
其中,基线偏移量计算模块 230根据脉沖重建模块 220获取的重建参数计 算闪烁脉沖的基线偏移量,然后对一段时间范围内的基线偏移量进行统计直方 图分析, 获得闪烁脉沖的平均基线偏移量并传送到基线校正模块 300。
对所得到的各个闪烁脉沖的基线偏移量进行分析的方法为均值计算或统 计分析。
以下通过一个具体的实施例数据对本发明的数字化闪烁脉沖的基线校正 方法及系统做进一步验证, 其中, 涉及到若干参数, 这些参数需要针对具体处 理数据进行调节以达到良好的性能, 下面列出本应用实例处理数据的参数。
步骤( 1 ) 中多阈值参考电压采样法中设置 6个阈值参考电压, 具体电压 幅值分别为 1.5mV, 21.5mV, 41.5 mV, 61.5 mV, 81.5 , 101.5; 导入的数字 脉沖为使用单根 LYSO (硅酸钇镥闪烁晶体 ) 晶体条和 Hammatsu R9800 PMT (光电倍增管)耦合采集到的闪烁脉沖。 采样率为 20GSps, 脉沖样本个数为 3000个, 每个脉沖波形样本点为 4000个。 能谱范围在 511 keV, 平均脉沖上 升沿时间约为 Ins, 探测器衰减时间常数约为 47ns。
步骤( 2 ) 中方法示例一中的闪烁脉沖是由 LYSO 晶体条与 Hamamatsu
R9800 PMT耦合产生的, 闪烁脉沖模型采用直线上升沿和指数衰减下降沿组 成;
步骤(2 ) 中方法示例二中的闪烁脉沖是由 LYSO晶体条与 SensL Array4 SiPM耦合产生的, 闪烁脉沖模型采用指数上升沿和指数衰减下降沿组成; 步骤(2 ) 中对由 LYSO/PMT探测器输出的 3000个闪烁脉沖的基线漂移 量进行统计分析, 基线偏移量是 -10mv;
具体请参照图 8所示,图 8为采用本发明方法下对能量分辨率的校正情况, ( a ) 为 MVT采样方法下未进行基线校正时得到的能量分辨率 22.3%, ( b ) 为 MVT采样方法下已进行基线校正得到的能量分辨率 18.1%。
对于本领域技术人员而言, 显然本发明不限于上述示范性实施例的细节, 而且在不背离本发明的精神或基本特征的情况下,能够以其他的具体形式实现 本发明。 因此, 无论从哪一点来看, 均应将实施例看作是示范性的, 而且是非 限制性的, 本发明的范围由所附权利要求而不是上述说明限定, 因此旨在将落 在权利要求的等同要件的含义和范围内的所有变化嚢括在本发明内。不应将权 利要求中的任何附图标记视为限制所涉及的权利要求。
此外, 应当理解, 虽然本说明书按照实施方式加以描述, 但并非每个实施 方式仅包含一个独立的技术方案, 说明书的这种叙述方式仅仅是为清楚起见, 本领域技术人员应当将说明书作为一个整体,各实施例中的技术方案也可以经 适当组合, 形成本领域技术人员可以理解的其他实施方式。

Claims

权 利 要 求
1、 一种数字化闪烁脉沖的基线校正方法, 其特征在于: 所述基线校正方 法步骤如下:
( 1 )对若干闪烁脉沖信号进行时间轴向采样, 获取若干闪烁脉沖信号对 应的闪烁脉沖波形的数字化采样点;
( 2 )选取各个闪烁脉沖波形的至少两个数字化采样点, 根据闪烁脉沖波 形的形状特性及选取的数字化采样点,重建出该些闪烁脉沖波形并辨识重建出 的闪烁脉沖波形的模型参数,据此获得各个闪烁脉沖波形的基线漂移量, 然后 对所得的基线漂移量进行分析, 获得基线漂移均值;
( 3 )将若干重建后的闪烁脉沖波形减去步骤( 2 )中获得的基线漂移均值, 完成基线校正。
2、 根据权利要求 1所述的数字化闪烁脉沖的基线校正方法,其特征在于: 所述步骤(1 ) 中, 采用多阈值电压采样法对若干闪烁脉沖信号进行时间轴向 采样, 所述多阈值电压采样法的具体步骤为: 预先设置若干阈值参考电压, 记 录各个闪烁脉沖的电压幅度达到各个阈值参考电压的具体时刻,每个阈值参考 电压和对应的触发时间即组成一个时间 -阈值采样点。
3、 根据权利要求 1所述的数字化闪烁脉沖的基线校正方法,其特征在于: 所述步骤(2 ) 中闪烁脉沖波形的形状特性根据所耦合的闪烁晶体和光电转换 器件的类别进行建脉沖模型获得,所建的该脉沖模型的表现形式为一组可以刻 画闪烁脉沖形状的函数表达式。
4、 根据权利要求 1至 3任一所述的数字化闪烁脉沖的基线校正方法, 其 特征在于: 所述步骤(2 )中辨识模型参数是指根据闪烁脉沖波形的形状特性, 对所选取闪烁脉沖波形的时间 -阈值采样点进行函数逼近得到特征数学模型, 并从中提取刻画单个闪烁脉沖特性的参数值。
5、 根据权利要求 4任一所述的数字化闪烁脉沖的基线校正方法, 其特征 在于: 所述步骤(2 ) 中闪烁脉沖基线偏移量是从提取的刻画闪烁脉沖特性的 参数值中找到描述基线水平的量得到,获得闪烁脉沖基线偏移量的具体步骤包 括: 在重建出的闪烁脉沖波形的模型参数中加入能够表示基线漂移的特征量, 通过多个数字化的采样点, 利用函数逼近的方法, 求得能够刻画单个脉沖的参 数值, 并将表示基线漂移特征的参数值作为闪烁脉沖的基线漂移量。
6、 根据权利要求 1至 3任一所述的数字化闪烁脉沖的基线校正方法, 其 特征在于: 所述步骤(2 ) 中对所得到的各个闪烁脉沖的基线偏移量进行分析 的方法为均值计算或统计分析。
7、 一种数字化闪烁脉沖的基线校正系统, 其特征在于: 包括:
数字化采样模块,用于使用多阈值电压采样方法对闪烁脉沖信号进行时间 轴向采样, 获取闪烁脉沖信号对应的闪烁脉沖波形的数字化采样点;
基线偏移计算模块,用于根据闪烁脉沖波形的形状特性及选取的数字化采 样点, 重建出闪烁脉沖波形并辨识模型参数, 获得闪烁脉沖的基线漂移量, 然 后对所得的基线漂移量进行分析, 获得基线漂移均值;
基线校正模块, 用于对闪烁脉沖进行基线校正,还原闪烁脉沖的原始数据 信息。
8、 根据权利要求 7所述的数字化闪烁脉沖的基线校正系统,其特征在于: 所述数字化采样模块包括阈值电压设置模块、 阈值甄别器模块和时间标记模 块, 其中,
阈值电压设置模块用于设定阈值参考电压,并将阈值参考电压送到阈值甄 别器模块和时间标记模块;
阈值甄别器模块用于比较闪烁脉沖阈值电压与阈值参考电压之间的大小 关系, 并在闪烁脉沖电压穿过阈值参考电压时产生逻辑脉沖, 并将产生的逻辑 脉沖送到时间标记模块进行时间打标;
时间标记模块用于对阈值甄别器模块输出的逻辑脉沖进行时间标记,并将 所得的时间戳与其相应的阈值参考电压组成时间-阈值采样点并传送到基线偏 移计算模块。
9、 根据权利要求 7所述的数字化闪烁脉沖的基线校正系统,其特征在于: 所述基线偏移计算模块包括事件堆积剔除模块、脉沖重建模块和基线偏移量计 算模块, 其中,
事件堆积剔除模块用于鉴别及剔除闪烁脉沖中的堆积事件; 脉沖重建模块用于重建所述闪烁脉沖波形, 辨识模型参数, 并将重建参数 值传送到基线偏移量计算模块;
基线偏移量计算模块根据脉沖重建模块获取的重建参数计算闪烁脉沖的 基线偏移量, 然后对一段时间范围内的基线偏移量进行统计直方图分析, 获得 闪烁脉沖的平均基线偏移量并传送到基线校正模块。
10、根据权利要求 7所述的数字化闪烁脉沖的基线校正系统,其特征在于: 对所得到的各个闪烁脉沖的基线偏移量进行分析的方法为均值计算或统计分
PCT/CN2013/073126 2013-02-05 2013-03-25 数字化闪烁脉冲的基线校正方法及系统 WO2014121548A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201310045576.5A CN103969675B (zh) 2013-02-05 2013-02-05 数字化闪烁脉冲的基线校正方法及系统
CN201310045576.5 2013-02-05

Publications (1)

Publication Number Publication Date
WO2014121548A1 true WO2014121548A1 (zh) 2014-08-14

Family

ID=51239402

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2013/073126 WO2014121548A1 (zh) 2013-02-05 2013-03-25 数字化闪烁脉冲的基线校正方法及系统

Country Status (2)

Country Link
CN (1) CN103969675B (zh)
WO (1) WO2014121548A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105866011A (zh) * 2016-03-31 2016-08-17 艾康生物技术(杭州)有限公司 脉冲基线值计算方法及血球分析仪的粒子计数方法

Families Citing this family (20)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105824817B (zh) * 2015-01-05 2019-05-21 苏州瑞派宁科技有限公司 一种闪烁脉冲的数字化方法
CN105212954B (zh) * 2015-11-05 2018-03-16 苏州瑞派宁科技有限公司 一种脉冲堆积事件实时处理方法与系统
CN105629290B (zh) * 2016-02-16 2018-12-04 北京中科坤润科技有限公司 一种数字核脉冲信号墨西哥草帽小波成形方法
CN105842544B (zh) * 2016-03-18 2018-09-18 南京瑞派宁信息科技有限公司 一种迭代的闪烁脉冲时间标记及其交叉验证方法
CN106125127B (zh) * 2016-06-18 2018-12-04 武汉京邦科技有限公司 一种硅光电倍增器的时间标记方法及其微元阵列编码系统
CN106361361A (zh) * 2016-08-26 2017-02-01 沈阳东软医疗系统有限公司 一种信息处理方法和装置
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
CN107422364B (zh) 2017-06-26 2019-06-25 东软医疗系统股份有限公司 一种检测器的时间定时校正方法及装置
CN107260197B (zh) 2017-07-10 2020-11-10 东软医疗系统股份有限公司 时间校正方法、装置及设备
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
CN109444559B (zh) * 2018-10-26 2021-02-05 苏州瑞迈斯医疗科技有限公司 脉冲信号的采样方法、重建方法和装置
CN110327063B (zh) * 2019-06-10 2023-05-12 沈阳智核医疗科技有限公司 能量校正方法、装置、控制台设备及pet系统
CN110286405B (zh) * 2019-07-10 2020-09-15 中国科学院近代物理研究所 一种深空探测器系统的标定装置的应用
CN111025373B (zh) * 2019-12-26 2023-03-31 中广核久源(成都)科技有限公司 数字化实时校正碘化钠晶体衰减时间方法
CN111307304B (zh) * 2020-03-09 2021-09-07 上海航天计算机技术研究所 红外钟形脉冲特征参量提取验证方法及平台
CN113204044A (zh) * 2021-03-16 2021-08-03 华中科技大学 一种用于核测井仪器的核脉冲能量测量方法
CN115345208B (zh) * 2022-10-19 2023-02-03 成都理工大学 基于顶帽变换的中子-伽马脉冲堆积甄别方法
CN115778337B (zh) * 2022-10-28 2023-06-16 北京津发科技股份有限公司 人体生物信号数据处理方法、装置、设备及存储介质
CN115980821B (zh) * 2023-01-16 2023-09-05 中国科学院近代物理研究所 一种基于fpga的自适应能谱测量方法及系统

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0508498A2 (en) * 1985-04-17 1992-10-14 Kabushiki Kaisha Toshiba Scintillation camera apparatus including autotuning
US20110095173A1 (en) * 2009-10-23 2011-04-28 Saint-Gobain Ceramics & Plastics, Inc. Radiation detector and method of using a radiation detector
CN102141632A (zh) * 2009-12-29 2011-08-03 株式会社东芝 提高γ射线检测中的定时分辨率的装置和相关方法
CN102262238A (zh) * 2011-04-19 2011-11-30 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
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

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8751195B2 (en) * 2008-09-24 2014-06-10 Inotera Memories, Inc. Method for automatically shifting a base line
CN102073059B (zh) * 2010-12-31 2013-05-22 华中科技大学 一种数字化pileup波形处理方法及系统
CN102499670A (zh) * 2011-11-23 2012-06-20 北京理工大学 基于稳健估计和固态模函数的心电基线漂移校正方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0508498A2 (en) * 1985-04-17 1992-10-14 Kabushiki Kaisha Toshiba Scintillation camera apparatus including autotuning
US20110095173A1 (en) * 2009-10-23 2011-04-28 Saint-Gobain Ceramics & Plastics, Inc. Radiation detector and method of using a radiation detector
CN102141632A (zh) * 2009-12-29 2011-08-03 株式会社东芝 提高γ射线检测中的定时分辨率的装置和相关方法
CN102262238A (zh) * 2011-04-19 2011-11-30 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
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

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN105866011A (zh) * 2016-03-31 2016-08-17 艾康生物技术(杭州)有限公司 脉冲基线值计算方法及血球分析仪的粒子计数方法

Also Published As

Publication number Publication date
CN103969675B (zh) 2017-08-04
CN103969675A (zh) 2014-08-06

Similar Documents

Publication Publication Date Title
WO2014121548A1 (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
US9772408B2 (en) Method and device for extracting scintillation pulse information
WO2016110141A1 (zh) 一种闪烁脉冲的数字化方法
WO2012088781A1 (zh) 一种数字化pileup波形处理方法及系统
EP3073292B1 (en) Method and system for recovering scintillation pulse information
US8969814B2 (en) System and method of determining timing triggers for detecting gamma events for nuclear imaging
US20150025852A1 (en) Signal data processing for radiation detection
WO2021135337A1 (zh) 信号采样电路、探测装置及成像系统
US9945962B2 (en) Signal processor and radiation detection device
CN106842277A (zh) 一种堆积事件处理方法及装置
Lin et al. An MVT based all-digital DAQ for energy determination
Ruiz-Gonzalez et al. Fisher information analysis of digital pulse timing
Xie et al. A new approach for pulse processing in PET
CN111948695A (zh) 一种pet信号的事件属性计算方法
CN117826228A (zh) 康普顿锥获取方法、装置、系统、设备及存储介质
Xie et al. An investigation of baseline calibration method for digitally sampling scintillation pulses in PET
CN113729751A (zh) 一种示踪剂活度的检测方法、设备及pet扫描仪
Du et al. Analysis of digital timing methods with DRS4 module
Garbino et al. A pulse modeling tool for PET scanners
OA16634A (en) Method and device for extracting scintillation pulse information

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: 13874849

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 13874849

Country of ref document: EP

Kind code of ref document: A1