Energy spectrum processing method for quasi-hemispherical CdZnTe detector

Abstract

The invention relates to the technical field of nuclear radiation detection, in particular to an energy spectrum processing method for a quasi-hemispherical CdZnTe detector, and the method comprises the steps: obtaining an original signal collected by the quasi-hemispherical CdZnTe detector; performing normalization processing on the amplitude of the original signal from the baseline to the highest point in the rising stage, dynamically selecting a plurality of energy threshold values, and performing rising time discrimination on the normalized signal in each energy threshold value; screening out signals with the rising time greater than the RTflag and determining a rising time range corresponding to the energy threshold value; constructing an intermediate energy spectrum corresponding to each energy threshold value according to the signal with the rise time greater than the RTflag and the rise time range; performing post-processing on the intermediate energy spectrum to obtain a target energy spectrum corresponding to each energy threshold value; according to the invention, energy threshold discrimination, rise time discrimination and an energy spectrum post-processing algorithm are combined, so that the energy resolution of the quasi-hemispherical CdZnTe detector to 662 keV gamma rays is stably improved from conventional 2-3% to less than 1%.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of nuclear radiation detection, and particularly relates to a spectrum processing method for a quasi-hemispherical CdZnTe detector. BACKGROUND

[0002] CdZnTe semiconductor material has become a detection material with relatively outstanding comprehensive performance in the field of nuclear radiation detection due to its high atomic number, wide band gap, high resistivity, excellent electron mobility, and other advantages. Since the product of the hole mobility and lifetime in CdZnTe material is about two orders of magnitude lower than that of the electron, such a detector is usually designed as a unipolar device to optimize its charge collection performance. Among them, the quasi-hemispherical electrode structure can effectively improve the energy resolution of the detector by realizing the "small pixel effect". The gamma ray spectrum detector prepared based on this structure not only has high energy resolution, small volume, and convenient use, but also can work stably at room temperature, and thus has been widely used in the fields of nuclear safety inspection, radiation dose monitoring, and high-energy physics experiments.

[0003] Currently, in the high-end applications such as deep trace element detection, new nuclear drug research and development, and nuclear safety monitoring, the energy resolution of the detector at the 662 keV energy point is required to be better than 1%. In order to improve the energy resolution, the rise time discrimination of the output signal of the detector has become an important means. The rise time discrimination technology is to filter out the event signals with shorter rise time and closer physical action position to the cathode region of the detector according to the rising edge waveform of the output signal of the charge-sensitive preamplifier. Since the weight corresponding to such signals is close to the ideal range from 0 to 1, for a complete energy deposition event, theoretically, higher energy resolution can be obtained. However, the energy resolution of the quasi-hemispherical CdZnTe detector still stays at the level of 2% to 3%, and it is difficult to break through the bottleneck of 1%. The main reason is that the actual effectiveness of the technology depends on the discrimination parameters such as the energy threshold and the rise time threshold, and these parameters face the problem of universality in practical application. On the one hand, due to the inherent differences caused by material growth and preparation process, different quasi-hemispherical CdZnTe detectors have individual inconsistencies in key parameters such as electron and hole collection efficiency, and different discrimination parameters need to be accurately matched, which leads to the fact that the system algorithm cannot effectively filter out the interaction signals with faster rise time and closer action position to the cathode region in the same test environment. On the other hand, after the hardware system is replaced, the individual performance difference of the electronic components (such as the feedback capacitor in the charge-sensitive preamplifier, whose tolerance can usually reach ±10%) will directly change the rise time characteristics of the output signal. This makes the fixed discrimination parameters in the original algorithm invalid, and cannot accurately adapt to the new signal waveform.

[0004] Therefore, a spectrum processing method capable of overcoming individual differences of systems and realizing adaptive matching of parameters is proposed, which is crucial for promoting stable and reliable high-energy resolution of quasi-hemispherical CdZnTe detectors. SUMMARY

[0005] In view of the problems in the prior art, the present application provides a spectrum processing method for a quasi-hemispherical CdZnTe detector to solve the problem that fixed-time discrimination parameters cannot be compatible with different detectors and hardware systems, and to realize the purpose of significantly and stably improving the resolution of quasi-hemispherical CdZnTe detectors.

[0006] To achieve the above object, the present application provides the following technical scheme: The present application provides a spectrum processing method for a quasi-hemispherical CdZnTe detector, comprising the following steps: Step 1: obtaining original signals collected by a quasi-hemispherical CdZnTe detector; Step 2: after normalizing the amplitude of the original signals from the baseline to the highest point in the rising stage, dynamically selecting a plurality of energy thresholds, performing rise time discrimination on the normalized signals within each energy threshold, and screening out signals with a rise time greater than RTflag and determining the rise time range corresponding to the energy threshold; Step 3: constructing an intermediate energy spectrum corresponding to each energy threshold according to the signals with a rise time greater than RTflag and the rise time range; Step 4: post-processing the intermediate energy spectrum to obtain a target energy spectrum corresponding to each energy threshold.

[0007] Further, the energy thresholds include 5%-95%, 10%-95%, 5%-90%, and 10%-90%.

[0008] Further, in step 2, the determination method of RTflag is as follows: dividing the sampling time into groups with m sampling intervals, 6≤m≤15, counting the integral counts of the pulse amplitudes of all normalized signals in each group in the rise time interval in the full spectrum, and drawing a relationship diagram of the rise time interval and the integral counts of the signals in the interval to the full spectrum; determining the rise time threshold RTflag in the energy threshold based on the relationship diagram; and the rise time range corresponding to the energy threshold is (RTflag+1ns)~(RTflag+effective interval width), wherein the effective interval is a continuous rise time interval with a contribution ratio of integral counts of ≥70%.

[0009] Further, in step 3, the method for constructing the intermediate energy spectrum comprises: accumulating signals with the rise time greater than RTflag in each energy threshold to obtain a total original energy spectrum corresponding to the energy threshold, screening signals corresponding to the rise time range from the total original energy spectrum, and accumulating the signals corresponding to the rise time range to obtain the intermediate energy spectrum corresponding to the energy threshold.

[0010] Further, in step 4, the post-processing of the intermediate energy spectrum comprises: sequentially performing filtering and smoothing processing and background subtraction processing on the intermediate energy spectrum.

[0011] Further, the filtering and smoothing processing adopts a Gaussian filter function and is calculated according to the following formula:

[0012] In the formula, x is the channel number of the energy spectrum, and σ is the coefficient of the Gaussian filter function in the window width. S i The energy spectrum after the filtering and smoothing processing is the energy spectrum of the first channel. i O i+j The intermediate energy spectrum is the energy spectrum of the first channel. i + j C j The Gaussian filter function is the coefficient of the Gaussian filter function in the window width. k to k

[0013] Further, the background subtraction processing adopts the SNIP algorithm.

[0014] Further, the background subtraction processing comprises: performing digital filtering on the smoothed energy spectrum to obtain a numerical transformed energy spectrum, determining a background spectrum to be peeled off based on the numerical transformed energy spectrum, and performing numerical inverse transformation on the background spectrum to be peeled off to obtain a target energy spectrum.

[0015] Further, the smoothed energy spectrum is first subjected to digital filtering according to the following formula:

[0016] In the formula, x is the channel number of the energy spectrum, and σ is the coefficient of the Gaussian filter function in the window width. G i The numerical transformed energy spectrum is the energy spectrum of the first channel. i

[0017] Further, the background spectrum to be peeled off is subjected to numerical inverse transformation according to the following formula:

[0018] In the formula, x is the channel number of the energy spectrum, and σ is the coefficient of the Gaussian filter function in the window width. b i The background spectrum to be peeled off is the energy spectrum of the first channel. i B i The target energy spectrum is the energy spectrum of the first channel.​​​​​i The count of the path.

[0019] Compared with the prior art, the application has the following beneficial effects: (1) The application combines energy threshold discrimination, rise time discrimination and energy spectrum post-processing algorithm, and the energy resolution of a quasi-hemispherical CdZnTe detector for 662 keV gamma rays is stably improved from 2% to 3% to less than 1%, which meets the stringent requirements of high-end nuclear radiation detection applications for high-performance detectors.

[0020] (2) After the amplitude of the original signal from the baseline to the highest point in the rising stage is normalized, the application selects four energy threshold dynamic changes of 5%-95%, 10%-95%, 5%-90% and 10%-90%. The energy contribution of the effective signal of the quasi-hemispherical CdZnTe detector is mainly concentrated in the middle and front to the middle and rear of the rising stage. The application can cover the core energy contribution interval through normalization processing and the combination of wide and narrow threshold values. Different quasi-hemispherical CdZnTe detectors can capture effective signals with high charge collection efficiency by dynamically selecting matching energy thresholds, flexibly cope with individual differences in carrier collection efficiency of different CdZnTe detectors, and changes in signal characteristics caused by different hardware systems (such as preamplifier feedback capacitor tolerance), thereby stably improving the energy resolution of the quasi-hemispherical CdZnTe detector.

[0021] (3) The application sets a specific rise time range for each energy threshold. The set energy threshold and rise time range work together to accurately select "elite" signals with high charge collection efficiency and originating from the cathode region of the detector, while effectively eliminating noise signals and signals caused by long hole drift time and serious charge loss. The background count and tailing effect of the energy spectrum are significantly reduced from the data source, further improving the energy resolution of the quasi-hemispherical CdZnTe detector.

[0022] (4) The application integrates signal discrimination and subsequent energy spectrum post-processing (including filtering and smoothing and background subtraction). Filtering and smoothing suppresses the statistical fluctuations of the energy spectrum, and background subtraction can maximize the Compton plateau and radiation background. This combined processing further extracts pure net energy spectrum, which together ensures the realization of the energy resolution index of the final quasi-hemispherical CdZnTe detector. BRIEF DESCRIPTION OF DRAWINGS

[0023] Figure 1 The method flowchart for realizing high energy resolution of the quasi-hemispherical CdZnTe detector provided by the embodiment of the application; Figure 2 The energy spectrum diagram of different rise time ranges when the energy threshold is 5%-95% in embodiment 1; Figure 3 is a schematic diagram of the energy spectrum in different rise time ranges when the energy threshold is 10%-95% in Example 1; Figure 4 is a schematic diagram of the energy spectrum in different rise time ranges when the energy threshold is 5%-90% in Example 1; Figure 5 is a schematic diagram of the energy spectrum in different rise time ranges when the energy threshold is 10%-90% in Example 1; Figure 6 is a graph of the number of iterations for background subtraction and the target energy spectrum when the energy threshold is 10%-90% in Example 1; Figure 7 is a graph of the number of iterations-peak efficiency (A) and the number of iterations-energy resolution (B) in Example 1; Figure 8 is a schematic diagram of the total original energy spectrum (A) and the target energy spectrum (B) corresponding to different energy thresholds obtained in Example 1; Figure 9 is a schematic diagram of the energy spectrum in different rise time ranges when the energy threshold is 10%-90% in Example 2; Figure 10 is a schematic diagram of the energy spectrum after rise time discrimination (A), the energy spectrum after filtering and smoothing (B), and the target energy spectrum (C) corresponding to the energy threshold of 10%-90% obtained in Example 2; Figure 11 is a resolution spectrum obtained according to the target energy spectrum in Example 2; Figure 12 is a schematic diagram of the energy spectrum in different rise time ranges when the energy threshold is 5%-95% in Example 3; Figure 13 is a schematic diagram of the energy spectrum after rise time discrimination (A), the energy spectrum after filtering and smoothing (B), and the target energy spectrum (C) corresponding to the energy threshold of 10%-90% obtained in Example 3; Figure 14 is a resolution spectrum obtained according to the target energy spectrum in Example 3; DETAILED DESCRIPTION

[0024] The present application will be described in detail below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of the present application, and not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by a person of ordinary skill in the art without creative labor fall within the protection scope of the present application.

[0025] Example 1 The present embodiment proposes a spectrum processing method for a quasi-hemispherical CdZnTe detector, comprising the following steps: Step 1: using137 A Cs radioactive source, gamma rays generated by the source are incident on a cathode surface of a quasi-hemispherical CdZnTe detector A, and a charge signal output by the quasi-hemispherical CdZnTe detector A is converted into a voltage signal via a charge-sensitive preamplifier A, and then sampled and digitized by an analog-to-digital converter to obtain an original signal; Step 2: After normalizing the amplitude of the original signal from the baseline to the highest point in the rising phase, four energy thresholds of 5%-95%, 10%-95%, 5%-90% and 10%-90% are dynamically selected, and the rising time discrimination is performed on the normalized signal in each energy threshold; signals with a rising time greater than RTflag (rising time threshold) are screened out, and the rising time range matched with the energy threshold is determined.

[0026] The specific rising time discrimination process is as follows: divide the sampling time into groups with m sampling intervals as a group, m is 10 in this embodiment, count the integral counts of the pulse amplitude corresponding to all normalized signals in each group of rising time intervals in the full spectrum, and draw a relationship diagram of the rising time interval and the integral counts of the signals in the interval to the full spectrum; based on the relationship diagram, the rising time threshold RTflag in the energy threshold is determined, Figures 2-5 are energy spectrum diagrams of different rising time ranges in the four energy thresholds of this embodiment; refer to Figures 2-5 It can be observed that in the region with shorter rising time, the integral count is very low and stable, this part is mainly contributed by noise, there is a critical rising time point (RTflag), after which the integral count starts to increase significantly and continuously, which indicates that the effective event signal starts to dominate, that is, RTflag can be determined by finding the inflection point of the energy spectrum diagram from the "flat platform" to the "rapid growth slope". refer to Figure 2 It can be seen that the RTflag corresponding to the energy threshold 5%-95% is 170ns; refer to Figure 3 It can be seen that the RTflag corresponding to the energy threshold 10%-95% is 160ns; refer to Figure 4 It can be seen that the RTflag corresponding to the energy threshold 5%-90% is 150ns; refer to Figure 5 It can be seen that the RTflag corresponding to the energy threshold 10%-90% is 140ns. The rising time range of the signal mainly contributing to the energy spectrum and the full energy peak count is concentrated in the effective interval greater than RTflag, wherein the effective interval is a continuous rising time interval with a contribution ratio of integral count ≥70%, based on this, (RTflag+1)~(RTflag+effective interval width) is taken as the rising time range matched with the energy threshold, and the effective interval width is usually about 20ns, from Figures 2-5As can be seen, in this embodiment, the rise time discrimination range corresponding to 5%-95% is 171~190ns, the rise time discrimination range corresponding to 10%-95% is 161~180ns, the rise time discrimination range corresponding to 5%-90% is 151~170ns, and the rise time discrimination range corresponding to 10%-90% is 141~160ns.

[0027] Step 3: For each energy threshold, accumulate signals with rise times greater than RTflag by energy channel address to obtain the total raw energy spectrum corresponding to that energy threshold. Filter signals from the total raw energy spectrum that correspond to the rise time range, and then accumulate these signals by energy channel address to obtain the intermediate energy spectrum corresponding to that energy threshold. For example, filter signals with rise times greater than 140ns within the energy threshold range of 5%-95%, accumulate these signals by energy channel address to obtain the total raw energy spectrum corresponding to the energy threshold range of 5%-95%, then filter all signals with rise times in the range of 171-190ns from the total raw energy spectrum corresponding to the energy threshold range of 5%-95%, and accumulate these signals by energy channel address to obtain the intermediate energy spectrum corresponding to the energy threshold range of 5%-95%. Similarly, obtain the intermediate energy spectra corresponding to energy thresholds of 10%-95%, 5%-90%, and 10%-90%, respectively.

[0028] Step 4: Perform point-by-point filtering and smoothing and background subtraction on each intermediate energy spectrum to obtain the target energy spectrum corresponding to each energy threshold.

[0029] The filtering and smoothing process uses a Gaussian filtering function and is calculated according to the following formula:

[0030] In the formula S i The energy spectrum after filtering and smoothing is at the 1st i The counting of the Tao; O i+j For the intermediate energy spectrum at the 1st i + j The counting of the Tao; C j For Gaussian filtering functions with a window width of - k arrive k The coefficients within, in this embodiment, k It is 20.

[0031] After smoothing, the background is subtracted by a statistically sensitive nonlinear iterative peak stripping algorithm (SNIP) for each energy spectrum.

[0032] First, perform digital filtering according to the following formula:

[0033] In the formula, G i The energy spectrum after numerical transformation is at the th i The counting of channels, where logarithmic and square root operations can reduce the difference between the maximum and minimum counts at different channel addresses, enhancing sensitivity to peaks. Then, the background spectrum to be stripped is determined, and the count at the current channel address is... G i The smaller value is used as the background count compared to the mean count at both edges of the window width. b i To minimize the impact on peak count while maximizing background subtraction, the final step is to... b i The background spectrum is obtained by performing the inverse numerical transformation according to the following formula. B i :

[0034] Since the SNIP background subtraction algorithm can significantly affect the Compton plateau and the low-energy side of the full-energy peak, it is the main post-processing step affecting the full-energy peak efficiency. Choosing an appropriate number of iterations can reduce the impact of incomplete charge collection on the full-energy peak counting, while simultaneously improving the peak shape quality and energy resolution of the spectrum. Furthermore, the effect of the number of iterations on peak efficiency follows an exponential decay law, which facilitates quantitative energy spectrum analysis while achieving high energy resolution.

[0035] In this embodiment, the method for selecting the number of iterations is as follows: First, plot the relationship between the number of iterations and the target energy spectrum. Then, based on the plot of the number of iterations versus the target energy spectrum, plot the relationship between the number of iterations versus the peak efficiency and the energy resolution. Next, select the point of maximum peak efficiency from the points of overlap between the plot of the number of iterations versus the peak efficiency and the plot of the number of iterations versus the energy resolution. The number of iterations corresponding to this point is the optimal number of iterations. Figure 6 This is a graph showing the relationship between the number of iterations for background subtraction and the target energy spectrum when the energy threshold is 10%-90% in this embodiment. Figure 7 This is a graph showing the relationship between the number of iterations, peak efficiency, and energy resolution when the energy threshold is 10%-90% in this embodiment; where A represents the relationship between the number of iterations and peak efficiency, and B represents the relationship between the number of iterations and energy resolution. Figure 7 It can be seen that in this embodiment, the optimal number of iterations for the SNIP algorithm corresponding to an energy threshold of 10%-90% is 20.

[0036] Figure 8 In this context, ET represents the energy threshold, RT represents the rise time, and the reference is... Figure 8It can be seen that, after the energy spectrum post-processing of energy threshold selection, rise time discrimination, energy spectrum smoothing and background deduction, the energy resolution of the total original energy spectrum for 662 keV is improved from 1.56% to 0.78%.

[0037] Embodiment 2 The difference between the embodiment and embodiment 1 is that, in the embodiment, the quasi-hemispherical CdZnTe detector A is replaced by a quasi-hemispherical CdZnTe detector B.

[0038] Reference Figures 9-11 It can be seen that, in the embodiment, when the energy threshold is 10%-90% and the rise time range is selected as 141-160 ns, the energy resolution of the target energy spectrum for 662 keV reaches 0.77%. Figure 9 The energy spectrum (A) after rise time discrimination is the energy spectrum formed by signals with a rise time greater than RTflag.

[0039] Embodiment 3 The difference between the embodiment and embodiment 1 is that, in the embodiment, the charge-sensitive preamplifier A is replaced by a charge-sensitive preamplifier B.

[0040] Reference Figures 12-14 It can be seen that, in the embodiment, when the energy threshold is 5%-95% and the rise time range is selected as 171-190 ns, the energy resolution of the target energy spectrum for 662 keV reaches 0.97%, Figure 12 In the embodiment, the energy spectrum (A) after rise time discrimination is the energy spectrum formed by signals with a rise time greater than RTflag.

[0041] From the energy resolution results of the above embodiments, it can be seen that the energy spectrum processing method proposed in the present application can overcome individual differences of the system, realize adaptive matching of parameters, and stably improve the energy resolution of the quasi-hemispherical CdZnTe detector for 662 keV gamma rays to less than 1% by combining energy threshold discrimination, rise time discrimination and energy spectrum post-processing algorithm, thereby meeting the stringent requirements of high-end nuclear radiation detection applications for high-performance detectors.

[0042] The above is only a specific embodiment of the present application, enabling those skilled in the art to understand or implement the present application. Various modifications to these embodiments will be apparent to those skilled in the art, and the general principles defined herein can be implemented in other embodiments without departing from the spirit or scope of the present application.

[0043] It is to be understood that the application is not limited to what has been described above and that modifications and / or additions can be made thereto without departing from the scope of the application. The scope of the application should, therefore, be determined not with reference to the above description, but instead with reference to the appended claims along with their full scope of equivalents.

Claims

1. A method for energy spectrum processing of a quasi-semi-spherical CdZnTe detector, characterized in that, The method comprises the following steps: Step 1: obtaining original signals collected by a quasi-hemispherical CdZnTe detector; Step 2: after normalizing the amplitude of the original signals from the baseline to the highest point in the rising stage, dynamically selecting multiple energy thresholds, discriminating the rising time of the normalized signals in each energy threshold, screening the signals with the rising time greater than RTflag and determining the rising time range corresponding to the energy threshold; Step 3: constructing the intermediate energy spectrum corresponding to each energy threshold according to the signals with the rising time greater than RTflag and the rising time range; Step 4: post-processing the intermediate energy spectrum to obtain the target energy spectrum corresponding to each energy threshold.

2. The method for processing energy spectrum of quasi-hemispherical CdZnTe detector according to claim 1, characterized in that, The energy thresholds include 5%-95%, 10%-95%, 5%-90% and 10%-90%.

3. The method of claim 1, wherein the quasi-semi-spherical CdZnTe detector is a detector having a shape of a semi-sphere with a flat surface. In step 2, the determination method of RTflag is as follows: dividing the sampling time into groups with m sampling intervals, 6≤m≤15, counting the integral counts of the pulse amplitude of all normalized signals in each group in the rising time interval in the full spectrum, and drawing a relationship diagram of the rising time interval and the integral counts of the signals in the interval to the full spectrum; determining the rising time threshold RTflag in the energy threshold based on the relationship diagram; and the rising time range corresponding to the energy threshold is (RTflag+1ns)~(RTflag+effective interval width), wherein the effective interval is a continuous rising time interval with a contribution ratio of integral counts of greater than or equal to 70%.

4. The method for processing energy spectrum of quasi-semi-spherical CdZnTe detector according to claim 1, characterized in that, In step 3, the method for constructing the intermediate energy spectrum comprises: accumulating the signals with the rising time greater than RTflag in each energy threshold to obtain the total original energy spectrum corresponding to the energy threshold, screening the signals corresponding to the rising time range from the total original energy spectrum, and accumulating the signals corresponding to the rising time range to obtain the intermediate energy spectrum corresponding to the energy threshold.

5. The method for processing energy spectrum of quasi-hemispherical CdZnTe detector according to claim 1, wherein, In step 4, the post-processing of the intermediate energy spectrum comprises: sequentially performing filter smoothing processing and background subtraction processing on the intermediate energy spectrum.

6. The method for processing energy spectrum of quasi-semi-spherical CdZnTe detector according to claim 5, characterized in that, The filter smoothing processing adopts a Gaussian filter function and is calculated according to the following formula: In the formula S i The energy spectrum after filtering and smoothing is at the 1st i The counting of the Tao; O i+j For the intermediate energy spectrum at the 1st i + j The counting of the Tao; C j For Gaussian filtering functions with a window width of - k arrive k The coefficient within.

7. The method for processing energy spectrum of quasi-semi-spherical CdZnTe detector according to claim 5, wherein, The background subtraction processing adopts the SNIP algorithm.

8. The method for processing energy spectrum of quasi-semi-spherical CdZnTe detector according to claim 7, characterized in that, The background subtraction processing comprises: performing digital filtering on the smoothed energy spectrum to obtain a numerically transformed energy spectrum, determining a background spectrum to be stripped based on the numerically transformed energy spectrum, and performing numerical inverse transformation on the background spectrum to be stripped to obtain the target energy spectrum.

9. The method for energy spectrum processing of a quasi-semi-spherical CdZnTe detector according to claim 8, wherein, The smoothed energy spectrum is first subjected to digital filtering according to the following formula: In the formula, G i The energy spectrum after numerical transformation is shown in FIG.

4. i The counts of the channel.

10. The method for processing energy spectrum of quasi-semi-spherical CdZnTe detector according to claim 8, characterized in that, The background spectrum to be stripped is subjected to numerical inverse transformation according to the following formula: wherein b i counts in the first channel for the background spectrum to be peeled off, i counts in the first channel for the target spectrum, B i counts in the first channel for the background spectrum to be peeled off, i counts in the first channel for the target spectrum.

Images