A data processing method and system for neutron energy-resolved imaging and a storage medium

Abstract

This invention relates to the field of neutron imaging detection and nuclear electronics data processing technology, specifically to a data processing method, system, and storage medium for neutron energy-resolved imaging. After a neutron event occurs, this application clusters the hit anode bars in the X and Y directions within the global trigger window of the acquisition board based on the bar number and the difference in overthreshold arrival time, obtaining X candidate clusters and Y candidate clusters. Then, the position coordinates of each hit anode bar in the X and Y candidate clusters are calculated respectively; the flight time and energy of neutrons in a real neutron event are calculated; finally, a full-energy spectrum and a selected energy region map are generated and output based on the position coordinates of each hit anode bar in the X and Y candidate clusters, the neutron flight time, and the neutron energy, and are then output. Using the scheme of this application, a data processing architecture with lower data volume and higher real-time performance is achieved, making it more suitable for high-throughput, wide-spectrum neutron source experiments.

Description

Technical Field

[0001] This invention relates to the field of neutron imaging detection and nuclear electronics data processing technology, specifically to a data processing method, system, and storage medium for neutron energy-resolution imaging. Background Technology

[0002] Neutron energy-resolved imaging (ERI) is an advanced non-destructive testing technique capable of simultaneously acquiring spatial structure information and nuclide characteristic information of a sample. Compared to conventional transmission imaging, ERI not only records the neutron flux variation at each pixel location but also the time information of neutron arrival at the detector. It then uses the time-of-flight method to invert the neutron energy, thereby obtaining transmission maps, resonance absorption maps, and nuclide distribution maps in different energy ranges. For white-light neutron source experiments in the eV to MeV range, the detector needs to simultaneously possess high spatial resolution, high temporal resolution, high count rate tolerance, and high neutron detection efficiency.

[0003] Neutron-sensitive microchannel plates (nMCPs) are considered one of the most promising neutron energy-resolved imaging technologies due to their high gain, high time resolution, high position resolution, and event-specific readout capabilities. Currently, nMCP detectors combined with cross-anode strip arrays can achieve high pixel-equivalent resolution with fewer readout channels and obtain positional accuracy superior to strip width using the charge centroid method. However, most existing data processing methods revolve around waveform sampling, relying on peak extraction, baseline estimation, and offline reconstruction of the entire waveform. On the one hand, these methods involve large data volumes, complex processing chains, and limited real-time performance, which is detrimental to long-term stable operation on high-flux white-light neutron sources. On the other hand, waveform modes are prone to trigger ambiguity, threshold bias, and incomplete neutron event collection when the count rate increases, multiple hits occur frequently, and the electronic dynamic range is limited. Therefore, establishing a systematic data processing method that can operate stably for a long time without relying on the entire waveform output, without affecting neutron event collection, and can improve detection efficiency in the high-energy region has become a key issue for nMCP energy-resolved imaging to move from experimental verification to stable engineering applications. Summary of the Invention

[0004] The present invention provides a data processing method, system and storage medium for neutron energy-resolved imaging, which effectively solves the problem that existing neutron energy-resolved imaging data processing systems cannot simultaneously achieve high-purity neutron event identification, position resolution better than stripe width and accurate time-of-flight energy resolution under low data throughput pressure.

[0005] According to a first aspect, one embodiment provides a data processing method for neutron energy-resolved imaging, comprising:

[0006] After a neutron event occurs, the event stream information in the acquisition board is obtained; the event stream information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number;

[0007] Within the global trigger window of the acquisition board, the anode bars in the X and Y directions are clustered according to the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars, to obtain at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment;

[0008] A joint discrimination strategy is adopted to pair and judge the X candidate clusters and Y candidate clusters that appear in each trigger window to obtain the real neutron events in each trigger window;

[0009] Obtain the center position of each hit anode strip in the X candidate cluster in the real neutron event, and the weight of each hit anode strip after threshold duration correction. Calculate the position coordinates of each hit anode strip in the X candidate cluster based on the center position of each hit anode strip in the X candidate cluster and the weight of each hit anode strip after threshold duration correction.

[0010] Obtain the center position of each hit anode strip in the Y candidate cluster in the real neutron event, and the weight of each hit anode strip after threshold duration correction. Calculate the position coordinates of each hit anode strip in the Y candidate cluster based on the center position of each hit anode strip in the Y candidate cluster and the weight of each hit anode strip after threshold duration correction.

[0011] The start time and threshold arrival time of the neutrons in the actual neutron event of the acquisition board are obtained, and the flight time of the neutrons in the actual neutron event is calculated based on the start time and threshold arrival time.

[0012] The energy of the neutron is calculated based on its flight time and flight path.

[0013] Based on the real neutron event, the system outputs the position coordinates of each hit anode bar in the X candidate cluster, the position coordinates of each hit anode bar in the Y candidate cluster, the flight time of the neutron, and the full-energy spectrum and energy selection region map of the neutron energy generation.

[0014] In one embodiment, after acquiring the event stream information from the acquisition board, the method further includes:

[0015] The acquisition board is clock-synchronized to form a unified global time series.

[0016] In one embodiment, the clustering of anode bars in the X and Y directions based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars includes:

[0017] Obtain the bar number of the hit anode bar in the X direction within each trigger window, assign the bar number of the hit anode bar in the X direction within each trigger window according to a preset proximity relationship, and determine whether there is a hit anode bar in the X direction among the bar numbers that meets the preset assignment conditions;

[0018] Obtain the threshold arrival time of the hit anode bar in the X direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the X direction, and determine whether the difference is less than the first preset time threshold width;

[0019] The bar numbers that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width are grouped into the same candidate cluster, denoted as the X candidate cluster;

[0020] Obtain the bar number of the hit anode bar in the Y direction within each trigger window, allocate the bar number of the hit anode bar in the Y direction within each trigger window according to a preset proximity relationship, and determine whether there is a hit anode bar in the Y direction among the bar numbers that meets the preset allocation conditions;

[0021] Obtain the threshold arrival time of the hit anode bar in the Y direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the Y direction, and determine whether the difference is less than the first preset time threshold width;

[0022] The bar numbers that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width are grouped into the same candidate cluster, denoted as the Y candidate cluster.

[0023] In one embodiment, the clustering of anode bars in the X and Y directions based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars further includes:

[0024] The X-direction hit anode bars that do not meet the preset allocation conditions, or the X-direction hit anode bars whose difference in arrival time is greater than the preset time gate width, are regarded as independent candidate clusters.

[0025] The Y-direction hit anode bars that do not meet the preset allocation conditions, or the Y-direction hit anode bars whose difference in arrival time is greater than the preset time gate width, are regarded as independent candidate clusters.

[0026] In one embodiment, the method of employing a joint discrimination strategy to pair and judge the X candidate clusters and Y candidate clusters appearing in each trigger window to obtain the real neutron events in each trigger window includes:

[0027] When the same trigger window includes multiple X candidate clusters and multiple Y candidate clusters, determine whether the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster is less than the second preset time gate width.

[0028] If so, determine whether the ratio between the sum of the threshold-crossing durations of the X candidate clusters and the sum of the threshold-crossing durations of the Y candidate clusters is within a preset range;

[0029] If so, determine whether the maximum position and cluster width of the threshold-crossing duration of candidate X cluster, and the maximum position and cluster width of candidate Y cluster are within the preset threshold.

[0030] If so, then it is determined that a real neutron event has occurred within the trigger window; and / or,

[0031] When the same trigger window includes one X candidate cluster and multiple Y candidate clusters, or the same trigger window includes multiple X candidate clusters and one Y candidate cluster, the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster, the ratio between the sum of the threshold arrival times of the X candidate clusters and the sum of the threshold arrival times of the Y candidate clusters, the cluster width of the X candidate cluster, and the cluster width of the Y candidate cluster are substituted into the preset cost function for calculation.

[0032] The combination of candidate clusters X and Y with the lowest cost is selected as the actual neutron event occurring within the trigger window.

[0033] In one embodiment, after obtaining the actual neutron events within each trigger window, the method further includes:

[0034] Calculate the number of secret marks and the discrimination score for each of the real neutron events;

[0035] Determine whether the discrimination score of the secret mark number is higher than the preset discrimination score threshold. If so, remove or mark the real neutron event corresponding to the discrimination score of the secret mark number as a low confidence event.

[0036] In one embodiment, after obtaining the actual neutron events within each trigger window, the method further includes:

[0037] The real neutron events are input into a preset saturation working area, and the real neutron events are output in a saturation manner so that the overthreshold duration in the real neutron events tends to be consistent.

[0038] In one embodiment, calculating the flight time of the neutron in the real neutron event based on the start time and the overthreshold arrival time includes:

[0039] The start time of the hit anode bar in the real neutron event was periodically corrected using γ-flash to obtain the corrected start time;

[0040] The threshold arrival time is segmented and compensated based on the threshold duration of the hit anode bar in the real neutron event to obtain the corrected threshold arrival time.

[0041] The flight time of neutrons in the real neutron event is calculated based on the corrected overthreshold arrival time and the corrected start time.

[0042] In one embodiment, after obtaining the actual neutron events within each trigger window, the method further includes:

[0043] The flight time of the neutron is calculated based on the threshold arrival time and start time of each of the said hit anode bars;

[0044] The neutron events are divided into different energy ranges based on the flight time, and the effective high-energy neutron events in the neutron events are evaluated by a preset high-energy response weight model.

[0045] According to a second aspect, one embodiment provides a data processing system for neutron energy-resolution imaging, comprising:

[0046] The event stream information acquisition module is used to acquire event stream information from the acquisition board after a neutron event occurs; the event stream information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number;

[0047] The clustering module is used to cluster the anode bars in the X and Y directions within the global trigger window of the acquisition board, based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars, to obtain at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment;

[0048] The neutron event judgment module is used to perform pairing judgment on the X candidate clusters and Y candidate clusters appearing in each trigger window using a joint discrimination strategy, so as to obtain the real neutron events in each trigger window;

[0049] The X-coordinate reconstruction module is used to obtain the center position of each hit anode strip in the X candidate cluster in the real neutron event, the weight of each hit anode strip after threshold duration correction, and to calculate the position coordinates of each hit anode strip in the X candidate cluster based on the center position of each hit anode strip in the X candidate cluster and the weight of each hit anode strip after threshold duration correction.

[0050] The Y-coordinate reconstruction module is used to obtain the center position of each hit anode strip in the Y candidate cluster in the real neutron event, the weight of each hit anode strip after threshold duration correction, and to calculate the position coordinates of each hit anode strip in the Y candidate cluster based on the center position of each hit anode strip in the Y candidate cluster and the weight of each hit anode strip after threshold duration correction.

[0051] The flight time calculation module is used to obtain the start time and threshold arrival time of the neutron in the actual neutron event of the acquisition board, and calculate the flight time of the neutron in the actual neutron event based on the start time and the threshold arrival time.

[0052] An energy calculation module is used to calculate the energy of the neutron based on its flight time and flight path.

[0053] The processing module is used to output the following data based on the real neutron event: the position coordinates of each hit anode bar in the X candidate cluster, the position coordinates of each hit anode bar in the Y candidate cluster, the flight time of the neutron, and the energy generation full-energy spectrum and energy selection region map of the neutron.

[0054] According to a third aspect, one embodiment provides a computer-readable storage medium including a program that can be executed by a processor to perform the methods as described in any of the embodiments herein.

[0055] According to the above embodiment of a neutron energy-resolved imaging data processing method / system, after a neutron event occurs, the bar number, threshold arrival time, threshold duration, and trigger batch number of each hit anode bar are acquired. Then, within the global trigger window of the acquisition board, the hit anode bars in the X and Y directions are clustered according to the difference between the hit anode bar number and the threshold arrival time to obtain X candidate clusters and Y candidate clusters. Subsequently, the actual neutron event occurring within each trigger window is obtained based on the X candidate clusters and Y candidate clusters appearing within each trigger window. Subsequently, based on the center positions of candidate clusters X and Y in each hit anode bar in real neutron events and the weighted after correction for the overthreshold duration of each hit anode bar, the position coordinates of each hit anode bar in candidate cluster X and candidate cluster Y are calculated respectively. The flight time of neutrons in real neutron events is calculated based on the start time and overthreshold arrival time of the hit anode bars. The energy of the neutrons is calculated based on their flight time and flight path. Finally, a full-energy spectrum and a selected energy region map are generated and output based on the position coordinates of each hit anode bar in candidate cluster X, the position coordinates of each hit anode bar in candidate cluster Y, the neutron flight time, and the neutron energy. The above-described scheme of this application provides a real-time or near-real-time data processing method that does not rely on full-waveform output but uses TOA+TOT event streams for processing. This allows for high-purity neutron event identification, position resolution superior to bar width, accurate flight time, and energy resolution simultaneously under lower data throughput pressure. It achieves a data processing architecture with lower data volume and higher real-time performance, making it more suitable for high-throughput, wide-spectrum neutron source experiments. Attached Figure Description

[0056] Figure 1 This is a flowchart of a data processing method for neutron energy-resolution imaging according to this embodiment;

[0057] Figure 2 This is a flowchart illustrating the clustering of anode strips in the X and Y directions according to this embodiment;

[0058] Figure 3 This is a flowchart illustrating the calculation of neutron flight time in a real neutron event in this embodiment;

[0059] Figure 4 This is a graph (signal waveform diagram) showing the change in signal amplitude over time after a neutron hits the anode bar in this embodiment.

[0060] Figure 5 This is a schematic diagram of the adjacent cluster signal features in this embodiment, wherein (a) is a schematic diagram of the adjacent cluster signal visualized in three-dimensional space; and (b) is a schematic diagram of the adjacent cluster signal visualized in planar space.

[0061] Figure 6This is a schematic diagram of the XY cluster hit in this embodiment;

[0062] Figure 7 This is a schematic diagram of the detector system when the neutron beam spot impacts the detector in this embodiment;

[0063] Figure 8 This is a schematic diagram of the structure of a data processing system for neutron energy-resolution imaging according to this embodiment.

[0064] Figure reference numerals: 10, Event Stream Information Acquisition Module; 20, Clustering Module; 30, Neutron Event Judgment Module; 40, X-Coordinate Reconstruction Module; 50, Y-Coordinate Reconstruction Module; 60, Time-of-Flight Calculation Module; 70, Energy Calculation Module; 80, Processing Module. Detailed Implementation

[0065] The present invention will now be described in further detail with reference to specific embodiments and accompanying drawings. Similar elements in different embodiments are referred to by associated similar element reference numerals. In the following embodiments, many details are described to facilitate a better understanding of the invention. However, those skilled in the art will readily recognize that some features may be omitted in different situations, or may be replaced by other elements, materials, or methods. In some cases, certain operations related to the present invention are not shown or described in the specification. This is to avoid obscuring the core parts of the invention with excessive description. For those skilled in the art, detailed description of these related operations is not necessary; they can fully understand the related operations based on the description in the specification and general technical knowledge in the art.

[0066] Furthermore, the features, operations, or characteristics described in the specification can be combined in any suitable manner to form various embodiments. At the same time, the steps or actions in the method description can be rearranged or adjusted in a manner obvious to those skilled in the art. Therefore, the various orders in the specification and drawings are only for the clear description of a particular embodiment and do not imply a necessary order, unless otherwise stated that a particular order must be followed.

[0067] The serial numbers assigned to components in this document, such as "first" and "second," are used only to distinguish the described objects and have no sequential or technical meaning. The terms "connection" and "linkage" used in this invention, unless otherwise specified, include both direct and indirect connections (linkages).

[0068] In previous nMCP energy-resolved imaging experiments, the applicant discovered several key common problems with cross-anode strip array readout. First, the amplitude range of neutron events is extremely wide. Due to differences in the primary electron generation location, multiplication path, and subsequent gain of ordinary MCPs in nMCPs, the same type of real neutron event may exhibit pulse amplitudes differing by tens, hundreds, or even higher factors at the output. If linear amplitude precise readout is still used, it will place excessive demands on the dynamic range of the front-end electronics, resulting in both small signal omissions and large signal saturation. Second, MCP dark counts are difficult to distinguish directly from real neutron events. Dark counts originate from random single-electron emission and hot electron multiplication within the MCP, and their temporal, positional, and amplitude distributions overlap to some extent with low-amplitude neutron events. Improper handling can significantly contaminate the TOF spectrum and energy-selective imaging results. Third, in the low-energy neutron region with high count rates, multiple-hit events increase significantly. Within the same sampling time window, there may be one X cluster corresponding to multiple Y clusters, multiple X clusters corresponding to one Y cluster, or multiple candidate clusters appearing in both the X and Y directions. Simply selecting the highest peak or retaining only single-peak events will result in the loss of low-energy events and image bias. Fourth, high-energy neutrons above keV, especially above 30keV, may not only interact with the 10B conversion layer but also undergo scattering or inelastic reactions with components such as Si, O, Pb, and Bi in the MCP matrix, forming response characteristics different from those in the low-energy region. If this high-energy response cannot be identified and utilized appropriately in data processing, the effective detection efficiency of nMCP in the high-energy region will be limited.

[0069] Furthermore, the goal of the cross-anode bar readout system is not simply to provide presence / absence counts, but rather to reconstruct a position resolution superior to the anode bar width and recover an accurate time-of-flight energy spectrum, even in TOA (Time of Arrival) and TOT (Time over Threshold) data modes. (Reference) Figure 4 Time of arrival (TOA) reflects the pulse arrival time and can be directly involved in time-of-flight (TOF) reconstruction. Time of transit (TOT) is related to pulse charge, local gain, and electron cloud coupling width. Although it is not equivalent to linear amplitude, after normalization, clustering, and calibration, it can still serve as an important basis for location reconstruction and event identification. Therefore, how to establish a systematic data processing method that does not rely on full waveform output but is based on TOA+TOT event stream for multiple hit differentiation, dark count suppression, amplitude dynamic range compression, and high-energy efficiency improvement has become a key issue for nMCP energy-resolved imaging to move from experimental verification to stable engineering applications.

[0070] In view of this, this application proposes a data processing method for neutron energy-resolved imaging to solve the aforementioned problems. The specific implementation scheme is described below.

[0071] refer to Figure 1 This embodiment provides a data processing method for neutron energy-resolved imaging, which includes the following steps:

[0072] Step 100: After a neutron event occurs, the event flow information acquisition module acquires the event flow information from the acquisition board; the event flow information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number.

[0073] Specifically, after a neutron event occurs, the front-end electronics readout unit is connected to the cross-anode strip array. This unit is responsible for recording the strip number, TOA, TOT, and trigger batch number for each hit anode strip in real time, and transmitting this information to the back-end data processing unit as a digital event stream. The back-end data processing unit consists of a clustering module, a neutron event discrimination module, an X-coordinate reconstruction module, a Y-coordinate reconstruction module, a time-of-flight calculation module, an energy calculation module, and a processing module, all sequentially connected to form a complete data processing pipeline. Each module can be implemented in real-time or near-real-time on a field-programmable gate array (FPGA), a digital signal processor (DSP), or a general-purpose processor. By replacing full waveform sampling with a TOA+TOT discrete event stream as the data processing input, the front-end electronics only needs to record the strip number, TOA, TOT, and trigger batch number, without needing to output the complete waveform. This reduces data throughput by more than 90% compared to the full waveform method, achieving a real-time or near-real-time data processing architecture and significantly improving the system's ability to operate stably for extended periods on a high-throughput, wide-spectrum neutron source.

[0074] Step 200: Within the global trigger window of the acquisition board, the clustering module clusters the anode bars in the X and Y directions according to the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars, to obtain at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment.

[0075] Specifically, before performing this step, the acquisition boards need to be clock-synchronized to form a unified global time series. Specifically, the time correction module receives the bar number, TOA, TOT, and trigger batch number of each hit anode bar from the front-end electronics readout unit. This module first performs unified clock alignment on the timestamps of different acquisition boards to eliminate inter-board synchronization offset; then, based on the channel delay calibration table, it independently corrects the inherent delay and threshold timing error of each channel, forming a global time series. After the global time series is established, the time correction module corrects T0 using gamma-ray flashes or a known reference target to obtain the Time of Flight (TOF) for each event. T0 correction employs a periodic online calibration mechanism, updating the T0 reference value periodically by receiving gamma-ray flash signals or reference target signals to suppress the impact of clock drift and long-term inter-channel deviations on TOF accuracy. The time correction module outputs the corrected global time series and TOF to the clustering module and energy reconstruction module.

[0076] refer to Figure 2 The specific clustering method is as follows: Step 210: Obtain the bar number of the hit anode bar in the X direction within each trigger window, allocate the bar number of the hit anode bar in the X direction within each trigger window according to the preset proximity relationship, and determine whether there is a hit anode bar in the X direction that meets the preset allocation conditions among the bar numbers.

[0077] Step 220: Obtain the threshold arrival time of the hit anode bar in the X direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the X direction, and determine whether the difference is less than the first preset time gate width;

[0078] Step 230: Group the hit anode bars in the X direction that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width into the same candidate cluster, and denot it as the X candidate cluster.

[0079] Step 240: Obtain the bar number of the hit anode bar in the Y direction within each trigger window, assign the bar number of the hit anode bar in the Y direction within each trigger window according to the preset proximity relationship, and determine whether there is a hit anode bar in the Y direction that meets the preset assignment conditions among the bar numbers;

[0080] Step 250: Obtain the threshold arrival time of the hit anode bar in the Y direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the Y direction, and determine whether the difference is less than the first preset time threshold width;

[0081] Step 260: Group the hit anode bars in the Y direction that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width into the same candidate cluster, and denot them as the Y candidate cluster.

[0082] After clock calibration of the acquisition board, the clustering module independently clusters the hit anode strips in the X and Y directions according to both spatial and temporal proximity criteria, using a single one-dimensional clustering method. Figure 6 Specifically, the clustering module can be signal-connected to the aforementioned time correction module to receive the global time series within each global trigger window. Multiple hit bars with adjacent bar numbers and a TOA difference less than a preset first time gate width are grouped into the same candidate cluster.

[0083] In addition, the X-direction hit anode bars that do not meet the preset allocation conditions, or the X-direction hit anode bars whose threshold arrival time difference is greater than the preset time gate width, are regarded as independent candidate clusters; the Y-direction hit anode bars that do not meet the preset allocation conditions, or the Y-direction hit anode bars whose threshold arrival time difference is greater than the preset time gate width, are regarded as independent candidate clusters.

[0084] Specifically, hit bars with isolated bar numbers or significantly deviated TOA are considered independent candidate clusters. For each candidate cluster, the clustering module calculates six attributes: cluster center bar number, total TOT, maximum TOT, TOT-weighted average TOA, cluster width, and number of bars within the cluster. These six attributes compress the raw multi-channel hit information into a structured feature vector for subsequent multi-hit discrimination and location reconstruction. The clustering module outputs a list of candidate clusters in the X direction (X candidate clusters) and a list of candidate clusters in the Y direction (Y candidate clusters) to the neutron event judgment module.

[0085] Step 300: The neutron event judgment module uses a joint discrimination strategy to pair and judge the X candidate clusters and Y candidate clusters that appear in each trigger window to obtain the real neutron events in each trigger window.

[0086] Specifically, the neutron event judgment module receives the candidate cluster list in the X direction (X candidate cluster) and the candidate cluster list in the Y direction (Y candidate cluster), and uses a joint discrimination strategy of time consistency, TOT symmetry, cluster width consistency and local topology constraints to distinguish multiple hits and perform optimal X / Y pairing.

[0087] Specifically, the actual neutron events within each trigger window are obtained in the following manner:

[0088] When the same trigger window includes multiple X candidate clusters and multiple Y candidate clusters, determine whether the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster is less than the second preset time gate width.

[0089] If so, determine whether the ratio between the sum of the threshold-crossing durations of candidate X clusters and the sum of the threshold-crossing durations of candidate Y clusters is within a preset range;

[0090] If so, determine whether the maximum position and cluster width of the threshold-crossing duration of candidate X cluster, and the maximum position and cluster width of the threshold-crossing duration of candidate Y cluster are within the preset threshold.

[0091] If so, then it is determined that a real neutron event has occurred within the trigger window; and / or,

[0092] When the same trigger window includes one X candidate cluster and multiple Y candidate clusters, or the same trigger window includes multiple X candidate clusters and one Y candidate cluster, the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster, the ratio between the sum of the threshold arrival times of the X candidate clusters and the sum of the threshold arrival times of the Y candidate clusters, the cluster width of the X candidate cluster, and the cluster width of the Y candidate cluster are substituted into the preset cost function for calculation.

[0093] The combination of candidate clusters X and Y with the lowest cost is selected as the actual neutron event occurring within the trigger window.

[0094] Specifically, such as Figure 5 As shown in Figure b (where the relationship between TOT value and signal amplitude is clearly visible: the larger the amplitude, the larger the TOT), when multiple X candidate clusters and multiple Y candidate clusters are included within the same trigger window, the threshold arrival times of multiple X candidate clusters are basically the same, and the threshold arrival times of multiple Y candidate clusters are also basically the same. Therefore, candidate matching pairs are first established using the TOA difference between X candidate clusters and Y candidate clusters as the primary criterion, requiring the TOA difference to be less than the second preset time threshold, which is usually set to less than 3ns. For candidate pairs that meet the TOA difference condition, the ratio of the sum of TOT in both directions and the cluster width difference are further compared, and candidate pairs with ratios within the allowable range and cluster width differences not exceeding the preset threshold are retained. When the same trigger window includes one X candidate cluster and multiple Y candidate clusters, or multiple X candidate clusters and one Y candidate cluster, the matching cost function combining the TOA residual, TOT ratio deviation, and cluster width deviation is calculated, and the X / Y combination with the smallest total cost is selected as the valid case. Cases that still cannot be uniquely determined are marked as multiple overlapping cases and retained with reduced weight, rather than being directly deleted. The above scheme is particularly suitable for multiple hits under high count rate conditions. Compared with the traditional approach of "only taking the highest peak" or "only retaining single-peak events", this scheme significantly reduces event loss and improves effective statistics and energy spectrum fidelity.

[0095] Step 400: The X-coordinate reconstruction module obtains the center position of each hit anode strip in the X candidate cluster of the real neutron event and the weight of the overthreshold duration of each hit anode strip after correction. Based on the center position of each hit anode strip in the X candidate cluster and the weight of the overthreshold duration of each hit anode strip after correction, the position coordinates of each hit anode strip in the X candidate cluster are calculated.

[0096] Specifically, using the corrected TOT as the local charge weight and TOA consistency as the screening constraint, a weighted centering calculation is performed on the candidate clusters in each direction. Suppose a candidate cluster in a certain direction contains k adjacent hits, and the center position of the i-th hit is... The corrected weights are Then the reconstructed coordinates in that direction are Before performing the weighted center of gravity calculation, the X-coordinate reconstruction module also calculates the TOA dispersion of each bar within the cluster and removes abnormal bars that deviate from the cluster average TOA by more than a preset threshold, in order to avoid occasional noise bars pulling the center of gravity off.

[0097] Step 500: The Y-coordinate reconstruction module obtains the center position of each hit anode strip in the Y candidate cluster of the real neutron event and the weight of the overthreshold duration of each hit anode strip after correction. Based on the center position of each hit anode strip in the Y candidate cluster and the weight of the overthreshold duration of each hit anode strip after correction, the position coordinates of each hit anode strip in the Y candidate cluster are calculated.

[0098] Similar to the X-coordinate reconstruction method described above, weighted centering calculations are performed on candidate clusters in each direction, using the corrected TOT as the local charge weight and TOA consistency as the screening constraint. Let a candidate cluster in a certain direction contain k adjacent hits, with the center position of the i-th hit being... The corrected weights are Then the reconstructed coordinates in that direction are Before performing the weighted center of gravity calculation, the Y-coordinate reconstruction module also calculates the TOA dispersion of each bar within the cluster and removes abnormal bars that deviate from the cluster average TOA by more than a preset threshold, in order to avoid occasional noise bars pulling the center of gravity off.

[0099] Using the TOT (Total Optical Trace) correction centroid method and TOA (Total Optical Aspect) consistency screening described above, even with an anode strip width of 0.4 mm to 0.5 mm, positional resolution superior to that of the strip width can be obtained. (Reference) Figure 7 The physical basis for this is that after the electron cloud reaches the anode surface, it usually spans multiple adjacent stripes, and the TOT of different stripes statistically reflects the local weight of that distribution. As long as sufficient relative weight information is retained and outliers are suppressed, substrip width localization can be achieved.

[0100] Step 600: The flight time calculation module obtains the start time and over-threshold arrival time of the neutrons hitting the anode bar in the real neutron event on the acquisition board, and calculates the flight time of the neutrons in the real neutron event based on the start time and over-threshold arrival time.

[0101] For details, please refer to Figure 3 The flight time of neutrons in real neutron events is calculated based on the start time and the time of arrival above the threshold. The specific steps include the following:

[0102] Step 610: Use γ-flash to periodically correct the start time of the anode bar hit in real neutron events to obtain the corrected start time;

[0103] Step 620: Based on the overthreshold duration of the neutron event hitting the anode bar, the overthreshold arrival time is segmented and compensated to obtain the corrected overthreshold arrival time.

[0104] Step 630: Calculate the flight time of neutrons in real neutron events based on the corrected overthreshold arrival time and the corrected start time.

[0105] After establishing the global time series, T0 is corrected using a γ-flash or a known reference target to obtain the Time of Flight (TOF) for each event. T0 correction employs a periodic online calibration mechanism, periodically updating the T0 reference value by receiving γ-flash or reference target signals to suppress the impact of clock drift and long-term inter-channel deviations on TOF accuracy. Specifically, the neutron time of flight (TOF) is obtained by calculating the difference between the corrected TOA and the corrected start time (TO). This mechanism enables the system to continuously suppress clock drift and inter-channel deviations during long-term continuous acquisition, maintaining long-term stability of TOF accuracy.

[0106] Step 700: The energy calculation module calculates the energy of the neutron based on its flight time and flight path.

[0107] Specifically, firstly, TOT-related compensation is performed on the TOA threshold travel time error: based on the magnitude of the corrected TOT, a quadratic polynomial or piecewise linear compensation is applied to the TOA to correct the system bias introduced by the threshold travel time error as the pulse amplitude changes, thus obtaining the compensated TOA. Then, the neutron energy E is calculated using the compensated TOA and the known flight path L: in the non-relativistic energy region according to... Conversion: Relativistic correction terms are introduced in the high-energy region for conversion. When constructing the TOF spectrum, the event confidence level is used as the weight to reduce the contribution of multiple overlapping events and low-confidence dark count boundary events, resulting in a confidence-weighted TOF spectrum.

[0108] Through the above processing, the accurate TOF energy spectrum can be recovered without preserving the complete waveform, and can be used for energy zone integral imaging, resonance peak identification, and nuclide distribution analysis.

[0109] Step 800: The processing module, based on real neutron events, outputs the position coordinates of each hit anode bar in candidate cluster X, the position coordinates of each hit anode bar in candidate cluster Y, the flight time of neutrons, and the full-energy spectrum and energy selection region map of neutron generation.

[0110] Specifically, the processing module generates a structured output for each valid neutron event based on the processing results of each step. The structured output includes at least the X-coordinate, Y-coordinate, TOF (Time-of-Flight), energy E, event confidence level, and multiple hit flag. Based on this structured event list, a full-energy transmission spectrum, a selected-energy transmission spectrum, and a net neutron count map after dark count subtraction are generated for subsequent resonance absorption imaging and nuclide distribution analysis. Events can also be flexibly filtered by TOF window, TOT confidence interval, or high-energy response enhancement flag to generate elemental resonance imaging maps and region of interest (ROI) energy spectrum maps, allowing the same batch of acquired data to be used for multiple imaging modes and analytical purposes.

[0111] This application proposes a systematic data processing method and corresponding system for nMCP neutron energy-resolved imaging detection systems based on cross-anode bar arrays, replacing full waveform sampling with a discrete event stream of TOA+TOT. The core of this method lies in the following: the front-end electronics only record the bar number, over-threshold arrival time (TOA), over-threshold duration (TOT), and trigger batch number for each hit anode bar, without outputting the complete waveform; the back-end sequentially performs time reference correction, independent one-dimensional clustering in the X / Y directions, joint discrimination and optimal pairing of multiple hits, multidimensional suppression and statistical subtraction of dark counts, saturated dynamic range compression and TOT correction, sub-bar width position reconstruction, and TOF precise energy reconstruction, ultimately outputting a structured list of neutron events with coordinates, energy, confidence level, and various flags, and generating a full-energy spectrum transmission map, a selected energy region transmission map, and a net neutron count map. This achieves a data processing architecture with lower data volume and higher real-time performance, making it more suitable for high-throughput, wide-spectrum neutron source experiments.

[0112] In some implementations, after obtaining the actual neutron events within each trigger window, the method further includes:

[0113] Calculate the number of secret marks and the discrimination score for each real neutron event;

[0114] Determine whether the discrimination score of the secret mark number is higher than the preset discrimination score threshold. If so, remove or mark the real neutron event corresponding to the discrimination score of the secret mark number as a low confidence event.

[0115] This solution divides dark count processing into MCP electronic cleaning, voltage divider control, online discrimination, and statistical deduction. First, both nMCPs and ordinary MCPs undergo electronic cleaning and gain monitoring in the factory cleanroom. While maintaining minimal gain impairment, the dark count of the cleaned MCPs shows a significant decrease. Most dark counts originate from the nMCPs, with minor contributions from the MCPs. When distributing voltage to the nMCP components, the voltage divider for the nMCPs is typically controlled at around 500V-600V, depending on the volume resistance differences of different acquisition boards, far lower than the 900-1100V voltage divider for ordinary MCPs. By reducing the nMCP voltage divider and appropriately increasing the MCP voltage divider (within a safe threshold), the dynamic range of the dark count and neutron event signal output amplitude can be effectively reduced.

[0116] In practical applications, the dark count suppression module calculates the dark count discrimination score for each real neutron event. Specifically, the dark count suppression module receives a list of paired events and calculates a multi-dimensional dark count discrimination score for each candidate event. This discrimination score includes at least five dimensions: single-direction hit index (hit only in the X or Y direction, with no corresponding directional cluster), extremely low TOT index (center bar TOT is below the lower limit of the dark count reference distribution), random TOF index (TOF falls into the beam closure region or the TOF distribution is inconsistent with the beam structure), spatial hotspot index (the spatial location of the event coincides with the known dark count hotspot region), and TOF structure consistency index (the degree of inconsistency between the event's TOF and the reference neutron TOF spectrum). Events with discrimination scores higher than a preset discrimination score threshold are directly eliminated or marked as low-confidence events; for events with discrimination scores in the critical range, the net neutron event count is obtained by normalizing and statistically subtracting from the independently measured dark count baseline rate based on the acquisition time and effective area.

[0117] For systems employing saturated output structures, since the amplitude difference between dark counts and real neutron events is reduced, this scheme further utilizes "TOF structure consistency" to enhance the dark count suppression effect, that is, preferentially retaining events that simultaneously have X / Y bidirectional cluster hits and TOF falls within the beam allowable range.

[0118] In some implementations, after obtaining the actual neutron events within each trigger window, the method further includes:

[0119] Real neutron events are input into a preset saturated working area, and the real neutron events are output in a saturated manner to make the overthreshold duration of the real neutron events tend to be consistent.

[0120] In practical applications, the net neutron event list from the dark count suppression module is output to a preset saturation operating region. This region is used for interval correction of the TOT (Transmission of Tunneling) values: because the detector uses a three-stage MCP or equivalent high-gain structure operating in the saturated or quasi-saturated output region, the center bar TOT values ​​of neutron events with different original charge values ​​have been compressed into a narrower interval, but nonlinear errors introduced by saturation still exist. This region applies piecewise nonlinear compensation to the center bar TOT values ​​using a TOT interval calibration table, uniformly mapping the TOT values ​​of each event to a uniformly processed amplitude range. The corrected TOT list is then output to the X-coordinate reconstruction module, Y-coordinate reconstruction module, and energy calculation module.

[0121] The advantages of this scheme are twofold: firstly, it avoids the overload effect of excessively large pulses on linear electronics, allowing both small and large signals to be recorded, effectively preventing the loss of small signals; secondly, although the TOT is not a strictly linear amplitude, it can still be used for charge centroid position reconstruction and multiple hit detection after interval calibration. The saturated output of the MCP does not significantly affect the TOT distribution of independent events.

[0122] In some implementations, after obtaining the actual neutron events within each trigger window, the method further includes:

[0123] The flight time of the neutron is calculated based on the threshold arrival time and start time of each anode bar hit;

[0124] Neutron events are divided into different energy ranges based on flight time, and effective high-energy neutron events are evaluated using a pre-set high-energy response weighting model.

[0125] In practical applications, after obtaining the Time-of-Flight (TOF) and energy E for each event, the events are categorized into thermal neutron, cold neutron, eV, low keV, and high-energy regions above 30 keV based on their TOF. For events categorized as high-energy regions above 30 keV, a relaxed TOF lower limit and cluster width criterion are applied to the event acceptance strategy, allowing the retention of locally multi-cluster but temporally coherent candidate events to fully utilize the high-energy responses induced by Si, O, and heavy nuclei in the MCP matrix. A high-energy response weighting model integrating TOF, TOF, cluster width, and bidirectional simultaneous hit probability is established to estimate the confidence level of each event as a valid high-energy neutron event. For samples with known resonance peaks, valid high-energy responses are identified through TOF difference spectral correction between samples with and without samples, confirming the validity of the high-energy responses.

[0126] refer to Figure 8This embodiment provides a data processing system for neutron energy-resolved imaging, including an event flow information acquisition module 10, a clustering module 20, a neutron event judgment module 30, an X-coordinate reconstruction module 40, a Y-coordinate reconstruction module 50, a time-of-flight calculation module 60, an energy calculation module 70, and a processing module 80. The event flow information acquisition module 10 acquires event flow information from the acquisition board after a neutron event occurs; the event flow information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number. The clustering module 20 clusters the anode bars in the X and Y directions within the global trigger window of the acquisition board based on the bar number of the hit anode bars and the difference in threshold arrival times between the hit anode bars, obtaining at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment. The neutron event judgment module 30 is used to perform pairing judgment on the X candidate clusters and Y candidate clusters appearing in each trigger window using a joint discrimination strategy to obtain the real neutron events in each trigger window. The X coordinate reconstruction module 40 is used to obtain the center position of each hit anode strip in the X candidate cluster of the real neutron event and the weight of each hit anode strip after threshold duration correction, and calculate the position coordinates of each hit anode strip in the X candidate cluster based on the center position of each hit anode strip in the X candidate cluster and the weight of each hit anode strip after threshold duration correction. The Y coordinate reconstruction module 50 is used to obtain the center position of each hit anode strip in the Y candidate cluster of the real neutron event and the weight of each hit anode strip after threshold duration correction, and calculate the position coordinates of each hit anode strip in the Y candidate cluster based on the center position of each hit anode strip in the Y candidate cluster and the weight of each hit anode strip after threshold duration correction. The time-of-flight calculation module 60 is used to acquire the start time and over-threshold arrival time of the neutrons hitting the anode strips in real neutron events on the acquisition board, and calculates the flight time of the neutrons in the real neutron events based on the start time and over-threshold arrival time. The energy calculation module 70 is used to calculate the energy of the neutrons based on the flight time and flight path of the neutrons. The processing module 80 is used to generate and output a full-energy spectrum and a selected energy region map based on the real neutron events, according to the position coordinates of each hitting anode strip in candidate cluster X, the position coordinates of each hitting anode strip in candidate cluster Y, the flight time of the neutrons, and the energy of the neutrons.

[0127] This embodiment presents a data processing system for neutron energy-resolved imaging, proposing a systematic data processing method and corresponding system that replaces full waveform sampling with a discrete event stream of Time of Arrival (TOA) and Time of Transmission (TOT). The core of this method lies in the following: the front-end electronics only record the strip number, time of arrival (TOA), duration of transmission (TOT), and trigger batch number for each hit anode strip, without outputting the complete waveform; the back-end sequentially performs time reference correction, independent one-dimensional clustering in the X / Y directions, joint discrimination and optimal pairing of multiple hits, multidimensional suppression and statistical subtraction of dark counts, saturated dynamic range compression and TOT correction, sub-strip width position reconstruction, and precise energy reconstruction using Time of Flight (TOF). Finally, it outputs a structured list of neutron events with coordinates, energy, confidence level, and various flags, and generates a full-energy transmission map, a selected-energy transmission map, and a net neutron count map. This achieves a data processing architecture with lower data volume and higher real-time performance, making it more suitable for high-throughput, wide-spectrum neutron source experiments.

[0128] Since the functions and roles of each module have been described in detail in the above embodiments of the data processing method for neutron energy-resolution imaging, this embodiment will not repeat them in detail here.

[0129] The following describes the method steps corresponding to each module of the above data processing system with reference to specific embodiments.

[0130] Taking a neutron energy-resolved imaging detector composed of 10B-doped nMCP and a cross-anode strip array as an example, the detector has 128 anode strips in each direction, with a strip width of 0.5 mm. The front-end electronics only output the strip number, TOA, and TOT, and do not output the complete waveform. The detector adopts a three-stage MCP structure, which allows neutron events to enter a controllable saturated or quasi-saturated operating region at the output end, compressing the TOT of the center strip of neutron events with original charge differences of tens to hundreds of times into a narrower processable range.

[0131] In the preprocessing stage of dark count suppression, nMCPs and ordinary MCPs underwent electronic cleaning preprocessing in a clean environment, and the cleaning effect was confirmed by gain monitoring. Regarding operating voltage allocation, the operating voltage allocated to the nMCP assembly was controlled between 500V and 600V, while the operating voltage allocated to a single ordinary MCP chip was controlled between 900V and 1100V. The significantly lower operating voltage of the nMCP compared to a single ordinary MCP chip effectively reduces the dark count rate of the nMCP while maintaining neutron detection gain, and further compresses the dynamic range of the neutron event signal output amplitude. Background data was acquired under beamless conditions to establish a multidimensional reference distribution of dark counts in TOA, TOT, spatial distribution, and cluster width distribution, serving as a benchmark for subsequent online discrimination.

[0132] At the start of the experiment, full-channel time alignment and T0 correction were performed using neutron source gamma-ray flash and sample-free beam data. The time correction module read the channel-specific delay calibration table and independently corrected the inherent delay, threshold timing error, and inter-plate synchronization offset of each channel to form a global time series. The T0 reference value was determined using the gamma-ray flash signal, and T0 was periodically updated by a periodic online correction trigger unit to suppress clock drift during long-term acquisition.

[0133] For each global trigger window, the clustering module 20 performs adjacent bar clustering on the hit anode bars in both the X and Y directions. Taking the X direction as an example: if the nth, (n+1)th, and (n+2)th bars are all hit, and the TOA difference between adjacent bars is less than the preset time threshold, then the three bars are merged into the same X-direction candidate cluster; if the mth bar is hit but none of its adjacent bars are hit, or its TOA difference with the adjacent hit bars exceeds the time threshold, then the mth bar forms an independent candidate cluster. For each candidate cluster, the cluster center bar number, total TOT, maximum TOT, weighted average TOA of TOT, cluster width, and number of bars within the cluster are calculated to form the X-direction candidate cluster list. The Y direction is processed in the same way to form the Y-direction candidate cluster list.

[0134] For cases with only one candidate cluster in the X direction and one candidate cluster in the Y direction, the multi-hit discrimination module directly verifies whether the TOA difference between the two directions satisfies the first time-gap constraint; if so, it directly pairs the cases as valid. For cases with multiple candidate clusters in the X direction or multiple candidate clusters in the Y direction, refer to... Figure 3 The multi-hit discrimination process shown is as follows: First, candidate matching pairs are established based on the condition that the TOA difference between the X cluster and the Y cluster is less than the first time threshold, which is usually set to less than 3ns. For each candidate matching pair, the ratio of the sum of TOT in both directions is compared to see if it is within the allowable range and whether the cluster width difference does not exceed the preset threshold. Candidate pairs that meet the conditions are retained. When there are still multiple candidate pairs retained, the cost function calculation unit calculates the matching cost of each candidate pair. The cost function is the weighted sum of the TOA residual, the TOT ratio deviation, and the cluster width deviation. The X / Y combination with the smallest total cost is selected as the valid case. For cases where the optimal pair cannot be uniquely determined after the above steps, the multi-overlap case deweighting marking unit marks them as multi-overlap cases and deweights them for retention, and they proceed to the subsequent processing flow.

[0135] The dark count suppression module calculates a five-dimensional dark count discrimination score for each candidate event. Taking the single-direction hit index as an example: if a candidate event has a candidate cluster hit only in the X direction and no corresponding candidate cluster in the Y direction, the single-direction hit index score is high, indicating that the event has a high dark count probability. Taking the TOF structure consistency index as an example: if the TOF of a candidate event falls into the beam-closing region, or its TOF value is inconsistent with the beam structure reference pattern, the TOF structure consistency index score is high. After combining the five-dimensional scores to obtain the total discrimination score, events exceeding the threshold are eliminated or marked as low-confidence events; for events with a total discrimination score in the critical range, the independently measured dark count baseline rate is normalized and statistically subtracted based on the acquisition time and effective area to obtain the net neutron event count. In the statistical subtraction step, events with simultaneous two-way cluster hits in both the X and Y directions and TOF falling within the beam-allowed range are preferentially retained, and TOF structure consistency is used to further distinguish dark counts from real neutron events. Dark counts typically produce hits in only one direction and have random Time-of-Flight (TOF) distributions, while real neutron events have physical characteristics of simultaneous X / Y bidirectional hits and TOF consistent with the beam structure. This difference can be used to further improve the dark count suppression effect.

[0136] The preset saturated working area receives the TOTs of each event in the net neutron event list. Piecewise nonlinear compensation is applied to the center bar TOTs through the TOT interval calibration table to compress the TOTs of the neutron events into a uniformly processed interval, and the corrected TOT list is output.

[0137] The coordinate reconstruction module uses the corrected TOT (Total Charge) as the local charge weight and TOA (Total Aspect) consistency as the screening constraint to perform weighted centroid calculation on candidate clusters in each direction. Before performing the weighted centroid calculation, the TOA dispersion of each line within the cluster is calculated, and hit lines that deviate from the cluster average TOA by more than a preset threshold are removed to reduce the centroid bias caused by occasional noise lines. Taking the X direction as an example, suppose a candidate cluster contains k adjacent hit lines, and the center position of the i-th line is... The corrected weights are Then the reconstructed coordinates in the X direction are The Y direction is calculated independently using the same formula. After reaching the anode surface, the electron cloud typically spans multiple adjacent strips. The TOT of different strips statistically reflects the local weight of this distribution. By retaining sufficient relative weight information and suppressing outliers, position resolution better than the strip width can be achieved on anode strips with a strip width of 0.4 mm to 0.5 mm.

[0138] The energy calculation module 70 performs TOT-related compensation on the TOA threshold travel time error: Based on the corrected TOT magnitude, a quadratic polynomial or piecewise linear compensation is applied to the TOA to correct the system bias introduced by the threshold travel time error as the pulse amplitude changes. After compensation, the neutron energy E is calculated using the corrected TOA and the known flight path L, and then calculated in the non-relativistic energy region according to... In this conversion, relativistic correction terms are introduced in the high-energy region. When constructing the TOF spectrum, the event confidence level is used as a weight to reduce the contributions of multiple overlapping events and low-confidence dark count boundary events, resulting in a confidence-weighted TOF spectrum.

[0139] For high-energy regions above 30 keV, the high-energy processing module 80 switches to a differentiated event acceptance strategy: based on the ordinary low-energy criteria, the lower limit of TOT is appropriately relaxed, allowing the retention of a small number of locally multi-cluster but temporally coherent candidate events; a high-energy response weight model integrating TOF, TOT, cluster width, and bidirectional simultaneous hit probability is established to estimate the confidence level of each high-energy event as a valid high-energy neutron event. For samples with known resonance peaks such as Fe, W, and Au, the validity of the high-energy response is confirmed through TOF difference spectral analysis with and without samples, further improving the reliability of the high-energy detection results. This processing transforms the high-energy responses induced by Si, O, and heavy nuclei in the MCP matrix from signals that would otherwise be easily discarded as background into effective detection information that can be modeled, discriminated, and utilized, improving the effective detection efficiency above 30 keV.

[0140] The processing module 80 generates a structured output for each valid neutron event based on the processing results of each step. The structured output includes at least the X-coordinate, Y-coordinate, TOF (Time-of-Flight), energy E, event confidence level, multiple hit flag, saturation correction flag, and high-energy response enhancement flag. Based on this structured event list, a full-energy transmission spectrum, a selected-energy transmission spectrum, and a net neutron count map after dark count subtraction are generated for subsequent resonance absorption imaging and nuclide distribution analysis. Events can also be flexibly filtered by TOF window, TOT confidence interval, or high-energy response enhancement flag to generate elemental resonance imaging maps and region of interest (ROI) energy spectrum maps, allowing the same batch of acquired data to be used for various imaging modes and analytical purposes.

[0141] In another implementation, the cross-anode bar array can be replaced with a rhombus-shaped pad coplanar readout structure, while still retaining the TOA+TOT event flow processing framework. In the clustering step, the proximity relationship of the pads is changed from rectangular adjacency to rhombus topological proximity, i.e., the four adjacent pads of the rhombus pad are used as the basis for determining proximity; the TOT weighting center formula remains unchanged, only the center position of each hit bar is changed. Replace with the center coordinates of the corresponding pad; the local topological constraint term in the multi-hit discrimination adjusts the proximity judgment logic according to the diamond pad layout. The above alternative implementation retains all the core functions of joint multi-hit discrimination, dark count multidimensional suppression, dynamic range compression, sub-resolution position reconstruction, and TOF energy spectrum reconstruction, and is suitable for imaging scenarios with higher requirements for position resolution isotropic.

[0142] In another implementation, the cost function in multi-hit discrimination can incorporate prior information about local speckle distribution or a machine learning classifier to assist in pairing decisions. Taking a machine learning classifier as an example, in the offline stage, the classifier is trained using labeled single-event and multi-hit event datasets. Input features include the TOA difference, TOT ratio, cluster width difference, and spatial location. In the online stage, the classifier outputs the probability value of each candidate pair, which is then weighted and fused with the cost function to select the optimal X / Y combination. A neural network classifier can also be introduced into the dark count discrimination score calculation, taking a five-dimensional feature vector as input and outputting a confidence score between 0 and 1, replacing or supplementing rule-based multi-dimensional discrimination score calculation. This extended scheme can further improve the accuracy of multi-hit pairing and the dark count suppression effect under experimental conditions of uneven speckle distribution or large fluctuations in dark count rate.

[0143] This embodiment provides a computer-readable storage medium including a program that can be executed by a processor to implement the method described in any of the embodiments herein. Since a data processing method for neutron energy-resolved imaging has been described in detail in the above embodiments, further elaboration is not required here.

[0144] In the above embodiments, implementation can be achieved, in whole or in part, by software, hardware, firmware, or any combination thereof. Furthermore, as those skilled in the art will understand, the principles herein can be reflected in a computer program product on a computer-readable storage medium pre-loaded with computer-readable program code. Any tangible, non-transitory computer-readable storage medium may be used, including magnetic storage devices (hard disks, floppy disks, etc.), optical storage devices (CDs, DVDs, Blu-ray discs, etc.), flash memory, and / or the like. These computer program instructions can be loaded onto a general-purpose computer, a special-purpose computer, or other programmable data processing apparatus to form a machine, such that instructions executing on the computer or other programmable data processing apparatus can generate means for performing a specified function. These computer program instructions can also be stored in a computer-readable storage medium that can instruct the computer or other programmable data processing apparatus to operate in a particular manner, such that instructions stored in the computer-readable storage medium can form an article of manufacture, including means for implementing the specified function. The computer program instructions can also be loaded onto a computer or other programmable data processing apparatus to perform a series of operational steps on the computer or other programmable apparatus to produce a computer-implemented process, such that instructions executing on the computer or other programmable apparatus can provide steps for implementing the specified function.

[0145] This document describes various exemplary embodiments with reference to them. However, those skilled in the art will recognize that changes and modifications can be made to the exemplary embodiments without departing from the scope of this document. For example, various operational steps and components for performing operational steps can be implemented in different ways depending on the specific application or considering any number of cost functions associated with the operation of the system (e.g., one or more steps can be deleted, modified, or combined with other steps).

[0146] While the principles herein have been illustrated in various embodiments, numerous modifications to the structures, arrangements, proportions, elements, materials, and components, particularly suited to specific environments and operational requirements, may be used without departing from the principles and scope of this disclosure. These modifications and other alterations or alterations will be included within the scope of this document. Those skilled in the art will recognize that many changes can be made to the details of the above embodiments without departing from the fundamental principles of the invention.

Claims

1. A data processing method for neutron energy-resolved imaging, characterized in that, include: After a neutron event occurs, acquire the event stream information from the acquisition board; The event stream information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number; Within the global trigger window of the acquisition board, the anode bars in the X and Y directions are clustered according to the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars, to obtain at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment; A joint discrimination strategy is adopted to pair and judge the X candidate clusters and Y candidate clusters that appear in each trigger window to obtain the real neutron events in each trigger window; Obtain the center position of each hit anode strip in the X candidate cluster in the real neutron event, and the weight of each hit anode strip after threshold duration correction. Calculate the position coordinates of each hit anode strip in the X candidate cluster based on the center position of each hit anode strip in the X candidate cluster and the weight of each hit anode strip after threshold duration correction. Obtain the center position of each hit anode strip in the Y candidate cluster in the real neutron event, and the weight of each hit anode strip after threshold duration correction. Calculate the position coordinates of each hit anode strip in the Y candidate cluster based on the center position of each hit anode strip in the Y candidate cluster and the weight of each hit anode strip after threshold duration correction. The start time and threshold arrival time of the neutrons in the actual neutron event of the acquisition board are obtained, and the flight time of the neutrons in the actual neutron event is calculated based on the start time and threshold arrival time. The energy of the neutron is calculated based on its flight time and flight path. Based on the real neutron event, the system outputs the position coordinates of each hit anode bar in the X candidate cluster, the position coordinates of each hit anode bar in the Y candidate cluster, the flight time of the neutron, and the full-energy spectrum and energy selection region map of the neutron energy generation.

2. The data processing method as described in claim 1, characterized in that, After acquiring the event stream information from the acquisition board, the process also includes: The acquisition board is clock-synchronized to form a unified global time series.

3. The data processing method as described in claim 1, characterized in that, The clustering of anode bars in the X and Y directions based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars includes: Obtain the bar number of the hit anode bar in the X direction within each trigger window, assign the bar number of the hit anode bar in the X direction within each trigger window according to a preset proximity relationship, and determine whether there is a hit anode bar in the X direction among the bar numbers that meets the preset assignment conditions; Obtain the threshold arrival time of the hit anode bar in the X direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the X direction, and determine whether the difference is less than the first preset time threshold width; The bar numbers that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width are grouped into the same candidate cluster, denoted as the X candidate cluster; Obtain the bar number of the hit anode bar in the Y direction within each trigger window, allocate the bar number of the hit anode bar in the Y direction within each trigger window according to a preset proximity relationship, and determine whether there is a hit anode bar in the Y direction among the bar numbers that meets the preset allocation conditions; Obtain the threshold arrival time of the hit anode bar in the Y direction within each trigger window, calculate the difference between the threshold arrival times of every two hit anode bars in the Y direction, and determine whether the difference is less than the first preset time threshold width; The bar numbers that meet the preset allocation conditions and whose threshold arrival time difference is less than the preset time gate width are grouped into the same candidate cluster, denoted as the Y candidate cluster.

4. The data processing method as described in claim 3, characterized in that, The clustering of anode bars in the X and Y directions based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars further includes: The X-direction hit anode bars that do not meet the preset allocation conditions, or the X-direction hit anode bars whose difference in arrival time is greater than the preset time gate width, are regarded as independent candidate clusters. The Y-direction hit anode bars that do not meet the preset allocation conditions, or the Y-direction hit anode bars whose difference in arrival time is greater than the preset time gate width, are regarded as independent candidate clusters.

5. The data processing method as described in claim 1, characterized in that, The method employs a joint discrimination strategy to pair and judge the X candidate clusters and Y candidate clusters appearing in each trigger window to obtain the actual neutron events in each trigger window, including: When the same trigger window includes multiple X candidate clusters and multiple Y candidate clusters, determine whether the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster is less than the second preset time gate width. If so, determine whether the ratio between the sum of the threshold-crossing durations of the X candidate clusters and the sum of the threshold-crossing durations of the Y candidate clusters is within a preset range; If so, determine whether the maximum position and cluster width of the threshold-crossing duration of candidate X cluster, and the maximum position and cluster width of candidate Y cluster are within the preset threshold. If so, then it is determined that a real neutron event has occurred within the trigger window; and / or, When the same trigger window includes one X candidate cluster and multiple Y candidate clusters, or the same trigger window includes multiple X candidate clusters and one Y candidate cluster, the difference between the threshold arrival time of the X candidate cluster and the threshold arrival time of the Y candidate cluster, the ratio between the sum of the threshold arrival times of the X candidate clusters and the sum of the threshold arrival times of the Y candidate clusters, the cluster width of the X candidate cluster, and the cluster width of the Y candidate cluster are substituted into the preset cost function for calculation. The combination of candidate clusters X and Y with the lowest cost is selected as the actual neutron event occurring within the trigger window.

6. The data processing method as described in claim 1, characterized in that, After obtaining the actual neutron events within each trigger window, the process also includes: Calculate the number of secret marks and the discrimination score for each of the real neutron events; Determine whether the discrimination score of the secret mark number is higher than the preset discrimination score threshold. If so, remove or mark the real neutron event corresponding to the discrimination score of the secret mark number as a low confidence event.

7. The data processing method as described in claim 1, characterized in that, After obtaining the actual neutron events within each trigger window, the process also includes: The real neutron events are input into a preset saturation working area, and the real neutron events are output in a saturation manner so that the overthreshold duration in the real neutron events tends to be consistent.

8. The data processing method as described in claim 1, characterized in that, The calculation of the neutron flight time in the real neutron event based on the start time and the threshold arrival time includes: The start time of the hit anode bar in the real neutron event was periodically corrected using γ-flash to obtain the corrected start time; The threshold arrival time is segmented and compensated based on the threshold duration of the hit anode bar in the real neutron event to obtain the corrected threshold arrival time. The flight time of neutrons in the real neutron event is calculated based on the corrected overthreshold arrival time and the corrected start time.

9. The data processing method as described in claim 1, characterized in that, After obtaining the actual neutron events within each trigger window, the process also includes: The flight time of the neutron is calculated based on the threshold arrival time and start time of each of the said hit anode bars; The neutron events are divided into different energy ranges based on the flight time, and the effective high-energy neutron events in the neutron events are evaluated by a preset high-energy response weight model.

10. A data processing system for neutron energy-resolution imaging, characterized in that, include: The event stream information acquisition module is used to acquire event stream information from the acquisition board after a neutron event occurs. The event stream information includes at least one of the following: the bar number of each hit anode bar, the threshold arrival time, the threshold duration, and the trigger batch number; The clustering module is used to cluster the anode bars in the X and Y directions within the global trigger window of the acquisition board, based on the bar number of the hit anode bar and the difference in the threshold arrival time between the hit anode bars, to obtain at least one X candidate cluster and / or at least one Y candidate cluster; the global trigger window is all time windows of the acquisition board at the same moment; The neutron event judgment module is used to perform pairing judgment on the X candidate clusters and Y candidate clusters appearing in each trigger window using a joint discrimination strategy, so as to obtain the real neutron events in each trigger window; The X-coordinate reconstruction module is used to obtain the center position of each hit anode strip in the X candidate cluster in the real neutron event, the weight of each hit anode strip after threshold duration correction, and to calculate the position coordinates of each hit anode strip in the X candidate cluster based on the center position of each hit anode strip in the X candidate cluster and the weight of each hit anode strip after threshold duration correction. The Y-coordinate reconstruction module is used to obtain the center position of each hit anode strip in the Y candidate cluster in the real neutron event, the weight of each hit anode strip after threshold duration correction, and to calculate the position coordinates of each hit anode strip in the Y candidate cluster based on the center position of each hit anode strip in the Y candidate cluster and the weight of each hit anode strip after threshold duration correction. The flight time calculation module is used to obtain the start time and threshold arrival time of the neutron in the actual neutron event of the acquisition board, and calculate the flight time of the neutron in the actual neutron event based on the start time and the threshold arrival time. An energy calculation module is used to calculate the energy of the neutron based on its flight time and flight path. The processing module is used to output the following data based on the real neutron event: the position coordinates of each hit anode bar in the X candidate cluster, the position coordinates of each hit anode bar in the Y candidate cluster, the flight time of the neutron, and the energy generation full-energy spectrum and energy selection region map of the neutron.

11. A computer-readable storage medium, characterized in that, Includes a program that can be executed by a processor to implement the method as described in any one of claims 1 to 9.

Images