CN114186166A

CN114186166A - Method and device for calculating reaction depth and computer readable storage medium

Info

Publication number: CN114186166A
Application number: CN202111338920.0A
Authority: CN
Inventors: 程冉; 肖鹏; 汪飞
Original assignee: Raycan Technology Co Ltd
Current assignee: Raycan Technology Co Ltd
Priority date: 2021-11-12
Filing date: 2021-11-12
Publication date: 2022-03-15
Anticipated expiration: 2041-11-12
Also published as: CN114186166B; WO2023082307A1

Abstract

The application provides a method, a device and a computer readable storage medium for calculating reaction depth, wherein the method comprises the steps of obtaining characteristic data related to photons from a photoelectric sensor/photoelectric sensor array; and inputting the characteristic data into a preset integrated learning model to calculate the reaction depth of the photons. According to some embodiments of the application, the scheme is simple, the accuracy rate is high, meanwhile, the FPGA is more convenient to use, and the method and the device are easier to popularize and use.

Description

Method and device for calculating reaction depth and computer readable storage medium

Technical Field

The present application relates to the field of data processing, and in particular, to a method and an apparatus for calculating a reaction depth, and a computer-readable storage medium.

Background

The response line in a PET (Position Emission Tomography, PET for short) system is a connection line for a pair of detectors that simultaneously receive gamma photons. Because the crystal has a certain length, for example, the PET of the small animal is 13mm, and the PET of the clinic is 20mm, the gamma photon has a certain reaction Depth, namely DOI (Depth of Interaction, DOI for short) information.

When no DOI information exists, the line of the response lines can only be determined as the line of the center of the detector surface, and this may cause the response lines to be positioned incorrectly, resulting in parallax effect, and affecting the spatial resolution of the system and the shape of the image. As shown in fig. 1a to 1c, parallax effect is caused due to uncertain depth of deposition of gamma photons in the crystal, resulting in low resolution and non-uniform resolution of PET images. Especially in the four-plate PET, the parallax effect between two adjacent plates is more obvious. A large number of research results show that the radial resolution gradually deteriorates from the center of the FOV (Field of View, FOV for short) to the edge of the FOV, which seriously affects the image resolution, and therefore the DOI measurement is very important for improving the PET image resolution.

Currently, DOI is calculated for a single strip of pixelated crystals, i.e. an array crystal, mainly by means of double-ended readout or single-ended readout.

The first two-terminal readout method is to use different photo sensors at both ends of the array crystal, for example, one end is coupled to PMT, and the other end is coupled to photodiode, and measure the reaction depth of gamma photons in the crystal by comparing the energy of the signals detected by the two photo sensors. Currently, it is commonly adopted to couple sipms (Silicon Photo-multiplexer) at both ends of the array crystal, process four sides of the crystal, for example, rough process, add a reflective layer, etc., and finally determine the DOI position according to the ratio of one-end energy to the sum of two-end energy.

For the single-ended readout method, a windowed optical sharing method is currently used in many cases. The method mainly utilizes the light distribution of crystals on SiPM to establish the relation with DOI. As shown in fig. 2, there is a light sharing window between crystal No. 3 and crystal No. 4 in the white position of the figure, and gamma photons strike crystal 3, and if deposited at the C position, visible light will propagate largely into crystal 4 and be detected by SiPM2, while little is detected by SiPM 1; if deposited at the A site, a significant amount of visible light will only propagate into the crystal 3 and be detected by the SiPM1, while almost none of the SiPM2 will be detected. At this point the light distribution over the SiPM array is no longer a point but is elongated, as shown in fig. 3a and 3 b. The DOI can be obtained by establishing a transfer function of the light distribution and the DOI.

Light distribution refers to the photon count on each SiPM in the SiPM array. As shown in fig. 4a and 4b, a gamma photon striking crystal array deposited at a location will have a light distribution on the coupled SiPM array, with the shallower the location (i.e., reaction depth) deposited, the broader the light distribution; the deeper the position of deposition, the more concentrated the light distribution.

The prior art methods of calculating DOI suffer from drawbacks, for example, the double-ended readout method doubles sipms and corresponding circuitry, and sipms and circuitry near one end of the FOV can affect the detection of gamma photons. The windowing light sharing method causes a small part of crystals to have no DOI capability due to the difference of windowing.

The method for calculating the DOI in the prior art increases the complexity of system design to a certain extent, needs extra space and has larger power consumption. For highly integrated small animal PET or PET/MR, however, the increased power consumption and additional space are not conducive to cost savings and optimized system design.

Disclosure of Invention

The present application provides a method, apparatus, and computer-readable storage medium for calculating a reaction depth, which addresses at least one of the problems set forth above.

According to an aspect of the present application, there is provided a method of calculating a reaction depth, the method including: acquiring characteristic data related to photons from a photosensor/photosensor array; and inputting the characteristic data into a preset integrated learning model to calculate the reaction depth of the photons.

According to some embodiments of the application, the method further comprises: and training the ensemble learning model by using sample data.

According to some embodiments of the application, the sample data includes feature data acquired from one or both ends of the photosensor/photosensor array.

According to some embodiments of the application, the sample data comprises feature data and reaction depth of the reaction depth.

According to some embodiments of the application, the method further comprises: combining a plurality of ensemble learning submodels to form the ensemble learning model.

According to some embodiments of the present application, the inputting the feature data into a preset ensemble learning model to calculate the reaction depth of the photon includes: in each integrated learning submodel, calculating by using the characteristic data respectively to generate the reaction depth of photons respectively; and calculating the reaction depth of the photons by voting or calculating an average value according to the calculation results of the integrated learning submodels.

According to some embodiments of the present application, the ensemble learning model is computed using one or more of a random forest algorithm, a boosted tree algorithm, or a gradient boosted tree algorithm.

According to some embodiments of the application, the characteristic data comprises photon corresponding energy information, pulse sampling point information, pulse rising edge time, pulse energy, pulse decay time and/or light distribution on the photosensor/photosensor array.

According to some embodiments of the present application, the acquiring photon-related feature data from a photosensor/photosensor array comprises: the characteristic data is acquired from the photosensor/photosensor array coupled to one or both ends of the scintillation crystal/scintillation crystal array.

According to some embodiments of the present application, the photosensor employs a PMT, SiPM; when the two ends of the scintillation crystal/scintillation crystal array are coupled with the photoelectric sensor/photoelectric sensor array, the two ends are simultaneously provided with PMT or SiPM, or one end is provided with PMT and the other end is provided with SiPM.

According to some embodiments of the application, the photons include high-energy photons of X-rays, gamma-rays, alpha-rays, beta-rays, neutron-rays.

According to an aspect of the present application, there is provided an apparatus for calculating a reaction depth, the apparatus including: the data acquisition unit is used for acquiring characteristic data related to photons from the photoelectric sensor/photoelectric sensor array; and the reaction depth calculating unit is used for inputting the characteristic data into a preset integrated learning model so as to calculate the reaction depth of the photons.

According to an aspect of the present application, there is provided an apparatus for calculating a reaction depth of a gamma photon, comprising one or more processors; a storage device for storing a computer program; the computer program, when executed by the one or more processors, causes the one or more processors to implement a method as in any one of the preceding.

According to an aspect of the application, a computer-readable storage medium is proposed, on which program instructions are stored, which program instructions, when executed, implement the method according to any of the preceding claims.

According to some example embodiments of the application, the photon deposition position in the crystal is subjected to reaction depth estimation through an integrated learning model by utilizing characteristic data related to photons, and the method is simple and easy to implement and has higher accuracy. And because the integrated learning model is simple to realize, the FPGA is more convenient to use to realize, and the integrated learning model is easier to popularize and use.

Drawings

In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings used in the description of the embodiments will be briefly introduced below.

FIG. 1a shows a schematic representation of response line repositioning according to simulated photon deposition coordinates.

FIG. 1b shows the reconstructed pattern at a reaction depth of 2 mm.

Fig. 1c shows the reconstructed image without the reaction depth information.

Figure 2 shows a schematic cross-sectional view of a detector employing a single-ended readout approach.

Fig. 3a shows the light distribution pattern of the SiPM array.

Figure 3b shows a photon count versus schematic diagram on a SiPM array.

Figure 4a shows a perspective view of the light distribution after gamma photons have been incident on the continuous crystal deposit.

Fig. 4b shows a schematic plan view of the light distribution on the coupled SiPM array after gamma photons have been incident on the continuous crystal.

FIG. 5 shows a flow chart of a method of calculating a depth of reaction according to an example embodiment of the present application.

Fig. 6 shows a decision tree structure.

Fig. 7a shows a random forest model according to an exemplary embodiment of the present application.

Fig. 7b shows a schematic diagram of estimating a reaction depth using a random forest model according to an exemplary embodiment of the present application.

FIG. 8 shows a flowchart of a method of training an ensemble learning model, according to an example embodiment of the present application.

Fig. 9 illustrates a screenshot of a method for obtaining sample data according to an exemplary embodiment of the present application.

FIG. 10 shows a block diagram of an apparatus for calculating a reaction depth according to an example embodiment of the present application.

FIG. 11 is a block diagram of another apparatus for calculating a reaction depth according to an embodiment of the present application.

Detailed Description

Example embodiments will now be described more fully with reference to the accompanying drawings. Example embodiments may, however, be embodied in many different forms and should not be construed as limited to the embodiments set forth herein; rather, these embodiments are provided so that this disclosure will be thorough and complete, and will fully convey the concept of example embodiments to those skilled in the art. The same reference numerals denote the same or similar parts in the drawings, and thus, a repetitive description thereof will be omitted.

The described features, structures, or characteristics may be combined in any suitable manner in one or more embodiments. In the following description, numerous specific details are provided to give a thorough understanding of embodiments of the disclosure. One skilled in the relevant art will recognize, however, that the embodiments of the disclosure can be practiced without one or more of the specific details, or with other means, components, materials, devices, or operations. In such cases, well-known structures, methods, devices, implementations, materials, or operations are not shown or described in detail.

The flow charts shown in the drawings are merely illustrative and do not necessarily include all of the contents and operations/steps, nor do they necessarily have to be performed in the order described. For example, some operations/steps may be decomposed, and some operations/steps may be combined or partially combined, so that the actual execution sequence may be changed according to the actual situation.

The terms "first," "second," and the like in the description and claims of the present application and in the above-described drawings are used for distinguishing between different objects and not for describing a particular order. Furthermore, the terms "include" and "have," as well as any variations thereof, are intended to cover non-exclusive inclusions. For example, a process, method, system, article, or apparatus that comprises a list of steps or elements is not limited to only those steps or elements listed, but may alternatively include other steps or elements not listed, or inherent to such process, method, article, or apparatus.

Specific embodiments according to the present application will be described in detail below with reference to the accompanying drawings.

FIG. 5 shows a flow chart of a method of calculating a depth of reaction according to an example embodiment of the present application. A method of calculating a reaction depth according to an exemplary embodiment of the present application will be described in detail with reference to fig. 5.

In step S501, characteristic data relating to photons is acquired from the photoelectric converter/photosensor array.

According to some embodiments of the present application, the photon-related characteristic data is acquired by a photo-electric converter/photo-sensor array in a detector comprising a scintillation crystal/scintillation crystal array and a photo-electric converter/photo-electric converter array coupled to at least one end of the scintillation crystal/scintillation crystal array, wherein the scintillation crystal converts photons to visible light, the photo-electric converter converts visible light to an electrical signal, and the photon-related characteristic data is acquired from the electrical signal.

According to some embodiments of the present application, the array of photoelectric converters comprises a PMT, an array of PMTs, an SiPM, an array of sipms, and individual elements of the array of photoelectric converters may be coupled with crystal elements of the array of scintillation crystals in a one-to-one, one-to-many manner. When the two ends of the scintillation crystal are both provided with photoelectric converters, the types of the photoelectric converters at the two ends can be the same or different, for example, both ends adopt PMT/SiPM, or one end adopts PMT and the other end adopts SiPM. Although PMTs perform less well than SiPMs, in certain applications, such as those where imaging quality requirements are not very high, the use of PMTs is more cost effective than SiPMs.

According to some example embodiments of the present application, the photons may be any high-energy photons capable of photoelectric conversion in the scintillation crystal, such as high-energy photons in X-rays, gamma-rays, alpha-rays, beta-rays, neutron rays, and the like.

According to some example embodiments of the present application, the acquired photon-related characteristic data includes energy information of gamma photons, pulse sampling point information, pulse rising edge time, pulse energy, pulse decay time, and/or light distribution over an array of photosensors, e.g., for gamma photons, whose characteristic energy is typically 511keV, the converted electrical pulse signal typically includes relatively fast rising edges and relatively slow falling edges.

According to some embodiments of the present application, photon-related characteristic data is obtained from electrical signals output by an array of photosensors at one or both ends of a scintillation crystal array.

In step S503, the photon-related feature data acquired in step S501 is input into a preset ensemble learning model to calculate the reaction depth of the gamma photons.

According to some embodiments of the present application, the ensemble learning model is calculated by using one or more of a random forest algorithm, a boosted tree algorithm, or a gradient boosted tree algorithm.

According to some embodiments of the present application, the ensemble learning model of step S503 includes a plurality of ensemble learning submodels, and is trained by using sample data. Wherein the sample data comprises feature data of the reaction depth and the reaction depth. In step S503, the reaction depth of the photon is calculated by voting or calculating an average value according to the calculation results of the plurality of ensemble learning submodels.

Taking a random forest model in a random forest algorithm as an example, because a random forest uses a plurality of decision trees to achieve a target, each decision tree gives a value. Step S503 is executed, that is, the feature data related to photons are input into the random forest model, and each branch of the random forest model will obtain an estimated value of the reaction depth, so that the final result is obtained by voting or averaging.

According to some embodiments, the ensemble learning model in step S503 may also select a plurality of different kinds of ensemble learning models, such as a lifting tree model and a gradient lifting tree model, and finally determine the final depth value of the reaction by voting or averaging.

According to some embodiments, if the feature data in step S501 is obtained by a single-ended photoelectric converter array, the ensemble learning model in step S503 is also obtained by single-ended sample feature data training.

According to some embodiments, if the feature data in step S501 is obtained by a two-terminal photoelectric converter array, the ensemble learning model in step S503 is also obtained by two-terminal sample feature data training.

The following describes how to calculate the reaction depth of photons according to the embodiment shown in fig. 5, taking the example of selecting a random forest in the ensemble learning model as an example.

The random forest is one of the integrated learning algorithms, the random forest achieves a target by utilizing a plurality of decision trees, each decision tree gives a value, and finally a final result is obtained in a voting or averaging mode. FIG. 6 illustrates a decision tree structure in which each internal node represents a test on an attribute, each branch represents a test output, and each leaf node represents a category or predictor.

The energy data in the gamma photon related feature data and the decision tree shown in fig. 7a and 7b are used to illustrate how to perform the reaction depth estimation of photons according to the embodiment shown in fig. 5.

As shown in fig. 7a and 7b, each node of the decision tree is a judgment condition. Assuming that both ends of the scintillation crystal bar/array are coupled to the photoelectric converter/array, the Energy input at both ends of a single crystal bar is CH1 Energy 4476.8 and CH2 Energy 6825.6, respectively, according to the decision tree illustrated in fig. 7a, when the first layer determines, since CH1 Energy 4476.8 is less than 5594.0, the determination condition "CH 1 Energy is less than or equal to 5594.0" that meets the node of the first layer is entered into the branch node on the left side of the second layer. Similarly, the branch trend of the second layer is judged, since the CH2 Energy is 6825.6 and does not meet the judgment condition of the node, namely that the CH2 Energy is not more than 6525.2, the branch node enters the right side of the next level, and finally the judgment route shown by the black curve in FIG. 7b is obtained, and the value of the reaction depth is 5.105 mm.

The embodiment shown in fig. 7a and 7b is energy data in the gamma photon-related feature data, and similarly, the reaction depth may also be estimated by simultaneously using feature data such as sampling points and pulse time, and the final reaction depth may be calculated by voting or averaging according to the reaction depth obtained from each feature data. The more feature data that is used, the more accurate the depth of reaction that is obtained.

In the embodiment shown in fig. 7a and 7b, the judgment conditions of each node of the decision tree and the values of the reaction depths at the leaf nodes are trained by sample data, features are randomly selected by randomly inputting data, and input samples of each tree are not necessarily all samples, so that the overfitting condition is not easy to occur. The tree is composed of different trees and can be used for processing nonlinear data and fitting a nonlinear model, so that the obtained judgment condition and the leaf node value are closer to the true value.

According to the example embodiment, the characteristic data related to photons is utilized, the position of the photon deposition in the crystal is estimated through the integrated learning model, DOI information can be obtained only by using single-ended data, the change of a back-end circuit to the existing system structure caused by a double-ended read-out detector is avoided, and the problem that photon detection is influenced by a front-end photoelectric conversion device, a circuit and the like in a double-ended read-out mode is effectively solved. In addition, on the system level, the cost is saved, extra load caused by increasing the number of channels can not be caused, and the problems of gaps and the like caused by changing the system structure are avoided.

At present, the average absolute error of the DOI estimation of the 3mm × 3mm × 20mm scintillation crystal array calculated by using the above embodiment is about 1mm, and the accuracy is higher compared with the double-end readout mode in the prior art.

The integrated learning model is simple to realize, is more convenient to realize by using the FPGA, and is easier to popularize and use in a real system.

FIG. 8 shows a flowchart of a method of training an ensemble learning model, according to an example embodiment of the present application. A method of training an ensemble learning model according to an exemplary embodiment of the present application will be described in detail with reference to fig. 8.

As shown in fig. 8, in step S801, an ensemble learning model category is selected.

According to some embodiments of the present application, the ensemble learning model is calculated using a random forest algorithm, a boosted tree algorithm, or a gradient boosted tree algorithm.

In step S803, sample data is acquired.

FIG. 9 illustrates a method of obtaining sample data. As shown in fig. 9, a Source is placed on the side of a crystal with a height of 20mm, SiPM arrays are coupled to the upper and lower ends of the crystal, gamma photons are input from a specific position, such as 30cm, on the side of the crystal, whether the incident gamma photons are collimated is monitored by the collimating crystal on the other side, then the incident position is sequentially moved, and then pulse signals with two ends at a known DOI position, i.e. a specific depth condition, can be obtained, and then characteristic data of the pulse signals, such as energy signals, etc., are obtained. It should be noted by those skilled in the art that the above-mentioned positions and distances are only used as examples and not as limitations, and those skilled in the art can select suitable positions and distances for data acquisition as required.

In step S805, an ensemble learning model is trained.

According to an exemplary embodiment of the present application, the ensemble learning model is trained using the two-end feature data and the corresponding DOI position data obtained in step S803.

According to some embodiments, if an ensemble learning model at one end is needed, when collecting the feature data, only the pulse signal feature data at one end needs to be collected, and the ensemble learning model is trained by using the pulse signal feature data at one end and the corresponding DOI position data.

FIG. 10 shows a block diagram of an apparatus for calculating a reaction depth according to an example embodiment of the present application. As shown in fig. 10, an apparatus for calculating a reaction depth includes a data acquisition unit 1001 and a reaction depth calculation unit 1003. The data acquisition unit 1001 is configured to acquire feature data related to photons from the photosensor/photosensor array, and the reaction depth calculation unit 1003 is configured to input the feature data into a preset ensemble learning model to calculate a reaction depth of the photons.

FIG. 11 is a block diagram of another apparatus for calculating a reaction depth according to an embodiment of the present application. The apparatus for calculating a reaction depth shown in fig. 11 is merely an example, and should not bring any limitation to the functions and the scope of use of the embodiments of the present application.

As shown in fig. 11, the apparatus for calculating a reaction depth is embodied in the form of a general-purpose computing device. The components of the apparatus for calculating the reaction depth may include, but are not limited to: at least one processor 210, at least one memory 220, a bus 230 connecting different system components (including the memory 220 and the processor 210), a display unit 240, and the like. Wherein the memory 220 stores program code that can be executed by the processor 210 to cause the processor 210 to perform the methods described herein according to various exemplary embodiments of the present application. For example, the processor 210 may perform a method as shown in fig. 5.

Memory 220 may include readable media in the form of volatile memory units, such as random access memory unit (RAM)2201 and/or cache memory unit 2202, and may further include read only memory unit (ROM) 2203.

The memory 220 may also include a program/utility 2204 having a set (at least one) of program modules 2205, such program modules 2205 including, but not limited to: an operating system, one or more application programs, other program modules, and program data, each of which, or some combination thereof, may comprise an implementation of a network environment.

Bus 230 may be one or more of several types of bus structures, including a memory unit bus or memory unit controller, a peripheral bus, an accelerated graphics port, a processing unit, or a local bus using any of a variety of bus architectures.

The means for calculating the depth of reaction may also be in communication with one or more external devices 300 (e.g., keyboard, pointing device, bluetooth device, etc.), one or more devices that enable a user to interact with the means for calculating the depth of reaction, and/or any device (e.g., router, modem, etc.) that enables the means for calculating the depth of reaction of photons to communicate with one or more other computing devices. Such communication may occur via an input/output (I/O) interface 250. Moreover, the device for calculating the depth of reaction may also communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public network, such as the Internet) via the network adapter 260. The network adapter 260 may communicate with other modules of the device that calculate the depth of reaction via the bus 230. It should be understood that although not shown in the figures, other hardware and/or software modules may be used in conjunction with the means for calculating the depth of reaction, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.

Through the above description of the embodiments, those skilled in the art will readily understand that the exemplary embodiments described herein may be implemented by software, or by software in combination with necessary hardware. The technical solution according to the embodiments of the present application may be embodied in the form of a software product, which may be stored in a computer-readable storage medium (which may be a CD-ROM, a usb disk, a removable hard disk, etc.) or on a network, and includes several computer program instructions to make a computing device (which may be a personal computer, a server, or a network device, etc.) execute the above-mentioned method according to the embodiments of the present application.

The software product may employ any combination of one or more readable media. The readable medium may be a readable signal medium or a readable storage medium. A readable storage medium may be, for example, but not limited to, an electronic, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination of the foregoing. More specific examples (a non-exhaustive list) of the readable storage medium include: an electrical connection having one or more wires, a portable disk, a hard disk, a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the foregoing.

A computer readable storage medium may include a propagated data signal with readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take many forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A readable storage medium may also be any readable medium that is not a readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device. Program code embodied on a readable storage medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.

Program code for carrying out operations of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, C + + or the like and conventional procedural programming languages, such as the C language or similar programming languages. The program code may execute entirely on the user's computing device, partly on the user's device, as a stand-alone software package, partly on the user's computing device and partly on a remote computing device, or entirely on the remote computing device or server. In the case of a remote computing device, the remote computing device may be connected to the user computing device through any kind of network, including a Local Area Network (LAN) or a Wide Area Network (WAN), or may be connected to an external computing device (e.g., through the internet using an internet service provider).

The computer-readable medium carries one or more program instructions that, when executed by a device, cause the computer-readable medium to perform the functions described above.

Those skilled in the art will appreciate that the modules described above may be distributed in the apparatus according to the description of the embodiments, or may be modified accordingly in one or more apparatuses unique from the embodiments. A plurality of modules in the above embodiments may be combined into one module, or one module may be further split into a plurality of sub-modules.

According to some example embodiments of the present application, the photon deposition position in the crystal is estimated by the integrated learning model using the photon-related feature data, so that the capability of acquiring the DOI only using single-ended data is realized without changing the structure of the existing detector, the change of the double-ended readout detector and the circuit to the structure of the existing system is avoided, and the influence on photon detection is avoided, and the accuracy is higher. And because the integrated learning model is simple to realize, the FPGA is more convenient to use to realize, and the integrated learning model is easier to popularize and use.

Although the present application provides method steps as described in the above embodiments or flowcharts, additional or fewer steps may be included in the method, based on conventional or non-inventive efforts. In the case of steps where no necessary causal relationship exists logically, the order of execution of the steps is not limited to that provided by the embodiments of the present application.

The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments.

The foregoing detailed description of the embodiments of the present application has been presented to illustrate the principles and implementations of the present application, and the description of the embodiments is only intended to facilitate the understanding of the methods and their core concepts of the present application. Meanwhile, a person skilled in the art should, according to the idea of the present application, change or modify the embodiments and applications of the present application based on the scope of the present application. In view of the above, the description should not be taken as limiting the application.

Claims

1. A method of calculating a depth of reaction, the method comprising:

acquiring characteristic data related to photons from a photosensor/photosensor array;

and inputting the characteristic data into a preset integrated learning model to calculate the reaction depth of the photons.

2. The method of claim 1, further comprising: and training the ensemble learning model by using sample data.

3. The method of claim 2, wherein the sample data includes feature data acquired from one or both ends of the photosensor/photosensor array.

4. The method of claim 2, wherein the sample data comprises feature data and reaction depth for the reaction depth.

5. The method of claim 1, further comprising: combining a plurality of ensemble learning submodels to form the ensemble learning model.

6. The method of claim 5, wherein the inputting the feature data into a preset ensemble learning model to calculate the reaction depth of the photon comprises:

in each integrated learning submodel, calculating by using the characteristic data respectively to generate the reaction depth of photons respectively;

and calculating the reaction depth of the photons by voting or calculating an average value according to the calculation results of the integrated learning submodels.

7. The method of claim 1, wherein the ensemble learning model is computed using one or more of a random forest algorithm, a boosted tree algorithm, or a gradient boosted tree algorithm.

8. The method of claim 1, wherein the characteristic data comprises photon corresponding energy information, pulse sampling point information, pulse rising edge time, pulse energy, pulse decay time, and/or light distribution on a photosensor/photosensor array.

9. The method of claim 1, wherein said obtaining photon-related feature data from a photosensor/photosensor array comprises: the characteristic data is acquired from the photosensor/photosensor array coupled to one or both ends of the scintillation crystal/scintillation crystal array.

10. The method of claim 9, wherein the photosensor employs a PMT, SiPM; when the two ends of the scintillation crystal/scintillation crystal array are coupled with the photoelectric sensor/photoelectric sensor array, the two ends are simultaneously provided with PMT or SiPM, or one end is provided with PMT and the other end is provided with SiPM.

11. The method of claim 1, wherein the photons comprise high-energy photons of X-rays, gamma-rays, alpha-rays, beta-rays, neutron-rays.

12. An apparatus for calculating a reaction depth, the apparatus comprising:

the data acquisition unit is used for acquiring characteristic data related to photons from the photoelectric sensor/photoelectric sensor array;

and the reaction depth calculating unit is used for inputting the characteristic data into a preset integrated learning model so as to calculate the reaction depth of the photons.

13. An apparatus for calculating a depth of reaction, comprising:

one or more processors;

a storage device for storing a computer program;

the computer program, when executed by the one or more processors, causes the one or more processors to implement the method of any one of claims 1-11.

14. A computer-readable storage medium having stored thereon program instructions that, when executed, implement the method of any of claims 1-11.