Information processing systems, information processing methods, and programs

Abstract

An information processing system comprising the following parts: an acquisition unit is configured to acquire projection data representing the projection image of an X-ray CT of a measurement target and an absorption model relating to the X-ray absorption pattern of the measurement target, the projection data includes information on the projection image for each direction in which the incident X-rays are incident on the measurement target; a generation unit is configured to generate corrected projection data for each candidate of a virtual incident X-ray assumed to be incident on the measurement target, the corrected projection data is projection data corrected based on the corresponding candidate of a virtual incident X-ray and the absorption model; a calculation unit is configured to calculate a consistency index for each of the generated corrected projection data, indicating the degree of consistency of the corrected projection image with respect to the direction; and an output unit is configured to output at least one corrected projection data from among the generated corrected projection data based on the consistency index. [Effect] Reduces beam hardening and artifacts.

Description

[Technical Field]

[0001] This invention relates to an information processing system, an information processing method, and a program. [Background technology]

[0002] Non-Patent Document 1 discloses a method for estimating the incident X-ray spectrum. For example, Non-Patent Document 1 discloses a method for estimating the spectrum of incident X-rays actually incident on a target by incident X-rays on a structure with a known linear absorption coefficient and measuring the intensity of the transmitted X-rays. Non-Patent Document 2 is a reference. [Prior art documents] [Non-patent literature]

[0003] [Non-Patent Document 1] Duan X, Wang, et. Al., CT scanner x-ray spectrum estimation from transmission measurements. Med Phys. 2011 Feb;38(2):993-7. [Non-Patent Document 2] T. Wurfl et. Al., 8th Conference on Industrial Computed Tomography, Wels, Austria (iCT 2018) [Overview of the project] [Problems that the invention aims to solve]

[0004] Incidentally, in X-ray CT (Computed Tomography), it is desirable to reduce adverse effects on the reconstructed image, such as beam hardening and scattering artifacts. However, even if correction is performed using the incident X-ray spectrum estimated by the method described above, it requires a structure with a known linear absorption coefficient, and there may be information (e.g., absorption edges and scattering) that cannot be taken into account by the model assumed for estimation. Therefore, there is still room for improvement in techniques for correcting projection data in X-ray CT to reduce adverse effects such as beam hardening according to the incident X-ray spectrum and the model of the object being measured. [Means for solving the problem]

[0005] According to one aspect of the present invention, an information processing system is provided, comprising the following parts: an acquisition unit configured to acquire projection data representing an X-ray CT projection image of a measurement target and an absorption model relating to the X-ray absorption pattern of the measurement target, the projection data including information on the projection image for each direction in which the incident X-rays are incident on the measurement target; a generation unit configured to generate corrected projection data for each candidate of virtual incident X-rays assumed to be incident on the measurement target, the corrected projection data being projection data corrected based on the corresponding candidate of virtual incident X-rays and the absorption model; a calculation unit configured to calculate a consistency index for each of the generated corrected projection data indicating the degree of consistency of the corrected projection image with respect to the direction; and an output unit configured to output at least one corrected projection data from among the generated corrected projection data based on the consistency index.

[0006] With this configuration, it is possible to obtain projection data that has been corrected to reduce adverse effects such as artifacts, while flexibly incorporating information that is difficult to include in existing models representing the X-ray absorption process by the object being measured, which may differ depending on the incident X-ray pattern. [Brief explanation of the drawing]

[0007] [Figure 1]It is a diagram showing an example of the system configuration and hardware configuration of the information processing system 1. [Figure 2] It is a flowchart showing an example of the flow of information processing executed in the information processing system 1. [Figure 3] It is a diagram showing an example of the spectrum of the actual incident X-ray as a comparison target. [Figure 4] It is a diagram showing an example of the spectrum of the virtual incident X-ray when the convergence condition is satisfied. [Figure 5] It is an example of a tomographic image of a sample reconstructed without correction. [Figure 6] It is an example of a tomographic image of a sample reconstructed after correction using the actual incident X-ray shown in FIG. 3. [Figure 7] It is an example of a tomographic image of a sample reconstructed after correction using the virtual incident X-ray shown in FIG. 4.

Mode for Carrying Out the Invention

[0008] Hereinafter, embodiments of the present invention will be described with reference to the drawings. Various features shown in the embodiments described below can be combined with each other.

[0009] By the way, the program for realizing the software appearing in one embodiment may be provided as a non-temporary computer-readable medium readable by a computer, may be provided so as to be downloadable from an external server, or may be provided so as to realize its function on a client terminal by starting the program on an external computer (so-called cloud computing).

[0010] Furthermore, in various information processing according to one embodiment, an input and an output corresponding to the input can be realized. Here, as long as an output is obtained as a result of the input, the form of the information referenced in such information processing (hereinafter referred to as "reference information") is not limited. The reference information may be, for example, rule-based information such as a database, a lookup table, or a predetermined function (including a decision formula such as a regression equation constructed by a statistical method), or a pre-trained model that has learned the correlation between input and output in advance, or a large-scale language model that can output a desired result by inputting a prompt.

[0011] Furthermore, in one embodiment, "part" may include, for example, hardware resources implemented by a circuit in a broad sense, and the information processing of software that can be specifically realized by these hardware resources. Also, in one embodiment, various types of information are handled, and this information can be represented, for example, by the physical values ​​of signal values ​​representing voltage and current, the high or low values ​​of signal values ​​as a set of binary bits composed of 0s or 1s, or by quantum superposition (so-called qubits), and communication and calculations can be performed on a circuit in a broad sense.

[0012] Furthermore, a circuit in a broad sense is a circuit realized by combining at least a suitable combination of circuits, circuits, processors, and memory. The processor may be a general-purpose processor or a dedicated circuit. In other words, it includes application-specific integrated circuits (ASICs), programmable logic devices (for example, simple programmable logic devices (SPLDs), complex programmable logic devices (CPLDs), and field programmable gate arrays (FPGAs)), etc.

[0013] 1. System configuration and hardware configuration of Information Processing System 1 First, the system configuration and hardware configuration of the information processing system 1 of this embodiment will be described with reference to Figure 1. Figure 1 is a diagram showing an example of the system configuration and hardware configuration of the information processing system 1.

[0014] <Information Processing System 1> The information processing system 1 shown in Figure 1 can process multiple projection images captured by the CT scanner 3. The information processing system 1 includes an information processing device 2 and a CT scanner 3. The information processing device 2 and the CT scanner 3 are configured to communicate with each other via a communication cable or network. This allows the information processing device 2 and the CT scanner 3 to transmit and receive various types of information from each other. Here, the system exemplified in the information processing system 1 consists of one or more devices or components. Therefore, even the information processing device 2 alone or the CT scanner 3 alone are included in the system exemplified in the information processing system 1. The information processing device 2 and the CT scanner 3 are operated, for example, by a user who is the measurer.

[0015] <Information Processing Device 2> The information processing device 2 is a PC (Personal Computer). The information processing device 2 may be a tablet computer, smartphone, or the like instead of a PC. The information processing device 2 can process multiple projection images captured by the CT scanner 3. Specifically, for example, the information processing device 2 is configured to perform arbitrary information processing on projection data acquired from the CT scanner 3, control radiation generated from the X-ray generator 35, acquire projection images detected by the detector 36, control the movement of the sample holder 34, control the rotation drive unit 37, and so on. The information processing device 2 only needs to be able to perform arbitrary information processing related to the CT scanner 3, and other information processing devices may be interposed between the information processing device 2 and the CT scanner 3. As shown in Figure 1, the information processing device 2 comprises a processor 21, a storage unit 22, a communication unit 23, an input unit 24, and an output unit 25, and these components are electrically connected within the information processing device 2 via a communication bus. The information processing device 2 performs processing according to the embodiment.

[0016] The processor 21 performs processing and control of the overall operation related to the information processing device 2. The processor 21 is, for example, a central processing unit (CPU). Information processing by programs stored in the memory unit 22 is concretely realized by the processor 21, which is an example of hardware, and can be executed as various functional units included in the processor 21. For example, the processing shown in Figure 2, which will be described later, is realized by each functional unit included in the processor 21. Note that the processor 21 is not limited to being a single unit, and may be implemented with multiple processors 21 for each function, or a combination thereof.

[0017] The storage unit 22 stores various types of information as defined above. This can be implemented, for example, as a storage device such as a solid-state drive (SSD) that stores various programs related to the information processing device 2 executed by the processor 21, or as memory such as random access memory (RAM) that stores temporarily necessary information (arguments, arrays, etc.) related to program calculations. The storage unit 22 stores various programs, variables, and data used by the processor 21 when it executes processing based on the programs related to the information processing device 2 executed by the processor 21. The storage unit 22 may also be an example of a storage medium.

[0018] The communication unit 23 preferably uses wired communication methods such as USB, IEEE1394, Thunderbolt®, and wired LAN network communication, but may also include wireless LAN network communication, mobile communication such as LTE / 3G / 4G / 5G, and Bluetooth® communication as needed. In other words, it is more preferable to implement it as a collection of these multiple communication methods. That is, the information processing device 2 may communicate various types of information from the outside via the communication unit 23.

[0019] The input unit 24 may be included in the housing of the information processing device 2 or it may be external. For example, the input unit 24 may be integrated with the output unit 25 and implemented as a touch panel. If it is a touch panel, the user can input tap operations, swipe operations, etc. Of course, a switch button, mouse, keyboard, etc. may be used instead of a touch panel. In other words, the input unit 24 receives input based on operations performed by the user. This input is transmitted as a command signal to the processor 21 via the communication bus, and the processor 21 can perform predetermined control or calculations as needed.

[0020] The output unit 25 can function as a display device of the information processing apparatus 2. The output unit 25 may be included in the housing of the information processing apparatus 2, or may be externally attached. The output unit 25 displays a screen of a graphical user interface (GUI) operable by a user. For example, it is preferable to selectively use display devices such as a CRT display, a liquid crystal display, an organic EL display, and a plasma display according to the type of the information processing apparatus 2.

[0021] <CT apparatus 3> The CT apparatus 3 is an apparatus capable of irradiating a sample with X-rays and acquiring a projection image of the sample from the amount of the transmitted X-rays. The CT apparatus 3 may include, but is not limited to, a sample rotation type CT apparatus that rotates the sample holding unit 34, a gantry type CT apparatus that rotates the X-ray generator 35 and the detector 36 with respect to the sample holding unit 34. The CT apparatus 3 includes a processor 31, a storage unit 32, a communication unit 33, a sample holding unit 34, an X-ray generator 35, a detector 36, and a rotation driving unit 37, and these components are electrically connected via a communication bus inside the CT apparatus 3. For the processor 31, the storage unit 32, and the communication unit 33 of the CT apparatus 3, refer to the processor 31, the storage unit 22, and the communication unit 23 of the information processing apparatus 2.

[0022] The sample holding unit 34 is configured to be able to hold a sample stage. The sample holding unit 34 may be configured to be able to move the sample stage in an arbitrary direction based on a movement instruction generated by the processor 21 or the processor 31. The sample stage is configured to be able to place a sample as a measurement object.

[0023] The X-ray generator 35 outputs incident X-rays toward a region including the sample disposed in the sample holding unit 34. The incident X-rays may be continuous X-rays or monochromatic X-rays.

[0024] The detector 36 is configured to detect X-rays that have passed through the sample placed in the sample holding section 34. For the sake of explanation, the X-rays detected by the detector 36 will be referred to as detected X-rays. The detector 36 is a two-dimensional detector, for example, using a CCD or imaging plate. The detected X-rays are transmitted as projection data to the information processing device 2, etc.

[0025] The projection data is data obtained by CT measurement using CT device 3 and represents the X-ray CT projection image of the sample. The projection data includes information about the projection image for each orientation in which the incident X-rays are incident on the sample. The orientation can be defined, for example, as the angle θ of the incident direction of the incident X-rays with respect to the reference direction of the sample. The orientation (in other words, angle θ) represents the amount of rotation of the coordinate system in each projection image. The angle θ can be defined, for example, in the range of 0 degrees or more and less than 360 degrees. The projection data can be represented, for example, as a dataset of the distribution of detected X-ray intensity I for each angle θ, where (x,y) is the position coordinate of the 2D detector. Hereinafter, the projection data as a dataset of measured orientation-specific intensity distributions I characterized by angle θ, where the distribution of detected X-ray intensity I I(x,y) at angle θ is represented as projection data I(x,y,θ). The coordinate system (x,y,θ) can be described after performing a Fan2para transformation so that it becomes a parallel beam coordinate system when the projection data is obtained using fan beam projection. In the case of a cone-beam optical system, only the detection results from the region near the central cross-section of the detector (center in the y-direction), where the incident X-ray beam can be considered a fan beam, can be used.

[0026] The rotation drive unit 37 is configured to control the incident direction of the incident X-rays to the sample. This allows the rotation drive unit 37 to change the angle θ (i.e., orientation) of the incident direction of the incident X-rays with respect to the reference direction of the sample. If the CT apparatus 3 is a sample rotation type CT apparatus, the rotation drive unit 37 is configured to rotate the sample holder 34 relative to the X-ray generator 35 and the detector 36, for example. If the CT apparatus 3 is a gantry type CT apparatus, the rotation drive unit 37 is configured to rotate the X-ray generator 35 and the detector 36 relative to the sample holder 34. The rotation drive unit 37 may include a mechanism that can adjust the magnification of the projected image during imaging (for example, a moving mechanism that can adjust the distance between the X-ray generator 35 or the detector 36 and the sample holder 34).

[0027] 2. Regarding information processing This section describes the information processing performed in the aforementioned information processing system 1. Figure 2 is a flowchart showing an example of the flow of information processing performed in information processing system 1. Note that this information processing may include arbitrary exception handling not shown. Exception handling includes interruption of the information processing or omission of each process. The selection or input performed in this information processing may be based on user operation or may be performed automatically without user operation.

[0028] [Step S1] First, in step S1, the processor 21 acquires projection data I(x,y,θ) from the CT device 3, etc. At this time, the processor 21 may also acquire incident information.

[0029] The incident information is information about the incident X-rays actually incident on the sample. The incident X-rays actually incident on the sample may be those actually detected by a detector 36, etc., from the incident X-rays output from the X-ray generator 35, or they may be spectra as standard values ​​guaranteed in advance by the manufacturer, etc., for the CT device 3. Alternatively, the actually incident X-rays may be spectra calculated inversely based on measurement results to reproduce the actually incident X-rays. For example, the incident information may include information about the energy range contained in the incident X-rays actually incident on the sample. This information about the energy range can be expressed, for example, as the lower and upper limits of the energy contained in the incident X-rays actually incident (in other words, the energy cutoff). The incident information may also include any information that characterizes the spectrum, such as the peak position (especially the maximum peak position), peak intensity, peak full width at half maximum, and weighted average value of energy of the incident X-rays actually incident. Furthermore, the incident information may be the spectrum of the incident X-rays actually incident. The incident information may also include information about the energy range of the incident X-rays. Information about the energy range of the incident X-rays means, for example, the upper or lower limit of the energy of the incident X-rays. The upper limit is preferably the same as the tube voltage, and the lower limit is preferably determined considering the lower limit of the detector's detection energy and the influence of absorption by the atmosphere.

[0030] [Step S2] Next, in step S2, the processor 21 acquires an absorption model. The absorption model is information about the X-ray absorption pattern of the sample, and can be represented, for example, as a spatial distribution of the absorption coefficient f of the sample. Steps S1 and S2 are examples of acquisition steps in this embodiment.

[0031] The absorption model can be expressed, for example, as a linear absorption coefficient model. The linear absorption coefficient model is the correspondence between incident X-rays and transmitted X-rays detected by the detector 36. The linear absorption coefficient model is a function chosen to approximate the energy dependence of the distribution of the linear absorption coefficient, and can be described, for example, as shown in equation (1) based on the Lambert-Beer law.

[0032]

number

[0033] In this model, the discretized energy is expressed using index b as E b It is expressed as follows: In equation (1), I on the left side represents the intensity of the detected X-rays detected by the detector 36. D on the right side of equation (1) b is the b-th energy E b The equation (1) represents the intensity of the X-rays, where l represents the transmission distance of the incident X-rays through the sample, and the integral with respect to l represents the line integral along the transmission path of the incident X-rays. For simplicity, the incident X-rays are assumed to be continuous X-rays and modeled as a superposition of multiple monochromatic X-rays discretized in the energy domain. In equation (1), the nonlinearity between the X-ray transmission distance and the X-ray decay is expressed by summing the decay of each monochromatic X-ray over the total energy.

[0034] As a more concrete example, the above linear absorption coefficient model can be expressed as follows, using the product of an energy-dependent scale factor s and a linear absorption coefficient f(E0) that depends only on a certain reference energy E0.

[0035]

number

[0036] Scale factor s(E b ) represents the energy dependence of the linear absorption coefficient and is a non-negative function with respect to energy E, with a value of 1 at the reference energy E0. The scale factor s can be expressed, for example, as a power function of energy as follows:

[0037]

number

[0038] The exponent α in equation (3) is an arbitrarily set parameter. Furthermore, the specific form of the scale factor s is not limited to this; any function that can approximate the energy dependence of the linear absorption coefficient, such as an exponential or logarithmic function, is acceptable.

[0039] [Step S3] Next, in step S3, the processor 21 sets the search conditions and convergence conditions. Step S3 is an example of a search condition setting step and a convergence condition setting step, and the processor 21 that performs the processing in step S3 can function as a search condition setting unit or a convergence condition setting unit.

[0040] The search conditions indicate the search range for virtual incident X-rays, which are input to the acquired correction model. Virtual incident X-rays are hypothetical incident X-rays assumed to be incident on the sample when calculating the line integral of the incident X-ray along the transmission path of the line absorption coefficient f based on the above correction model. For the sake of explanation, incident X-rays that are actually incident on the sample will be referred to as actual incident X-rays to distinguish them from the aforementioned virtual incident X-rays. For example, virtual incident X-rays have the energy E b Intensity D b It can be represented as a spectrum like a set of X-rays. Here, the virtual incident X-rays are defined as the actual incident X-rays, and the search conditions are defined, for example, based on the incident information described above. As an example, the processor 21 sets the search conditions so that the energy range included in the virtual incident X-rays is limited based on the energy range included in the actual incident X-rays (i.e., real incident X-rays). With such a configuration, the time required for correction can be shortened, the possibility of excessively adding effects that do not actually occur can be suppressed, and more appropriate correction can be performed. For example, the processor 21 sets a finite intensity D in the virtual incident X-rays set in step S4 described later. b Discrete energy E having bThe search conditions are set so that the range falls within the range defined by the lower and upper limits of the actual incident X-rays. The processor 21 may also set the search conditions so that the position of the maximum peak included in the virtual incident X-rays is limited based on the position of the maximum peak included in the actual incident X-rays (for example, so that the distance between the positions of both peaks falls within a specified range). Alternatively, the search conditions may be set independently of the actual incident X-rays.

[0041] The convergence condition is a condition for determining whether the consistency index, which will be described later, is converging or not. The convergence condition can be set in various ways, such as whether the value of the consistency index is less than or equal to a specified value indicating convergence, or whether the amount of change in the consistency index is less than or equal to a specified value.

[0042] [Step S4] Next, in step S4, the processor 21 sets the spectrum of the virtual incident X-rays. Step S4 is an example of a candidate setting step, and the processor 21 that performs the processing in step S4 can function as a candidate setting unit. The spectrum of the virtual incident X-rays set in step S4 is an example of a candidate for virtual incident X-rays that are assumed to be incident on the object to be measured. For example, the processor 21 sets the candidate for virtual incident X-rays to satisfy the search conditions set in step S3. With such a configuration, projection data that has been appropriately corrected to reduce inconsistencies with the actual measurement results can be obtained. Furthermore, some of the candidate for virtual incident X-rays can be generated based on the spectrum of incident X-rays that are actually incident on the object to be measured. With such a configuration, the correction range based on the virtual model can be appropriately narrowed down, and the time required for correction can be reduced. In this embodiment, in the first step S4, the processor 21 sets the spectrum of the virtual incident X-rays (i.e., the initial model of the virtual incident X-rays) to match the spectrum of the actual incident X-rays.

[0043] [Step S5] Next, in step S5, the processor 21 executes correction of the projection data acquired in step S1 based on the corresponding candidate of the virtual incident X-ray set in step S4 and the absorption model. As a result, the processor 21 generates corrected projection data. Hereinafter, for convenience of explanation, the projection data subjected to such correction is referred to as corrected projection data.

[0044] Here, an example of the process when the correction is beam hardening correction will be described. Beam hardening correction is a correction using a linear absorption coefficient model that represents the energy dependence of the linear absorption coefficient by a scale factor s including parameters. As an example, beam hardening correction is a correction performed based on the value of the line integral of the linear absorption coefficient at a reference energy E0. For example, the processor 21, as beam hardening correction, in the absorption model represented by Equation (2), based on the reference energy E0 set in the absorption model in step S2 and the spectrum (E b , D b ) of the virtual incident X-ray set in step S4, calculates the line integral of the linear absorption coefficient f from the detection result of the detected X-ray (that is, the projection data). Here, the processor 21 optimizes ∫dlf(E0) so as to reproduce the intensity distribution I(x, y, θ) of the detected X-ray. For example, the processor 21 calculates ∫dlf(E0) by, for example, the Newton method using the scale factor s (specifically, the power exponent α included in the scale factor) as a parameter. Note that the specific algorithm for obtaining ∫dlf(E0) is not limited to the Newton method, and may be the steepest descent method or the like. The processor 21 multiplies the line integral of the linear absorption coefficient f at the reference energy E0 by the scale factor s represented by Equation (3) to obtain the line integral of the linear absorption coefficient at the energy E b .

[0045] The processor 21 corrects the projection data obtained in step S1 based on the line integral of the line absorption coefficient f calculated in this way. This yields corrected projection data. That is, the corrected projection data is projection data that has been corrected based on the corresponding virtual incident X-ray candidate and absorption model. The corrected projection data may include the variable (x,y,θ) corresponding to the projection data before correction and the intensity distribution I(x,y,θ) corresponding to that variable.

[0046] [Step S6] Next, in step S6, the processor 21 calculates a consistency index based on the corrected projection data obtained in step S5. Step S6 is an example of a calculation step, and the processor 21 that performs the processing in step S6 can function as a calculation unit.

[0047] The consistency index is a value that indicates the degree of consistency of the corrected projection image with respect to the azimuth, and in this embodiment, it functions as an evaluation function for evaluating the spectrum of the virtual incident X-rays used for correction. The consistency index may be configured to increase when the consistency of the corrected projection image with respect to the azimuth is low and decrease when the consistency is high. The consistency index may be defined, for example, as NRMSD (Normalized Root Mean Square Deviation) as follows.

[0048]

number

[0049] M in equation (4) k It is defined as follows:

[0050]

number

[0051]

number

[0052] In equations (5) and (6), s is a parameter representing the detection position, for example, s=(x,y). k is a parameter corresponding to the direction of the incident X-ray, for example, k=θ. In equation (6), r is a coordinate system indicating the position of the sample and is a parameter representing the spatial distribution of the linear absorption coefficient f. In equation (6), m(k,s) is the line integral of the linear absorption coefficient f at reference energy E0 along the linear transmission path of a parallel beam of incident X-rays from an angle k to a detection position s. k This value represents the integral (or sum if the values ​​are obtained discretely) of m(k,s) defined by equation (6) at all detection positions s. k This value represents the total linear absorption coefficient (i.e., the absorption pattern of incident X-rays) of incident X-rays from an orientation characterized by a certain angle k, as a single sum. Note that the value in equation (6) can be obtained by using the line integral of the linear absorption coefficient f of the reference energy E0 obtained as a result of the optimization in step S5.

[0053] The processor 21 uses the correction result of the projection data according to equations (5) and (6) to calculate the M for each direction corresponding to angle k. k The processor then calculates the M for each direction calculated according to equation (4). k The consistency index (NRMSD) is calculated based on this. When the consistency for each direction is highest, the M for each direction k Since all values ​​are equal, the consistency index takes its minimum value (0 in this case). Note that the consistency index is not limited to this, M k The variance and standard deviation of the angle k may be parameters that represent the variation in the absorption pattern of incident X-rays for each angle k. In short, the consistency index can be arbitrary as long as it indicates the degree of consistency of the projected image after correction with respect to the direction.

[0054] [Step S7] Next, in step S7, the processor 21 performs a convergence condition determination step, determining whether the consistency index calculated in step S6 satisfies the convergence condition. Step S7 is an example of a convergence condition determination step, and the processor 21 executing step S7 can function as a convergence condition determination unit. The convergence condition may be the one set in step S3. For example, if it is determined that the consistency index calculated in step S6 does not satisfy the convergence condition, the processor 21 returns to step S4 and sets the virtual incident X-ray spectrum again. Then, based on the newly set virtual incident X-ray spectrum, it executes the processes in steps S5 to S7, and recursively generates corrected projection data for each virtual incident X-ray spectrum until the consistency index satisfies the convergence condition. Such processing (e.g., step S5) is an example of a generation step, and the processor 21 executing such processing can function as a generation unit. The virtual incident X-rays recursively set in step S4 in this way are also examples of candidate virtual incident X-rays, similar to the virtual incident X-rays set in the initial step S4. In this way, the processor 21 generates corrected projection data for each candidate virtual incident X-ray assumed to be incident on the sample, and calculates a consistency index for each of the generated corrected projection data. For example, if the processor 21 determines in step S7 that the consistency index does not satisfy the convergence condition, it recursively generates corrected projection data and calculates the consistency index in step S6 using a different candidate virtual incident X-ray than the one previously used. The processor 21 sets the virtual incident X-ray in the next step S4 based, for example, on a previously set candidate virtual incident X-ray (for example, a candidate virtual incident X-ray set in the previous step S4). Specifically, the processor 21 sets a virtual incident X-ray in the next step S4 that further optimizes the evaluation function such as the consistency index described above. The algorithm for searching for the next virtual incident X-ray is arbitrary, but examples include local search methods, successive improvement methods, and nearest neighbor search methods. Alternatively, such algorithms may be metaheuristic search methods such as cuckoo search, genetic algorithms, and particle swarm optimization methods.Furthermore, the processor 21 may randomly or sequentially set the spectrum of virtual incident X-rays from a pre-set list of virtual incident X-ray candidates. Such candidates may be set appropriately based on search conditions. In this case, at least one of the candidate virtual incident X-ray spectra may be set regardless of its similarity to the actual incident X-rays.

[0055] Furthermore, the processor 21 can set the virtual incident X-rays based on a consistency index, regardless of whether the virtual incident X-rays asymptotically approach the actual incident X-rays. With such a configuration, it becomes easier to incorporate the influence of factors that cause more complex artifacts occurring in the sample (e.g., the structure of the sample and higher-order scattering processes) into the virtual incident X-rays, thus potentially reducing artifacts that cannot be reduced when correcting using actual incident X-rays. The similarity between the two X-rays is determined by the variables that define the X-rays (e.g., each energy E). b Intensity D for each b This can be expressed as the distance between vectors. This distance can be defined by any index, such as the Manhattan distance, Euclidean distance, or cosine similarity.

[0056] [Step S8] On the other hand, if the processor 21 determines in step S7 that the consistency index satisfies the convergence condition, in step S8 the processor 21 accepts the specification of parameters to be used for the correction described later. These parameters are information regarding the spectrum of the virtual incident X-rays (E b ,D b ) may include. Furthermore, the parameter may include the reference energy E0 and the scale factor s(E b This may include parameters used in optimization based on the linear absorption coefficient model, such as the exponent α (or similar). This specification may be made, for example, by the user. In this case, the processor 21 may reconstruct a tomographic image of the sample based on the corrected projection data and display it on the output unit 25. With this configuration, the user can be prompted to evaluate the artifact reduction method by the correction selected based on the consistency index by comparing it with the tomographic image of the sample obtained from the correction. The parameter specification may be performed automatically by the processor 21 or the like.

[0057] [Step S9] Next, in step S9, the processor 21 performs the final correction of the projection data based on the parameters specified in step S8. This correction may be performed in the same manner as the correction of the projection data in step S5, or the line integral value of the line absorption coefficient f may be optimized again using the spectrum of the same virtual incident X-rays. In this case, the number of iterations of the optimization in step S9 may be increased compared to the optimization in step S5. As a result, the processor 21 can generate the corrected projection data. Therefore, step S9 can also function as a generation step.

[0058] [Step S10] Next, in step S10, the processor 21 outputs corrected projection data obtained by this correction as an example of corrected projection data that satisfies the convergence condition. This corrected projection data may become the final corrected projection data. Step S10 is an example of an output step in this embodiment, and the processor 21 that performs the processing in step S10 may be an example of an output unit. In other words, the processor 21 outputs at least one corrected projection data from among the corrected projection data generated in step S5, based on consistency metrics. For example, the processor 21 outputs corrected projection data that satisfies the convergence condition from among the generated corrected projection data. If there are multiple corrected projection data that satisfy the convergence condition, the processor 21 may output at least one of them.

[0059] [Step S11] Next, in step S11, the processor 21 reconstructs a tomographic image of the sample based on the corrected projection data output in step S10. The processor 21 then presents the reconstructed tomographic image of the sample to the user through the output unit 25.

[0060] Subsequently, the processor 21 terminates this information processing. With the above configuration, it is possible to obtain projection data that has been corrected to reduce adverse effects such as artifacts, while flexibly incorporating information that is difficult to take into account in existing models representing the X-ray absorption process by the object being measured, which may differ depending on the incident X-ray pattern. Furthermore, in the above embodiment, since the line integral of the linear absorption coefficient f of the sample is treated as an optimization parameter and the consistency of the absorption pattern for each direction is evaluated based on the line integral for each direction, even when information on the structure (shape) of the sample is insufficient (for example, the structure of the sample is unknown), it is possible to incorporate the influence of the structure on the distribution of the linear absorption coefficient (see Non-Patent Document 2, etc.) into the virtual incident X-ray and present a correction that has the potential to reduce artifacts.

[0061] 3. An example of the results of the above information processing Next, we will explain an example of the information processing described in the previous section. Figure 3 shows an example of the spectrum of actual incident X-rays for comparison. Figure 4 shows an example of the spectrum of virtual incident X-rays when the convergence conditions are met. In this section, the virtual incident X-rays shown in Figure 4 will simply be referred to as virtual incident X-rays.

[0062] As shown in Figure 3, the spectrum of the actual incident X-ray is, Energy E b Intensity D greater than 0 in the range of 10 keV to 80 keV b It has the following characteristics and includes a single and largest peak in the energy range of 10 keV (or less). The spectrum of the actual incident X-ray is the calibrated value (i.e., the measured value according to the standard) of the X-ray output from the X-ray generator 35.

[0063] On the other hand, as shown in Figure 4, in this embodiment, the virtual incident X-rays have an intensity D in increments of 10 keV within the range of 10 keV to 140 keV. bIt is represented as a discrete linear spectrum having a value of . As shown in Figure 4, the spectrum of the virtual incident X-ray has a finite value greater than 0 in the range of 10 keV to 80 keV, and has three peaks below 10 keV, at 50 keV, and at 80 keV. The energy range included in the spectrum of the virtual incident X-ray is within the range of the energy range included in the spectrum of the actual incident X-ray. Of the three peaks, the 50 keV peak has the greatest intensity. As shown in Figures 3 and 4, in this embodiment, the virtual incident X-ray corresponding to the consistency index that satisfies the convergence condition may have a spectrum that is completely different from that of the actual incident X-ray from various viewpoints such as peak position and peak intensity ratio.

[0064] Next, an example of the effect of the beam hardening correction described above will be explained using these incident X-rays, based on the reconstructed images. Figure 5 is an example of a tomographic image of a sample reconstructed without correction. Figure 6 is an example of a tomographic image of a sample reconstructed after correction using the actual incident X-rays shown in Figure 3. Figure 7 is an example of a tomographic image of a sample reconstructed after correction using the virtual incident X-rays shown in Figure 4. A multilayer ceramic capacitor (MLCC) was used as the sample. The exponent of the exponential function used in the correction to obtain the tomographic image shown in Figure 6 was -3.5, and the exponent of the exponential function used in the correction to obtain the tomographic image shown in Figure 7 was -4.0. For convenience of explanation, the tomographic image shown in Figure 5 will be denoted as IM1, the tomographic image shown in Figure 6 as IM2, and the tomographic image shown in Figure 7 as IM3. The white or gray rectangular shaded areas displayed in the center of these tomographic images IM1 to IM3 correspond to the tomographic cross-sections of the sample.

[0065] As shown in Figure 6, in the uncorrected tomographic image IM1, the outer edge of the sample and vertical linear regions within the sample appear blurred and white. This is thought to be due to artifacts during CT reconstruction. In contrast to the projection data that forms the basis of such a tomographic image, the tomographic image IM2, which uses actual incident X-rays, shows that the central vertical linear region is emphasized, while other regions become darker, forming an X-shaped bright region that extends diagonally from the center of the sample. Therefore, the visibility and accuracy of the tomographic image IM2 may actually be lower than that of the uncorrected tomographic image IM1. Thus, in this information processing, using actual incident X-rays may actually decrease the quality of the corrected tomographic image IM2 compared to the uncorrected tomographic image IM1.

[0066] On the other hand, as shown in Figure 7, when correction is performed using virtual incident X-rays that satisfy the convergence conditions, the outer edge of the sample in tomographic image IM3 is clearer than in the uncorrected tomographic image IM1, and there are fewer (or no) patterns that reduce accuracy, as seen in tomographic image IM2.

[0067] Thus, in the above information processing, the virtual incident X-ray is determined based on consistency indices, regardless of its similarity in shape to the actual incident X-ray (e.g., peak position, peak intensity ratio, full width at half maximum, etc.). Therefore, by incorporating various elements in the sample that require correction (e.g., morphological factors, higher-order scattering processes, etc.) into the virtual incident X-ray, it is possible to reduce artifacts more effectively compared to, for example, simply assuming a virtual incident X-ray that approximates the actual incident X-ray and performing corrections based on that assumption.

[0068] 4. Examples of absorption models and corrections Next, we will describe another example of the absorption model obtained in step S2, and another example of the correction performed in step S5 using that absorption model. In the following explanation, we will omit the explanation of the common parts of the above information processing described with reference to Figure 2.

[0069] The absorption model is Klein Nishina factor f KNIt may also be configured to include components related to other absorption or scattering processes, such as Klein Nishina factor f KN A correction model that includes components related to can be expressed as follows:

[0070]

number

[0071] In equation (7), ρ is a function representing the charge density distribution. The Klein-Nishina factor f KN (E b ) is set as appropriate. Equation (7) gives the spectrum D of the given incident X-rays. b , scale factor s, reference energy E0, and Klein Nishina factor f KN In contrast, two types of variables can be optimized towards the intensity I of the detected X-rays: the line absorption coefficient f(E0) at the reference energy E0 and the line integral of the charge density distribution along the X-ray transmission path. Therefore, the number of variables doubles compared to the case where the beam hardening correction described above is performed.

[0072] Therefore, in this embodiment, the processor 21, in step S1 of the flowchart shown in Figure 2, receives incident X-rays having two different energy spectra as projection data, and the intensity distribution of each detected X-ray I Low (x,y,θ),I High This problem can be treated as solving the following system of nonlinear equations using (x,y,θ).

[0073]

number

[0074]

number

[0075] The processor 21 can calculate solutions for two types of variables based on equations (8) and (9) above using various numerical analysis algorithms such as the multivariable Newton method. The processor 21 can then calculate the consistency index based on the line integral of the line absorption coefficient f(E0) at the reference energy E0 from among the calculated variables, and execute the process in step S7. Subsequently, if the consistency index satisfies the convergence condition, the processes from step S8 onward are performed, and the tomographic image of the sample is reconstructed. This type of correction, performed based on multiple projection data when CT imaging is performed using incident X-rays with multiple (especially two) different energy spectra, is called dual-energy correction. Therefore, the correction can be beam hardening correction or dual-energy correction. With such a configuration, projection data corrected to more appropriately reduce the effects of artifacts and the like can be obtained. Note that the specific forms of correction are not limited to these.

[0076] 5. Others The above embodiment may be modified as follows.

[0077] In the above embodiment, the virtual incident X-ray has discrete energy E b Each intensity D b Although it was described with histogram-like degrees of freedom, the method of defining virtual incident X-rays is not limited to this. For example, virtual incident X-rays may be described as a superposition of continuous peak distributions such as Gaussians. In this case, candidate spectra of virtual incident X-rays can be obtained, for example, by changing the peak position, peak intensity, and full width at half maximum of each Gaussian.

[0078] In the above embodiment, the processor 21 sequentially sets one virtual incident X-ray candidate at a time by repeatedly executing step S4 and searches for a virtual incident X-ray that satisfies the convergence condition. However, the method for setting virtual incident X-ray candidates is not limited to this. For example, the processor 21 may set a predetermined number of virtual incident X-ray candidates regularly or randomly from within a predetermined search range (for example, a range defined by the search conditions), and calculate a consistency index for each of the extracted candidates individually or in parallel.

[0079] Furthermore, the processor 21 does not need to set convergence conditions. In this case, the processor 21 corrects the projection image data based on the set virtual incident X-ray candidates and calculates a consistency index. Subsequently, the processor 21 may determine the corrected projection data to be output based on the calculated consistency index.

[0080] The processor 21 may omit the processing in steps S8 and S9 shown in Figure 2. For example, the processor 21 may output the result of the correction in step S5, which was performed using the parameters for which the consistency index was determined to satisfy the convergence condition in step S7, as the corrected projection data in step S10.

[0081] The processor 21 does not need to set the virtual incident X-ray candidates itself, and may obtain virtual incident X-ray candidates that have been set in advance by another device, and then perform projection image data correction and consistency index calculation for each candidate.

[0082] The correction model is not limited to a linear absorption coefficient model; any model representing the distribution of the sample's absorption coefficient, which describes the relationship between incident X-rays and transmitted X-rays, can be applied. For example, the correction model may be a mass absorption coefficient model that represents the distribution of mass absorption coefficients.

[0083] The information processing device 2 may be on-premise or in a cloud-based configuration. In the case of a cloud-based information processing device 2, for example, the above-mentioned functions and processing may be provided in the form of SaaS (Software as a Service) or cloud computing.

[0084] In the above embodiment, the information processing device 2 performed various storage and control functions, but instead of the information processing device 2, multiple external devices may be used. That is, various information and programs may be stored in a distributed manner across multiple external devices using blockchain technology or the like.

[0085] The above embodiment is not limited to the information processing system 1, but may also be an information processing method or a program. The information processing method includes each step of the information processing system 1. The program causes at least one computer to execute each step of the information processing system 1.

[0086] The above-mentioned information processing system 1, etc., may be provided in any of the following embodiments.

[0087] (1) An information processing system comprising the following parts: an acquisition unit configured to acquire projection data representing an X-ray CT projection image relating to a measurement target and an absorption model relating to the X-ray absorption pattern by the measurement target, wherein the projection data includes information relating to the projection image for each direction in which the incident X-rays are incident on the measurement target; a generation unit configured to generate corrected projection data for each candidate of virtual incident X-rays assumed to be incident on the measurement target, wherein the corrected projection data is the projection data that has been corrected based on the corresponding candidate of virtual incident X-rays and the absorption model; a calculation unit configured to calculate a consistency index for each of the generated corrected projection data indicating the degree of consistency of the corrected projection image with respect to the direction; and an output unit configured to output at least one of the generated corrected projection data based on the consistency index.

[0088] With this configuration, it is possible to obtain projection data that has been corrected to reduce adverse effects such as artifacts, while flexibly incorporating information that is difficult to include in existing models representing the X-ray absorption process by the object being measured, which may differ depending on the incident X-ray pattern.

[0089] (2) An information processing system as described in (1) above, wherein the correction is beam hardening correction or dual energy correction.

[0090] With this configuration, it is possible to obtain projection data that has been corrected to more effectively reduce the effects of artifacts and other factors.

[0091] (3) An information processing system as described in (1) or (2) above, wherein the candidate setting unit is further configured to set candidates for virtual incident X-rays so as to satisfy search conditions defined based on incident information relating to incident X-rays actually incident on the object to be measured.

[0092] With this configuration, it is possible to obtain projection data that has been appropriately corrected to minimize discrepancies with the actual measurement results.

[0093] (4) An information processing system as described in (3) above, wherein some of the candidates for the virtual incident X-rays are generated based on the spectrum of the incident X-rays actually incident on the object to be measured.

[0094] With this configuration, the correction range based on the virtual model can be appropriately narrowed, and the time required for correction can be reduced.

[0095] (5) An information processing system as described in (3) or (4) above, wherein the incident information includes information on the energy range of the incident X-ray, and the search condition setting unit is configured to set the search conditions such that the energy range of the virtual incident X-ray is limited based on the energy range of the incident X-ray.

[0096] This configuration allows for shorter correction times while suppressing the possibility of over-incorporating effects that are unlikely to occur in reality, thereby enabling more appropriate corrections.

[0097] (6) An information processing system according to any one of (1) to (5) above, wherein the convergence condition setting unit is configured to set convergence conditions, and the convergence condition determination unit is configured to determine whether the consistency index satisfies the convergence conditions, and if it is determined that the convergence conditions are not satisfied, the generation unit recursively generates the corrected projection data and the calculation unit calculates the consistency index using a candidate for a virtual incident X-ray different from the candidate for the virtual incident X-ray, and if it is determined that the convergence conditions are satisfied, the output unit is configured to output the corrected projection data that satisfies the convergence conditions.

[0098] (7) An information processing system according to any one of (1) to (6) above, wherein the reconstruction unit is configured to reconstruct a tomographic image of the object to be measured from the X-ray CT projection image of the object to be measured based on the output corrected projection data.

[0099] (8) An information processing method comprising the following steps: an acquisition step of acquiring projection data representing an X-ray CT projection image relating to a measurement target and an absorption model relating to the X-ray absorption pattern by the measurement target, wherein the projection data includes information relating to the projection image for each direction in which the incident X-rays are incident on the measurement target; a generation step of generating corrected projection data for each candidate of virtual incident X-rays assumed to be incident on the measurement target, wherein the corrected projection data is the projection data that has been corrected based on the corresponding candidate of virtual incident X-rays and the absorption model; a calculation step of calculating a consistency index for each of the generated corrected projection data that indicates the degree of consistency of the corrected projection image with respect to the direction; and an output step of outputting at least one of the generated corrected projection data based on the consistency index.

[0100] (9) A program that causes at least one computer to perform the following steps: an acquisition step, which acquires projection data representing an X-ray CT projection image of a measurement target and an absorption model relating to the X-ray absorption pattern of the measurement target, wherein the projection data includes information relating to the projection image for each direction in which the incident X-rays are incident on the measurement target; a generation step, which generates corrected projection data for each candidate of a virtual incident X-ray assumed to be incident on the measurement target, wherein the corrected projection data is the projection data that has been corrected based on the corresponding candidate of the virtual incident X-ray and the absorption model; a calculation step, which calculates a consistency index for each of the generated corrected projection data that indicates the degree of consistency of the corrected projection image with respect to the direction; and an output step, which outputs at least one of the generated corrected projection data based on the consistency index. Of course, this is not always the case.

[0101] Finally, while various embodiments relating to this disclosure have been described, these are presented as examples only and are not intended to limit the scope of the invention. These novel embodiments can be implemented in a variety of other forms, and various omissions, substitutions, and modifications can be made without departing from the spirit of the invention. These embodiments and their variations are included in the scope and spirit of the invention, as well as in the claims and their equivalents. [Explanation of Symbols]

[0102] 1: Information Processing System 2: Information Processing Device 21: Processor 22: Storage section 23: Communications Department 24: Input section 25: Output section 3: Equipment 31: Processor 32: Storage section 33: Communications Department 34: X-ray generator 35: Detector 36: Sample holding section 37: Rotary drive unit IM1: Tomographic image IM2: Tomographic image IM3: Tomographic image

Claims

1. An information processing system, It comprises the following parts: The acquisition unit is configured to acquire projection data representing the X-ray CT projection image of the object to be measured, and an absorption model relating to the X-ray absorption pattern of the object to be measured. The projection data includes information regarding the projection image for each direction in which the incident X-rays are incident on the object to be measured. The generation unit is configured to generate corrected projection data for each candidate of the virtual incident X-rays assumed to be incident on the object to be measured. The corrected projection data is the projection data that has been corrected based on the corresponding virtual incident X-ray candidate and the absorption model. The calculation unit is configured to calculate a consistency index for each of the generated corrected projection data, which indicates the degree of consistency of the corrected projection image with respect to the orientation. An information processing system in which the output unit is configured to output at least one of the corrected projection data from the generated corrected projection data based on the consistency index.

2. In the information processing system described in claim 1, The aforementioned correction is a beam hardening correction or a dual-energy correction, in this information processing system.

3. In the information processing system described in claim 1, Furthermore, the candidate setting unit is configured to set candidate virtual incident X-rays so as to satisfy search conditions defined based on incident information relating to incident X-rays actually incident on the measurement target, in this information processing system.

4. In the information processing system described in claim 3, An information processing system in which some of the candidate virtual incident X-rays are generated based on the spectrum of incident X-rays actually incident on the object to be measured.

5. In the information processing system described in claim 3, The incident information includes information on the energy region of the incident X-rays, Furthermore, the search condition setting unit is configured to set the search conditions such that the energy range of the virtual incident X-ray is limited based on the energy range of the incident X-ray, in this information processing system.

6. In the information processing system described in claim 1, Furthermore, the convergence condition setting unit is configured to set convergence conditions. Furthermore, the convergence condition determination unit, Determine whether the consistency index satisfies the convergence condition, If it is determined that the convergence conditions are not met, the generation of the corrected projection data by the generation unit and the calculation of the consistency index by the calculation unit are recursively performed using a candidate for a virtual incident X-ray different from the candidate for the virtual incident X-ray. An information processing system configured to output the corrected projection data satisfying the convergence conditions using the output unit when it is determined that the convergence conditions are met.

7. In the information processing system described in claim 1, Furthermore, the reconstruction unit is configured to reconstruct a tomographic image of the object to be measured from the X-ray CT projection image of the object to be measured, based on the outputted corrected projection data, in this information processing system.

8. Information processing method, The following steps are included: In the acquisition step, projection data representing the X-ray CT projection image of the object to be measured and an absorption model relating to the X-ray absorption pattern of the object to be measured are acquired. The projection data includes information regarding the projection image for each direction in which the incident X-rays are incident on the object to be measured. In the generation step, corrected projection data is generated for each candidate of the virtual incident X-rays assumed to be incident on the object to be measured. The corrected projection data is the projection data that has been corrected based on the corresponding virtual incident X-ray candidate and the absorption model. In the calculation step, for each of the generated corrected projection data, a consistency index is calculated that indicates the degree of consistency of the corrected projection image with respect to the orientation. An information processing method that, in the output step, outputs at least one of the corrected projection data from the generated corrected projection data based on the consistency index.

9. It is a program, Have at least one computer perform each of the following steps: In the acquisition step, projection data representing the X-ray CT projection image of the object to be measured and an absorption model relating to the X-ray absorption pattern of the object to be measured are acquired. The projection data includes information regarding the projection image for each direction in which the incident X-rays are incident on the object to be measured. In the generation step, corrected projection data is generated for each candidate of the virtual incident X-rays assumed to be incident on the object to be measured. The corrected projection data is the projection data that has been corrected based on the corresponding virtual incident X-ray candidate and the absorption model. In the calculation step, for each of the generated corrected projection data, a consistency index is calculated that indicates the degree of consistency of the corrected projection image with respect to the orientation. The program outputs at least one of the corrected projection data from the generated corrected projection data based on the consistency index.

Images