Disease analysis assistance program, disease analysis assistance method, and disease analysis assistance system

Abstract

To provide a disease analysis assistance program.SOLUTION: Subject brain data containing position information regarding a position in a brain three-dimensional model and a diagnosis value corresponding to each position information is acquired on the basis of a brain measuring result of a subject. The diagnosis value is calculated by comparing a measurement value with a predetermined reference value. On the basis of at least one diagnosis value inside an extraction region identified for each brain table position information corresponding to a brain table of the subject, an analysis value corresponding to each brain table information is calculated. Thus, brain table data representing a condition in the brain is generated on the basis of the analysis value corresponding to each brain table position information. The extraction region is defined so as to extend within a limited range in an inside direction toward the inside of the brain with at least the brain table position as a base point. Further, the analysis value is calculated so that the contribution of a minimum value of the diagnosis value inside the extraction region can be greater than that of the diagnosis value larger than an average value among the diagnosis values inside the extraction region.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a disease analysis support program, a disease analysis support method, and a disease analysis support system. [Background technology]

[0002] Patent Document 1 discloses a technique for superimposing and displaying the analysis results obtained by normalizing to a standard shape on an image of the shape of a subject's organ that has not been normalized.

[0003] The technology described in Patent Document 1 comprises a Z-value calculation unit that calculates Z-values by comparing standard brain data of healthy individuals with standard brain data of the subject; a normalization / de-normalization processing unit that normalizes the subject's individual brain data to the standard brain and de-normalizes the analysis results of the standard brain using the parameters used when normalizing to the individual brain; a brain surface data generation unit that generates 3D brain surface data having the values of the subject's individual brain shape data on the brain surface based on the subject's individual brain shape data and the de-normalized Z-value data; and a 2D image generation unit that generates a 2D image that two-dimensionally displays the Z-value distribution on the brain surface based on the 3D brain surface data, and displays the 2D image. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] Japanese Patent Application Laid-Open No. 2010-172511 Summary of the Invention [Problem to be solved by the invention]

[0005] In diagnosing some diseases, such as Alzheimer's disease, it may be effective to evaluate the specific accumulation of drugs in gray matter. However, due to the nature of drugs used for such diseases, drug accumulation may be observed nonspecifically in white matter regions located more medially than gray matter regions in the brain. Therefore, even if brain surface data used for analysis support is generated from subject brain data, if a portion of the brain surface data reflects a stronger influence of white matter than gray matter, it may be difficult to understand the state of the gray matter near the brain surface from the brain surface data, which may indicate the presence or absence of a disease and its progression. [Means for solving the problem]

[0006] According to one aspect of the present invention, a disease analysis support program is provided. The analysis support program causes at least one computer to execute the following steps: In the data acquisition step, subject brain data generated based on measurement results of the subject's brain is acquired. The measurement results include measurement values corresponding to brain positions. The subject brain data includes position information regarding positions in a three-dimensional model of the brain and diagnostic values corresponding to each of the position information. The diagnostic values are parameters that can be calculated by comparing the measurement values with predetermined reference values. In the generation step, an analytic value corresponding to each piece of brain surface position information is calculated based on at least one diagnostic value within an extraction region identified for each piece of brain surface position information representing a brain surface position corresponding to the subject's brain surface, thereby generating brain surface data that represents the state of the brain on the brain surface using the analytic value corresponding to each piece of brain surface position information. The extraction region is defined to extend within a finite range inward from the brain surface position as a base point toward the interior of the brain. In the generation step, the analytic value is calculated so that the contribution of the minimum diagnostic value within the extraction region is greater than the contribution of diagnostic values greater than the average diagnostic value within the extraction region.

[0007] This configuration provides a technology for generating brain surface data that makes it easier to understand the state of gray matter near the brain surface compared to conventional technology, thereby assisting doctors and others in diagnosing diseases. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is a configuration diagram illustrating an analysis support system 1. [Figure 2] FIG. 2 is a block diagram showing the hardware configuration of the analysis support device 3. [Figure 3] FIG. 2 is a block diagram showing the hardware configuration of a user terminal 4. [Figure 4] 1 is a diagram showing the contents of data used in information processing executed in the analysis support system 1 and an overview of changes in the data due to the information processing. [Figure 5] FIG. 10 is a conceptual diagram showing an example of the contents of subject brain data D3. [Figure 6] FIG. 10 is a diagram showing an example of an extraction region R2 in a case where the calculation conditions are set so that the minimum value within the extraction region R2 is calculated as the analytical value. [Figure 7] A case will be described where some voxels C1 in the extracted region R2 in FIG. 6 are not used in the calculation of the analytical values. [Figure 8] FIG. 10 is a diagram showing an example of an extracted region R2 when a base point O1 and a starting point O2 are different. [Figure 9] FIG. 10 is a diagram showing an example of a method for specifying an extraction region R2 so as to exclude diagnostic values P that are less than a lower limit value. [Figure 10] 10 is a diagram showing an example of an extraction region R2 when an analysis value S is calculated based on a plurality of diagnostic values P, and the weights of the diagnostic values P included in the extraction region R2. FIG. [Figure 11] FIG. 10 is a diagram showing an example of an extraction region R2 in which the weight of a voxel C1 storing a diagnostic value P greater than an upper limit is set to 0. [Figure 12] FIG. 10 is a diagram showing an example of an extraction region R2 in a case where the weight is changed based on the distance from a starting point O2. [Figure 13] FIG. 2 is an activity diagram showing an example of information processing executed in the analysis support system 1. [Figure 14]10 is an example of a user interface including a brain surface image of a subject who is negative for Alzheimer's disease, when a calculation condition is set such that the maximum value of the diagnostic value P within the extracted region R2 is adopted as the analysis value S. [Figure 15] 10 is an example of a user interface including a brain surface image of a subject who is positive for Alzheimer's disease, when a calculation condition is set such that the maximum value of the diagnostic value P within the extracted region R2 is adopted as the analysis value S. [Figure 16] FIG. 10 is a diagram showing an example of a display of a brain surface image 51 of a subject who is negative for Alzheimer's disease when the minimum value mode is adopted. [Figure 17] FIG. 10 is a diagram showing an example of a display of a brain surface image 51 of a subject who is positive for Alzheimer's disease when the minimum value mode is adopted. [Figure 18] FIG. 10 is a diagram showing an example of a display of a brain surface image 51 of a subject who is negative for Alzheimer's disease when the variable mode is adopted in the average mode. [Figure 19] FIG. 10 is a diagram showing an example of a display of a brain surface image 51 of a subject who is positive for Alzheimer's disease when the variable mode is adopted in the average mode. [Figure 20] FIG. 17 is a diagram showing a display example of a brain surface image 51 of a subject when smoothing processing is performed in the case shown in FIG. 16. [Figure 21] FIG. 18 is a diagram showing a display example of a brain surface image 51 of a subject when smoothing processing is performed in the case shown in FIG. 17. [Figure 22] 10A and 10B are diagrams showing examples of candidate regions R1 and the like when an extracted vector V1 differs from a normal direction V2. DETAILED DESCRIPTION OF THE INVENTION

[0009] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The present invention will be described below with reference to the accompanying drawings. Various features shown in the following embodiments can be combined with each other.

[0010] Incidentally, the program for realizing the software appearing in this embodiment may be provided as a non-transitory computer-readable medium, or may be provided so that it can be downloaded from an external server, or may be provided so that the program is started on an external computer and its functions are realized on a client terminal (so-called cloud computing).

[0011] In this embodiment, the term "unit" may also include, for example, a combination of hardware resources implemented by a circuit in the broad sense and software information processing that can be specifically realized by these hardware resources. In addition, this embodiment handles various types of information, which may be represented by, for example, physical values of signal values representing voltages and currents, high and low signal values as a binary bit set consisting of 0 or 1, or quantum superposition (so-called quantum bits), and communication and calculations may be performed on a circuit in the broad sense.

[0012] In addition, a circuit in the broad sense is a circuit realized by at least appropriately combining a circuit, circuitry, a processor, a memory, etc. That is, it includes an application specific integrated circuit (ASIC), a programmable logic device (e.g., a simple programmable logic device (SPLD), a complex programmable logic device (CPLD), and a field programmable gate array (FPGA)), etc.

[0013] 1. Hardware Configuration This section explains the hardware configuration.

[0014] <Analysis Support System 1> FIG. 1 is a configuration diagram showing an analysis support system 1. The analysis support system 1 is configured to perform tests on subjects and present information about the test results to a diagnosing expert such as a doctor. The analysis support system 1 can support the diagnosing expert in diagnosing the subject's disease by presenting information about the test results to the diagnosing expert according to the disease of interest. In this embodiment, the analysis support system 1 is a system for supporting the analysis of diseases in which the state of the brain of subjects administered with a specific drug differs significantly between those who are positive for the disease and those who are negative for the disease. Such diseases include Alzheimer's disease in particular. Examples of specific drugs include amyloid imaging agents and tau imaging agents. Examples of amyloid imaging agents include 18 F-florbetapir, 18 F-flutemetamol, 18 F-florbetaben, 18 F-NAV4694, 11 C-PiB (Pittsburgh compound-B), etc. Examples of tau imaging agents include: 18 F-RO69558948, 18 F-PI-2620, 18 F-PM-PBB3, 18 F-GTP1, 18 F-MK-6240, 18 F-JNJ-067, 18 F-CBD-2115, 18 F-JNJ-311 and the like.

[0015] The analysis support system 1 includes a PET / CT examination device 2, an analysis support device 3, and a user terminal 4. The analysis support device 3 and the user terminal 4 are configured to be able to communicate with each other via a telecommunications line. In one embodiment, the analysis support system 1 is made up of one or more devices or components. For example, if the analysis support system 1 is made up of only the analysis support device 3, the analysis support system 1 can be the analysis support device 3. These components will be described below.

[0016] <PET / CT Examination Device 2> PET / CT Examination Device 2 performs a PET (Positron Emission Tomograpy) / CT (Computed Tomograpy) examination on a subject. Thereby, PET / CT Examination Device 2 can generate a PET image representing the distribution of a drug in the subject's brain, which is an example of the measurement result of the subject's brain, and a CT image representing the shape of the subject's brain. That is, PET / CT Examination Device 2 can generate primary brain data D1.

[0017] <Analysis Support Device 3> FIG. 2 is a block diagram showing the hardware configuration of Analysis Support Device 3. Analysis Support Device 3 includes a communication unit 31, a storage unit 32, and a processor 33, and these components are electrically connected via a communication bus 30 inside Analysis Support Device 3. Each component will be further described.

[0018] The communication unit 31 preferably uses wired communication means such as USB, IEEE1394, Thunderbolt (registered trademark), wired LAN network communication, etc., but may include wireless LAN network communication, mobile communication such as 3G / LTE / 5G, BLUETOOTH (registered trademark) communication, etc. as needed. That is, it is more preferable to implement it as a collection of these multiple communication means. That is, Analysis Support Device 3 may communicate various information from the outside via the communication unit 31 and the network.

[0019] The memory unit 32 stores various pieces of information defined above. This can be implemented, for example, as a storage device such as a solid state drive (SSD) that stores various programs and the like related to the analysis support device 3 executed by the processor 33, or as a memory such as a random access memory (RAM) that stores temporarily required information (arguments, arrays, etc.) related to the program operations. The memory unit 32 stores various programs, variables, etc. related to the analysis support device 3 executed by the processor 33.

[0020] The processor 33 processes and controls the overall operations related to the analysis support device 3. The processor 33 is, for example, a central processing unit (CPU) not shown. The processor 33 realizes various functions related to the analysis support device 3 by reading out predetermined programs stored in the storage unit 32. In other words, information processing by software stored in the storage unit 32 is specifically realized by the processor 33, which is an example of hardware, and can be executed as each functional unit included in the processor 33. These will be described in more detail in the next section. Note that the processor 33 is not limited to being single, and multiple processors 33 may be provided for each function. A combination of these may also be used.

[0021] <User terminal 4> 3 is a block diagram showing the hardware configuration of the user terminal 4. A user of the user terminal 4 is, for example, a doctor acting as a diagnostician or a technician who operates the PET / CT examination apparatus 2. The user terminal 4 includes a communication unit 41, a storage unit 42, a processor 43, a display unit 44, and an input unit 45, and these components are electrically connected via a communication bus 40 inside the user terminal 4. The description of the communication unit 41, the storage unit 42, and the processor 43 will be omitted because they are the same as the description of each unit in the analysis support device 3.

[0022] The display unit 44 may be included in the housing of the user terminal 4 or may be externally attached. The display unit 44 displays a graphical user interface (GUI) screen that can be operated by the user. This is preferably implemented by selectively using display devices such as a CRT display, a liquid crystal display, an organic EL display, or a plasma display depending on the type of user terminal 4.

[0023] The input unit 45 is configured to be able to accept input from a user. The input unit 45 may be included in the housing of the user terminal 4 or may be externally attached. For example, the input unit 45 may be implemented as a touch panel integrated with the display unit 44. The touch panel allows the user to input tapping, swiping, and the like. Of course, instead of a touch panel, a switch button, a mouse, a QWERTY keyboard, a voice recognition device, a gesture detection device, a gaze detection device, a biosignal detection device, an imaging device, and the like may be used. That is, the input unit 45 accepts an operation input made by the user. In response, the input unit 45 transfers a signal corresponding to the operation input to the processor 43 via the communication bus 40. The processor 43 may execute predetermined control or calculation as necessary.

[0024] Here, an example of the function of the processor 33 will be described. The processor 33 is configured to be able to acquire information from the user terminal 4 or other devices. The processor 33 is configured to be able to acquire various pieces of information by reading out various pieces of information stored in a storage area that is at least a part of the memory unit 32 and writing the read out information in a working area that is at least a part of the memory unit 32. The storage area is, for example, an area of the memory unit 32 that is implemented as a storage device such as an SSD. The working area is, for example, an area that is implemented as a memory such as a RAM. Note that acquisition by the processor 33 includes acquiring output results of each functional unit included in the processor 33.

[0025] The processor 33 is configured to be able to display various types of information. The information can be presented to the user via the display unit 44 of the user terminal 4 or another device. In such a case, for example, the processor 33 controls the display unit 44 of the user terminal 4 to display visual information such as a screen, an image including a still image or a video, an icon, or a message. The processor 33 may generate only rendering information for displaying the visual information on the user terminal 4. Note that the processor 33 may present the output information to the user without going through the user terminal 4 or another device user.

[0026] Furthermore, the processor 33 can perform various calculations, judgments, etc. based on the acquired information and output new information.

[0027] 2. Information Processing This section describes the information processing executed in the aforementioned analysis support system 1. This information processing is configured to generate brain surface data D4 based on subject brain data D3 generated from primary brain data D1. The subject brain data D3 represents a three-dimensional model of the subject's brain. The brain surface data D4 represents a two-dimensional surface model that represents the state inside the subject's brain on the brain surface.

[0028] 2.1. Overview of information processing First, an overview of this information processing will be explained in relation to the transition of data used in this information processing. Figure 4 is a diagram showing the contents of data used in the information processing executed in the analysis support system 1 and an overview of the transition of the data due to the information processing.

[0029] First, the processor 33 acquires PET images and CT images from the PET / CT examination device 2. The PET images obtained as a measurement result by the PET / CT examination device 2 contain information regarding the distribution of a drug administered to the subject. The CT images obtained as a measurement result by the PET / CT examination device 2 contain information regarding the shape of the subject's brain. That is, the processor 33 acquires primary brain data D1 from the PET / CT examination device 2. That is, the primary brain data D1 obtained as a measurement result by the PET / CT examination device 2 includes a measurement value I corresponding to the position of the brain. The measurement value I can be expressed, for example, as brightness information of the PET image. Furthermore, the primary brain data D1 includes position information (x, y, z) in a three-dimensional model of the brain and the measurement value I at that position.

[0030] Next, the processor 33 superimposes the acquired PET image and the CT image. This allows the drug distribution to correspond to the shape of the subject's brain. The measurement values I included in the primary brain data D1 are stored at positions that result in a shape similar to the shape of the subject's brain. Therefore, the shape of the three-dimensional brain model represented by the primary brain data D1 still has individual differences.

[0031] Next, processor 33 adjusts the shape of the primary brain data D1 so that the data of the region of interest can be compared with the data of the reference region. Any shape adjustment method may be used, for example, by anatomical standardization to convert the shape of the primary brain data D1 into the shape of the secondary brain data D2. As a result, processor 33 generates secondary brain data D2 representing a three-dimensional brain model of a standard shape that can identify the reference region. Furthermore, among the position information of the three-dimensional brain model of the standard shape, the position information corresponding to its outer edge represents a position corresponding to the subject's brain surface. Hereinafter, for convenience of explanation, a position corresponding to the brain surface will be referred to as a brain surface position, and position information corresponding to a brain surface position will be referred to as brain surface position information.

[0032] The standard brain data represents a three-dimensional brain model obtained by converting primary brain data D1 of multiple subjects diagnosed as negative for the disease for which analysis support is being provided into a predetermined shape and statistically processing the measurement values of the converted primary brain data D1 for each position. The standard brain data includes the average value of the measurement values I associated with each position and the standard deviation of the measurement values I at that position.

[0033] Next, the processor 33 generates subject brain data D3 by comparing the measurement values of the region of interest with the measurement values of the reference region among the measurement values stored at each position in the secondary brain data D2. The subject brain data D3 includes position information regarding a position in the three-dimensional model and a diagnostic value P corresponding to each position information. The diagnostic value P is a parameter that can be calculated by comparing a measurement value I included in the measurement results of the subject's brain with a predetermined reference value, and is a standardized value representing the brain state, such as SUVr (standardized uptake value ratio). The subject brain data D3 can also be said to be data in which a diagnostic value expressing a quantitative difference (e.g., ratio) is stored instead of the measurement value in the secondary brain data D2. As an example, SUVr is expressed as the ratio of the SUV (Standardized Uptake Value) in the region of interest of the subject's brain to the SUV in the reference region of the subject's brain. In other words, the processor 33 generates subject brain data D3 by comparing the measurement values stored at each position in the secondary brain data D2 with the measurement values in the reference region of the subject's brain. SUV is a value that indicates the amount of drug accumulated per unit weight of tissue, assuming that the specific gravity of the body is 1 and that the radiation concentration when all administered drugs are uniformly distributed throughout the body is 1, and is a value defined by the measured value (PET value x calibration constant (1 / CCF)) / (administered radioactivity / subject's body weight). The reference region can be selected from the cerebellar cortex, the entire cerebellum, the entire cerebellum and brainstem, and the pons.

[0034] For example, the SUVr as a diagnostic value expresses the strength of accumulation in each part of the brain as a ratio with respect to the reference region in the secondary brain data D2, and serves as a basis for the diagnostician to determine whether an individual subject is positive or negative for the disease. The measurement values stored in association with each piece of position information of the reference region are an example of the reference value described above. Note that the form of the data of the reference region storing the reference value is not limited to this and can be arbitrary.

[0035] Next, the processor 33 calculates an analytical value S based on at least one of the diagnostic values P within the extraction region identified for each piece of brain surface position information among the position information of the generated subject brain data D3, thereby generating brain surface data D4. In this embodiment, the processor 33 identifies an extraction region for each piece of brain surface position information (x', y', z') among the position information of the subject brain data D3 according to predetermined or specified generation conditions for calculating the analytical value, and calculates an analytical value based on the diagnostic value included in the extraction region. The processor 33 associates the calculated analytical value S with the brain surface position information (x', y', z') to generate brain surface data D4 in which (x', y', z', S) is stored for each brain surface position. Therefore, the brain surface data D4 is configured to represent the state of the brain on the brain surface using the analytical value S corresponding to each piece of the above-mentioned brain surface position information (x', y', z'). In this embodiment, the processor 33 uses the brain surface data D4 to display a brain surface image representing the state of the subject's brain on the display unit 44. A doctor, as an example of a user, can use the brain surface image displayed on the display unit 44 as information for determining whether the subject is positive or negative for a disease. Brain surface images include not only those used to determine whether the subject is positive or negative for a disease, but also those used as information for determining whether the subject is likely to be positive for a disease. This can assist a diagnostician in more accurately determining whether the subject is positive or negative for a target disease based on the state of the subject's brain. The generated brain surface data D4 can be associated with the generation conditions and the generation results (e.g., whether the data is useful for the diagnostician to make a diagnosis) and then stored in the storage unit 32. The stored brain surface data D4 can be used, for example, as training data for automatically specifying the generation conditions. This can further improve the efficiency of analysis support.

[0036] Thereafter, the processor 33 ends this information processing in response to a user operation.

[0037] 2.2. Overview of the process for identifying the extracted region from the subject's brain data D3 Next, an overview of the process for identifying an extraction region from the subject brain data D3 will be described. FIG. 5 is a conceptual diagram illustrating an example of the contents of the subject brain data D3. In this embodiment, the subject brain data D3 includes a voxel C1 corresponding to position information and a diagnostic value P stored in the voxel C1. The multiple voxels C1 include multiple brain surface voxels C11 and multiple internal voxels C12. The brain surface voxel C11 represents the outer edge of the brain and is a voxel corresponding to brain surface position information. The internal voxel C12 may be a voxel associated with predetermined position information in a standard shape. For ease of explanation, FIG. 5 illustrates the voxel C1, through which the boundary line B1 representing the virtual brain surface passes, as corresponding to the brain surface voxel C11. The brain surface voxel C11 tends to correspond to at least a portion of an area corresponding to gray matter.

[0038] The internal voxels C12 are voxels within the region surrounded by the brain surface voxels C11. The internal voxels C12 can also be considered as voxels of the voxels C1 other than the brain surface voxels C11. Among the internal voxels C12, internal voxels C12 that are relatively close to the brain surface voxel C11, such as adjacent to the brain surface voxel C11, tend to correspond to at least a portion of gray matter, while internal voxels C12 that are relatively far from the brain surface voxel C11 tend to correspond to white matter.

[0039] The processor 33 sets one of the brain surface voxels C11 for which the analytic value S has not been calculated as the base point O1.

[0040] The processor 33 then calculates an extracted vector V1 from the brain surface at the base point O1. The extracted vector V1 is configured to extend inward from the brain surface position that serves as the base point O1. In this embodiment, the extracted vector V1 is configured to coincide with a normal vector to the brain surface at the base point O1. Any method for calculating the normal vector may be used. For example, the processor 33 may determine the extracted vector V1 appropriately based on the theory of vector analysis, such as by minimizing the absolute value of the sum of the inner product of the extracted vector V1 and elements of a vector set that represents the relative positional relationship of the brain surface voxel C11 near the base point O1 with respect to the brain surface voxel C11 that serves as the base point O1. In other words, the extracted region R2 extends along the extracted vector V1. For example, the inward direction (extracted vector V1) extends along the normal direction to the local plane of the brain surface at the brain surface position that serves as the base point. This configuration allows for more appropriate representation of the state inside the brain. The length of the extraction region R2 in the direction of the extraction vector V1 (hereinafter referred to as the extraction region length) is arbitrary as long as it does not include areas outside the brain surface. Specifically, for example, the length corresponding to real space may be 5, 5.5, 6, 6.5, 7, 7.5, 8, 8.5, 9, 9.5, or 10 mm, or may be within a range between any two of the values exemplified here. In this embodiment, the length of the extraction region R2 in the direction of the extraction vector V1 is preferably greater than 6.75 mm. The length may also be determined by the number of voxels C1. The number of voxels C1 corresponding to the extraction region length is arbitrary, but may be, for example, 3, 4, 5, 6, 7, 8, 9, or 10, or may be within a range between any two of the values exemplified here. In particular, the extraction region length may be determined by at least four consecutive voxels C1 starting from the starting point O2. With this configuration, the influence of gray matter and white matter can be appropriately evaluated by appropriately specifying the generation conditions. In this embodiment, the length of one side of the voxel C1 corresponds to approximately 2 to 3 mm (especially approximately 2 mm). The length of the extraction region may be determined by the length of the extraction vector V1. In this case, the extraction region R2 is made up of a set of voxels C1 through which the extraction vector V1 passes.In this case, the length of the extracted region may be defined as a length L2 from the edge of the voxel C1 including the start point O2 to the edge of the voxel C1 including the end point O3. Alternatively, for example, the length of the extracted region may be defined as a length L1 from the center of the voxel C1 that is the start point O2 to the center of the voxel C1 that is the end point O3.

[0041] Next, the processor 33 sets a set of voxels C1 (here, internal voxels C12) located in the direction from the base point O1 along the extraction vector V1 as a candidate region R1. Of course, the candidate region R1 may be set to include the brain surface voxel C11. Note that the length of the candidate region R1, in other words, the number of voxels C1 included in the candidate region R1, is arbitrary.

[0042] Next, the processor 33 identifies at least a portion of the candidate region R1 as the extraction region R2 in accordance with conditions (hereinafter referred to as extraction conditions) related to the extraction mode of the extraction region, which are included in the generation conditions. The extraction conditions may include, for example, the length of the extraction region R2 in the direction along the extraction vector V1 (the number of voxels C1), the range of diagnostic values P that can be included in the extraction region R2, the minimum value of diagnostic values P to be set as the starting point O2 of the extraction region R2, and the maximum value of diagnostic values P to be excluded from the extraction region R2. For example, the processor 33 searches for the voxel C1 that will be the starting point O2 of the extraction region R2, starting from the voxel C1 closest to the base point O1 in the direction along the extraction vector V1 (including the brain surface voxel C11 corresponding to the base point O1), and determines the voxel C1 that stores a diagnostic value that satisfies the extraction conditions as the starting point O2. In the example shown in FIG. 5, the starting point O2 coincides with the base point O1. The processor 33 then determines, as the extraction region R2, a region including the voxel C1 corresponding to the determined starting point O2 and the voxel C1 that has a predetermined positional relationship with the starting point O2. This defines the extraction region R2 to extend at least within a finite range in an inward direction (direction along the extraction vector V1) toward the interior of the brain, starting from the brain surface position as the base point O1. Alternatively, for example, the processor 33 may determine, as the extraction region R2, a candidate region R1 of a predetermined length. In this case, the starting point O2 coincides with the base point O1. The processor 33 then calculates an analytical value based on the diagnostic value included in the extraction region R2, based on a condition (hereinafter, for convenience of explanation, referred to as a calculation condition) regarding the contribution of the diagnostic value P included in the generation condition. For example, the processor 33 calculates the analytical value so that the contribution of the minimum diagnostic value P within the extraction region R2 is greater than the contribution of a diagnostic value P that is greater than the average diagnostic value P within the extraction region R2.

[0043] 2.3. Relationship between generation conditions and extraction area R2, etc. Next, a specific example of the relationship between the extraction region R2 and the analytical value for the above-mentioned generation conditions will be described. Note that the generation conditions and the corresponding extraction region R2 described here are merely examples.

[0044] <When calculating one of the diagnostic values P in the extracted region R2 as the analysis value> First, a case will be described in which one of the diagnostic values P within the extraction region R2 is calculated as the analytical value. FIG. 6 is a diagram illustrating an example of the extraction region R2 when the calculation conditions are set to calculate the minimum value within the extraction region R2 as the analytical value. In the example illustrated in FIG. 7, the extraction conditions are set so that the base point O1 and the starting point O2 coincide with each other and the extraction region R2 is composed of a total of six voxels C1 contiguous to the starting point O2. The diagnostic values P stored in the voxels C1 of the extraction region R2 are 2, 2, 3, 4, 3, and 5, starting from the voxel C1 closest to the base point O1, and tend to increase with distance from the base point O1, which is the brain surface position. This suggests that the diagnostic value P in gray matter tends to be smaller than that in white matter. In this case, the processor 33 calculates the minimum value (here, 2) of the diagnostic value P within the extraction region R2 as the analytical value corresponding to the brain surface position. This configuration further reduces the influence of white matter, which tends to have a larger diagnostic value than gray matter, on the analytical value. Specifically, for example, it is possible to reflect whether or not there is drug accumulation in the gray matter to the same extent as in the white matter. This is equivalent to setting the contribution of the voxel C1 storing the minimum diagnostic value to a finite value and setting the contribution of the voxel C1 storing a diagnostic value other than the minimum value to zero. Therefore, it can be said that the processor 33 calculates the analytical value so that the contribution of the maximum value of the diagnostic value P within the extracted region R2 is approximately zero. This configuration further reduces the possibility that the white matter will affect the analytical value.

[0045] FIG. 7 illustrates a case in which some voxels C1 in the extraction region R2 in FIG. 6 are not used in calculating the analysis value. The processor 33 may exclude from the extraction region R2 any voxels C1 in the extraction region R2 that store diagnostic values outside the ranges set according to their positions from the starting point O2, and then re-specify the extraction region R2. Here, voxels C1 whose diagnostic values are not between 2 and 4 are set as an exclusion region R3, and the diagnostic values P included in the exclusion region R3 are not used in calculating the analysis value S. Therefore, as shown in FIG. 7, even if the exclusion region R3 includes the minimum diagnostic value P (1) in the initial extraction region R2, the processor 33 can calculate the minimum value (2) included in the extraction region R2 outside the exclusion region R3 as the analysis value. This reduces the possibility of using outliers with extremely small diagnostic values in the white matter region as the analysis value, thereby more accurately reflecting the state of the gray matter. Furthermore, even voxels corresponding to gray matter regions (especially the outer edge of the standard shape) in a brain model represented by standard brain data do not necessarily reflect the gray matter regions of the subject's brain due to shape adjustment. Therefore, the likelihood of outliers with extremely small diagnostic values P being adopted as analysis values can be reduced, allowing for a more accurate reflection of the state of the gray matter. While an example has been shown in which a threshold is set in the extraction conditions and the analysis value S is calculated by excluding some diagnostic values P, this is not limiting. For example, the analysis value S may be calculated by setting a threshold in the calculation conditions rather than setting a threshold in the extraction conditions.

[0046] Furthermore, when the extraction condition is satisfied by setting the voxel C1 closest to the base point O1 as the starting point O2 among the voxels C1 storing a diagnostic value equal to or greater than a predetermined value, the base point O1 and the starting point O2 may be different. FIG. 8 is a diagram showing an example of an extraction region R2 when the base point O1 and the starting point O2 are different. The diagnostic values P stored in the candidate region R1 shown in FIG. 8 are 1, 2, 3, 3, 2, 4, 5, and 4, in order from the voxel C1 closest to the base point O1. When the predetermined value (e.g., the lower limit) included in the extraction condition is set to 3, the voxel C1 that satisfies the extraction condition for the starting point O2 is the third voxel C1 from the left in FIG. 8 (internal voxel C12), and therefore the processor 33 sets this voxel C1 as the starting point O2. The processor 33 then sets a predetermined number (6) of voxels C1 from the starting point O2 as the extraction region R2 and calculates the minimum value of the diagnostic values P stored in the extraction region R2 (here, the diagnostic value 2, which is the diagnostic value stored in the third voxel C1 from the starting point O2 in the extraction region R2) as the analytical value S. With this configuration, for example, when converting the primary brain data D1 to the secondary brain data D2, there is a difference between the brain surface and the standard shape, and a region with no measured value is set near the brain surface. In this case, when a process of interpolating the diagnostic value of the region with a relatively low value, such as a minimum value, is performed, the influence of the interpolated value on the analytical value can be reduced. The extraction region R2 does not need to be continuous from the starting point O2. If a voxel C1 storing a diagnostic value P below a predetermined value is present in the initial extraction region R2 specified based on the starting point O2, the processor 33 may exclude the voxel C1 from the extraction region R2. In other words, the extraction region R2 can be specified so that diagnostic values P below a predetermined lower limit are excluded. With this configuration, for example, when a diagnostic value near the brain surface of the subject's brain data is complemented by a minimum value or the like, the influence of such complemented value on the analytical value can be reduced. Furthermore, the extracted region R2 may be configured to be continuous from the starting point O2. In this case, for example, by being continuous from the starting point O2, the contribution of the diagnostic value P near the brain surface can be more easily reflected in the analytical value S.On the other hand, the extracted region R2 may be composed of a region continuous with the starting point O2 and a region disconnected from the starting point O2. In this case, the contribution of the diagnostic value P, which is a relatively low value inside the brain and is an outlier, to the analytical value S can be further reduced.

[0047] The manner in which the diagnostic values P are deleted is not limited to this. FIG. 9 is a diagram illustrating an example of a method for specifying the extraction region R2 so as to exclude diagnostic values P below a lower limit. The array of diagnostic values P illustrated in FIG. 9 is identical to the array of diagnostic values P illustrated in FIG. 8. As illustrated in FIG. 9, the processor 33 may match the base point O1 with the start point O2 regardless of the value of the diagnostic value P corresponding to the base point O1, and then exclude, as the exclusion region R3, voxels C1 within a range of a predetermined extraction region length from the start point O2 (in other words, the base point O1) that store diagnostic values P below a lower limit. This may specify, as the extraction region R2, an area in the candidate region R1 other than the exclusion region R3. In this way, the processor 33 can specify the extraction region R2 so as to exclude diagnostic values P below a predetermined lower limit in various ways. This method of setting the exclusion region R3 is similarly applicable to various ways in which the analysis value S is calculated based on multiple diagnostic values P, which will be described later.

[0048] <When calculating analytical values based on multiple diagnostic values> Next, a description will be given of an example where the calculation conditions are configured to calculate the analytical value S based on a plurality of diagnostic values P. Fig. 10 is a diagram showing an example of an extraction region R2 when the analytical value S is calculated based on a plurality of diagnostic values P, and weights of the diagnostic values P included in the extraction region R2.

[0049] Here, the analysis value S is calculated based on the diagnostic value at a position included in the extraction region and a weight set based on the distance of the position from the brain surface position serving as the base point. Specifically, the processor 33 calculates a weighted average of the diagnostic values P stored in each voxel C1 based on the weight set for each voxel C1, and adopts this weighted average as the analysis value S. The weight is set so that it increases as the distance of the position from the brain surface position serving as the base point decreases. With this configuration, positions closer to the base point are more likely to be gray matter, so the contribution of diagnostic values P that are more likely to correspond to gray matter can be strengthened, making analysis focusing on gray matter easier. Note that weights are an example of contribution. In the example shown in FIG. 10, the diagnostic values P within the extraction region R2 are 1, 2, 3, 4, 5, and 5, starting from the voxel C1 closest to the base point O1 (starting point O2), and the weights are 1, 0.9, 0.8, 0.7, 0.6, and 0.5. The weights are relative values between voxels C1 and may be greater than or equal to 1. Here, they are normalized based on the base point O1 (start point O2). In this case, the analysis value S is the weighted average of these values. In the case of FIG. 10, the analysis value S is approximately 2.3. Meanwhile, the average value of the diagnostic values P within the extracted region R2 is approximately 3.3, and the median value is 3.5. Therefore, the analysis value S is less than the average and median values of the diagnostic values P within the extracted region R2. This indicates that the analysis value S can more effectively emphasize the state of the gray matter than when the analysis value S is calculated simply by using the average, median, maximum, or other value of the diagnostic value P.

[0050] Even when calculating the analytic value S based on multiple diagnostic values P, the processor 33 can calculate the analytic value S so that the contribution of the maximum value of the diagnostic values P within the extracted region is approximately zero, just as when calculating one of the diagnostic values P as the analytic value S. FIG. 11 is a diagram showing an example of an extracted region R2 in which the weight of a voxel C1 storing a diagnostic value P greater than an upper limit is set to 0. As shown in FIG. 11, when the upper limit of the diagnostic value P is set to 4, the processor 33 changes the weight of a voxel C1 storing a diagnostic value P greater than the upper limit (5 or greater) to a default value (here, 0), regardless of the weight originally set for the voxel C1. As a result, the weighted average value before the change is 2.6, but after the change, it becomes 1.25. Therefore, the influence of outliers of the diagnostic values P near the brain surface, such as an extremely large diagnostic value P near the brain surface, on the analytic value S can be reduced, and the analytic value used for analyzing the area near the brain surface can be made more realistic. The upper limit may be set only for voxels C1 within a specified distance from the starting point O2. This can further reduce the influence of outliers near the brain surface. The region formed by voxels C1 whose weights are set to approximately zero is also an example of the exclusion region R3. The upper and lower limit values may also be set individually for each voxel C1 in the extraction region R2, for example, based on its positional relationship with the starting point O2. This configuration allows the influence of outliers to be appropriately considered for voxels C1 that are close to the starting point O2 and therefore likely to be gray matter, and voxels C1 that are far from the starting point O2 and therefore likely to be white matter.

[0051] The processor 33 may change the weight based on the distance from the base point O1 (start point O2). FIG. 12 is a diagram showing an example of the extraction region R2 when the weight is changed based on the distance from the start point O2. In this example, among the voxels C1 included in the extraction region R2, the weight of the voxel C1 that is the farthest from the start point O2 and the weight of the voxel C1 that is the second farthest from the start point O2 are changed to zero. As a result, the weighted average value of the diagnostic value P changes from approximately 2.05 to 1.21. This suggests that the influence of the diagnostic value P of the voxel C1, which is far from the brain surface and likely corresponds to white matter, is removed from the analysis value S, resulting in an analysis value S that is easier to determine the state of gray matter. The weight may also be changed based on the position of the voxel C1 and the diagnostic value P stored at that position. For example, the processor 33 may be configured to increase the upper limit value as the voxel C1 moves away from the start point O2. This configuration enables appropriate processing depending on the situation, using voxels C1 that are likely to correspond to gray matter and voxels C1 that are likely to correspond to white matter. Furthermore, processor 33 may change the weighting based on the position of voxels C1, regardless of the value of diagnostic value P. In other words, processor 33 may calculate analytic value S so that the contribution of diagnostic value P closest to the brain surface location corresponding to extracted region R2 is greater than the contribution of diagnostic value P farthest from the brain surface location. This configuration prioritizes the contribution of diagnostic values P included in extracted region R2 that are likely to correspond to gray matter closer to the brain surface than white matter. This reduces the impact of white matter measurement results on the analytic value, making it easier for diagnosticians, such as physicians, to analyze the state of gray matter near the brain surface. These conditions for changing the weighting are an example of calculation conditions.

[0052] Furthermore, even when calculating an analytical value based on multiple diagnostic values, the base point O1 and the starting point O2 may be set to be different, just as in the case where one of the diagnostic values P in the extraction region R2 is calculated as the analytical value.

[0053] In this way, the processor 33 specifies the extraction region R2 according to the generation conditions such as the extraction conditions and calculation conditions, and calculates the analysis value S based on the diagnostic value P within the extraction region R2.

[0054] 2.3. Information processing flow FIG. 13 is an activity diagram showing an example of information processing executed in the analysis support system 1. The information processing may include any exception processing not shown. Exception processing includes interrupting the information processing or omitting each process. The selection or input performed in the information processing may be based on a user operation or may be performed automatically without relying on a user operation. This information processing generates brain surface data D4 based on the subject's brain data D3 generated from the primary brain data D1, and displays a brain surface image that allows visual understanding of the analysis values for each brain surface position based on the generated brain surface data D4.

[0055] [Activity A1] First, in activity A1, the processor 33 generates and acquires subject brain data D3 based on primary brain data D1 obtained from the PET / CT examination device 2. When a quantitative value that requires normalization based on medical data such as the subject's weight or BMI is adopted as a diagnostic value, the processor 33 may acquire the subject's medical data. The medical data is information about the subject's body, and may include any information, such as information about the body (age, height, weight, BMI, etc.) and the results of a medical interview. The medical data is obtained, for example, by a doctor's interview or physical measurements of the subject.

[0056] [Activity A2] Next, the processor 33 acquires the designation of the generation conditions. In this embodiment, the initial generation condition is automatically set by the processor 33. For example, the processor 33 acquires the designation of at least one of the multiple generation conditions. The generation condition is a condition that specifies a method for calculating an analytic value from a diagnostic value. In this embodiment, the multiple generation conditions include at least one of a condition (extraction condition) related to the extraction mode of the extraction region R2 and a condition (calculation condition) related to the contribution of the diagnostic value P. The extraction conditions may include, for example, the direction of the extraction vector V1, the setting mode of the start point O2 (e.g., the lower limit value of the diagnostic value P that can be set as the start point O2), and the length of the extraction region. The calculation conditions may include, for example, a weight for each positional information within the extraction region R2, a condition for changing the weight, and a condition related to the diagnostic value P adopted as the analytic value S. In this embodiment, the generation conditions may include a condition related to whether or not to perform pre-processing on the subject's brain data D3. Examples of pre-processing include smoothing of each voxel C1 in the subject's brain data D3, removal of outliers, and standardization of the data format. The smoothing process is a process of changing the diagnostic value P of a certain voxel C1 based on the diagnostic value P of at least one voxel C1 included within a specified range from the certain voxel C1. For example, by the smoothing process, the processor 33 changes the diagnostic value P of a certain voxel C1 to the average value of the diagnostic values P of at least one voxel C1 included within a specified range from the certain voxel C1.

[0057] Specifically, for example, the processor 33 specifies and acquires at least one of a plurality of generation conditions based on the subject brain data D3, etc. As an example, the processor 33 specifies, as the initial generation conditions, predetermined generation conditions for test-related conditions, such as the disease to be analyzed, the type of medication administered to the subject, and the test method performed on the subject. The processor 33 may also input the test-related conditions into a trained model that has been trained in advance, thereby acquiring candidate generation conditions output from the trained model as specified generation conditions. The trained model is trained, for example, by supervised learning or semi-supervised learning using past diagnostic history as training data. The diagnostic history includes, for example, the test-related conditions and generation conditions specified when a doctor makes a diagnosis using the test-related conditions, and is configured to assign a correct label to such generation conditions. The specification may be performed, for example, by a user operating the input unit 45.

[0058] [Activity A3] If the acquired generation conditions include a command to perform pre-processing (pre-processing command), the process proceeds to activity A3, and the processor 33 executes the specified pre-processing on the subject brain data D3 and updates the diagnostic value P stored in the subject brain data D3. Then, the process proceeds to activity A4. The processor 33 may generate new subject brain data D3 including the updated diagnostic value P. On the other hand, if the acquired generation conditions do not include a pre-processing command, the process of activity A3 is omitted, and the process proceeds to activity A4.

[0059] [Activity A4] Next, in activity A4, the processor 33 identifies the brain surface position of the subject's brain data D3. As for the brain surface position, the voxel C1 that defines the outer edge of the standard shape as described above is identified as the brain surface voxel C11 that corresponds to the brain surface position.

[0060] Thereafter, the processor 33 performs the following processes of activities A5 to A12 for each identified brain surface position.

[0061] [Activity A5] First, in activity A5, the processor 33 selects one of the brain surface voxels C11 to which no analytic value S is associated as a base point O1.

[0062] [Activity A6] Next, in activity A6, the processor 33 sets an extracted vector V1 corresponding to the selected base point O1 based on the specified extraction conditions.

[0063] [Activity A7] Next, in activity A7, processor 33 identifies a candidate region R1 based on the specified extraction conditions. Specifically, processor 33 identifies multiple voxels C1 located in the direction of extraction vector V1 from selected base point O1 as candidate region R1.

[0064] [Activity A8] Next, in activity A8, processor 33 determines whether or not a diagnostic value P that can satisfy the extraction conditions is present within the identified candidate region R1. Specifically, processor 33 determines whether or not a starting point O2 can be set within the identified candidate region R1. For example, if a voxel C1 that stores a diagnostic value P equal to or greater than the lower limit exists within the candidate region R1, processor 33 determines that a diagnostic value P that can satisfy the extraction conditions is present. On the other hand, if only voxels C1 that store diagnostic values P below the lower limit exist, processor 33 determines that a diagnostic value P that can satisfy the extraction conditions is not present within the identified candidate region R1.

[0065] [Activity A9] First, the flow of information processing when it is determined that there is a diagnostic value P that can satisfy the extraction conditions will be described. In this case, processing proceeds to activity A9, and processor 33 identifies an extraction region R2 from candidate region R1 based on the extraction conditions. For example, processor 33 sets the voxel C1 included in candidate region R1 that satisfies the extraction conditions and that is the closest to base point O1 as start point O2, and identifies the region including voxel C1 that corresponds to a predetermined specific region length from start point O2 as extraction region R2.

[0066] [Activity A10] Thereafter, in activity A10, the processor 33 calculates the analysis value S from the diagnosis value P in the identified extraction region R2 based on the calculation conditions.

[0067] [Activity A11] Next, in activity A11, the processor 33 stores the calculated analytical value S in the storage unit 32 in association with the base point O1 of the extraction region R2.

[0068] [Activity A12] On the other hand, in activity A12, if the processor 33 determines that there is no diagnostic value P that can satisfy the extraction conditions within the candidate region R1 identified in activity A8, the processor 33 stores a predetermined set value in association with the base point O1 as the analysis value S in activity A12. The set value can be set arbitrarily, but for example, the start point O2 included in the extraction conditions is set to a value equal to or less than the settable lower limit value, such as the same value as the lower limit value or the smallest settable analysis value S.

[0069] In this way, in activity A11 or activity A12, if there is a brain surface voxel C11 to which no analytical value S has been associated after associating the analytical value with the base point, the processor 33 returns to activity A5 and repeats the processing of activities A5 to A12 until analytical values are associated with all brain surface voxels C11. After completing this repetition, the processing proceeds to activity A21.

[0070] [Activity A21] Next, in activity A21, the processor 33 generates brain surface data D4 defined by at least brain surface position information corresponding to the base point O1 and the analysis value S corresponding to the base point O1. In particular, for supporting the analysis of Alzheimer's disease, it is preferable that the processor 33 generates the brain surface data D4 based on brain data D3 of a subject administered with an amyloid imaging agent or a tau imaging agent. From a similar perspective, the processor 33 may also generate brain surface data based on data (e.g., the subject brain data D3) on the brain of a subject administered with a radiopharmaceutical that shows nonspecific accumulation in white matter regions and specific accumulation in gray matter regions. This configuration facilitates support for the analysis of Alzheimer's disease as a disease. [Activity A22] Next, in activity A22, the processor 33 performs a smoothing process on the generated brain surface data D4.

[0071] [Activity A23] Next, in activity 23, the processor 33 displays a brain surface image on the display unit 44, which allows the distribution of the analysis value S in the three-dimensional brain model to be visually recognized, based on the generated brain surface data D4. The user can make various diagnoses regarding diseases by referring to the brain surface image displayed on the display unit 44. Furthermore, the user can appropriately set the content and display mode of the brain surface image according to the purpose by performing a generation operation of the generation conditions on the input unit 45 while viewing the brain surface image.

[0072] When an operation to change the generation conditions is performed, the process returns to activity A2, and the processor 33 acquires the changed generation conditions again and performs the above-described process again to update the brain surface image.

[0073] The above information processing is merely an example and is not limiting. For example, activity A3 may be omitted in the above information processing. Activities A5 and A6 may be omitted when identifying the candidate region R1, the extracted region R2, or calculating the analysis value. For example, the processor 33 may identify a predetermined region in the subject brain data D3 as the candidate region R1, identify the extracted region R2 from the candidate region R1 based on extraction conditions, and generate the brain surface data D4 by associating the analysis value with the brain surface voxel C11. In activity A6, the processor 33 may not individually calculate the extracted vector V1 according to the subject brain data D3, but may instead read a predetermined standard vector (e.g., a vector of a fixed length extending in the normal direction from the brain surface position) from the storage unit 32 and set the standard vector as the extracted vector V1, regardless of the subject brain data D3.

[0074] Furthermore, when displaying a brain surface image, the pre-processing in activity A3 and the smoothing process in activity A22 may be omitted. In this case, for example, before acquiring the subject brain data D3 in activity A1, the processor 33 may acquire a specification of acquisition conditions regarding the acquisition mode of the subject brain data D3, and acquire the subject brain data D3 in activity A1 on which pre-processing (smoothing process, masking process, etc.) has been performed in advance. The acquisition conditions indicate, for example, whether the subject brain data D3 is acquired in a state before or after the smoothing process. The smoothing process may be executed by the processor 33 or by an external device other than the processor 33, for example, the PET / CT examination device 2 or its control terminal, etc.

[0075] 3. Examples of brain surface images displayed In this section, an example of a brain surface image displayed on the display unit 44 based on the brain surface data D4 generated by the above-described information processing will be described. In the following description, the brain surface image is used to assist in diagnosing whether a subject is positive or negative for Alzheimer's disease. More specifically, the brain surface image is used to assist in diagnosing whether a subject is positive or negative for amyloid accumulation.

[0076] 3.1. Example of user interface configuration In this section, for reference, an example of a brain surface image when a calculation condition is set to adopt the maximum value of the diagnostic values P within the extracted region R2 as the analysis value S will be described. FIG. 14 shows an example of a user interface including a brain surface image of a subject who is negative for Alzheimer's disease when a calculation condition is set to adopt the maximum value of the diagnostic values P within the extracted region R2 as the analysis value S. FIG. 15 shows an example of a user interface including a brain surface image of a subject who is positive for Alzheimer's disease when a calculation condition is set to adopt the maximum value of the diagnostic values P within the extracted region R2 as the analysis value S.

[0077] Before describing each brain surface image in detail, an example of the configuration of a UI on which the brain surface image is displayed will be described with reference to FIG. 14. The example of the UI configuration described here is also applicable to other examples of displaying brain surface images described below. As shown in FIG. 14, after generating brain surface data D4, the processor 33 displays an analysis support image 5 on the display unit 44. The analysis support image 5 includes a brain surface image 51, a first setting area 52, and a second setting area 53.

[0078] The brain surface image 51 is a brain surface image displayed in activity A23. It projects the brain surface data D4 onto a two-dimensional plane along a certain observation direction and visually displays the distribution of the analysis value S on the two-dimensional plane. The distribution of the analysis value S can be expressed using different colors. In this embodiment, as the analysis value S decreases, darker colors (black, purple, blue, etc.) are assigned, and as the analysis value S increases, warmer colors (yellow, orange, red, etc.) are assigned. In this embodiment, the brain surface image 51 includes two-dimensional images of the left medial surface (L-medial), right medial surface (R-medial), left lateral surface (L-lateral), right lateral surface (R-lateral), anterior surface (anterior), posterior surface (posterior), superior surface (superior), and inferior surface (inferior) of the brain. Each two-dimensional image can be displayed so that the brain surface image representing the distribution of the analysis value S is superimposed on a pre-standardized appearance image representing the brain's appearance. Furthermore, for example, each two-dimensional image may be displayed so that a brain surface image showing the distribution of the analytic value S is superimposed on a medical image such as a PET image, a CT image, or an MRI image of the subject.

[0079] The first setting area 52 is a UI for changing the generation conditions of SUVr, which is an example of the analytical value S. The first setting area 52 may include an extraction mode setting area 521, a smoothing instruction area 522, and a value range setting area 523.

[0080] The extraction mode setting area 521 is configured to allow specification of a mode for calculating the analysis value S based on the diagnostic values P included in the extraction area R2, including an average mode (tab buttons "Mean (>Low)" and "Mean (>0)") in which an average value is calculated as the analysis value based on multiple diagnostic values, a minimum mode (tab buttons "Min (>Low)" and "Min (>0)") in which the minimum value of the diagnostic values P within the extraction area R2 is calculated as the analysis value, and a maximum mode (tab button "Max") in which the maximum value of the diagnostic values P within the extraction area R2 is calculated as the analysis value.

[0081] In the averaging mode, the processor 33 calculates the average value (including a weighted average value) of the multiple diagnostic values P included in the extraction region R2 as the analysis value S. Here, the processor 33 calculates the analysis value S based on a predetermined weight. The specific aspects of the extraction region R2 based on the averaging mode are as shown in, for example, FIGS. 10 to 12.

[0082] In the minimum value mode, the processor 33 calculates the minimum value of the multiple diagnostic values P stored in the storage unit 42 as the analysis value S. The specific manner of the extraction region R2 based on the minimum value mode is as shown in, for example, FIGS. 6 to 8.

[0083] The average mode and minimum mode may each include a fixed mode in which the base point O1 is set as the starting point O2, and a variable mode ("Mean (>Low)" and "Min (>Low)") in which the lower limit value of the diagnostic value P that can be set as the starting point O2 can be set. The manner in which the extraction region R2 is specified in the fixed mode is as shown in, for example, FIGS. 6 and 10. The manner in which the extraction region R2 is specified in the variable mode is as shown in, for example, FIGS. 7 and 8.

[0084] The smoothing instruction area 522 is a UI for generating a pre-processing command that specifies whether or not to execute pre-processing performed in activity A3. If "No smooth" is specified, the processor 33 omits the processing of activity A3. If "Smoothed" is specified, the processor 33 executes the processing of activity A3 and performs smoothing processing as pre-processing. This changes each diagnostic value P stored in the subject's brain data D3, and accordingly changes the analysis value S stored in the brain surface data D4. As a result, the display mode of the brain surface image 51 changes.

[0085] The smoothing instruction area 522 may also be a UI for generating an instruction for smoothing the brain surface data D4 in activity A22. For example, if "No smooth" is specified, the processor 33 omits smoothing the brain surface data D4 in activity A22. If "Smoothed" is specified, the processor 33 executes smoothing on the brain surface data D4 in activity A22. Switching between these settings changes the display mode of the brain surface image 51. The smoothing instruction area 522 may also be configured to allow the user to specify whether to individually execute pre-processing on the subject's brain data D3 in activity A3 and smoothing on the brain surface data D4 in activity A22. The smoothing instruction area 522 may also be configured to allow the user to input specifications regarding the acquisition conditions described above. Furthermore, the generation conditions may be configured to be partially set for the brain surface data D4, etc. For example, the generation conditions may be set differently for each partial area of the brain surface.

[0086] The range setting area 523 is a UI for setting the range of the diagnostic value P (here, SUVr) used to calculate the analytical value S. For example, the range setting area 523 is configured to allow the upper limit value ("SUVr Upper") and the lower limit value ("Lower") of the diagnostic value P used to calculate the analytical value S to be individually specified.

[0087] In addition, the first setting area 52 may be configured to allow specification of the display mode of the brain surface image 51 (for example, a change in the display color, corresponding to a "Color map") other than changing the calculation method of the analysis value S, in addition to specifying the generation conditions described above.

[0088] The second setting area 53 is a UI for specifying generation conditions for the Z value, which is one of the diagnostic values P, and allows the user to specify generation conditions for the Z value, which is one of the diagnostic values P different from SUVr. The generation conditions that can be specified are the same as those for SUVr, for example. In other words, the processor 33 may acquire the specification of generation conditions for each of multiple types of diagnostic values and, based on this, generate brain surface images for each of the analysis values based on the multiple types of diagnostic values. This configuration makes it easier for the diagnostician to perform multifaceted analysis of the state of the brain.

[0089] 3.2. Example of brain surface image display when using maximum value mode Next, we will explain display examples of the brain surface image 51 according to the generation conditions set in the extraction mode setting area 521. First, we will explain a display example of the brain surface image 51 in maximum value mode. As shown in FIGS. 14 and 15 , the brain surface image 51 in maximum value mode reaches its maximum value over almost the entire area, saturating the analysis value S for both Alzheimer's disease-positive and -negative subjects. This makes it difficult to distinguish between Alzheimer's disease-positive and -negative subjects from the distribution of the analysis value S shown in the brain surface image 51. This phenomenon is not limited to when the maximum value of the diagnostic value P within the extraction area R2 is used as the analysis value S, but can also occur when using any calculation method in which the diagnostic value P in the white matter contributes significantly, or when using a simple average or median of the diagnostic value P as the analysis value S. On the other hand, the maximum value mode has the advantage of facilitating the diagnosis of diseases that are easily distinguished by the state of the white matter region. Therefore, by allowing the maximum value mode to be specified along with the minimum value mode and the average mode, we can provide an analysis support system 1 that can support the diagnosis of a wider variety of diseases.

[0090] 3.3. Example of brain surface image display when minimum value mode is used Next, a display example of a brain surface image 51 when the minimum value mode is employed will be described. FIG. 16 is a diagram showing a display example of a brain surface image 51 of a subject who is negative for Alzheimer's disease when the minimum value mode is employed. FIG. 17 is a diagram showing a display example of a brain surface image 51 of a subject who is positive for Alzheimer's disease when the minimum value mode is employed. As shown in FIG. 16, the brain surface image 51 is displayed so that no areas where the analysis value S is saturated are detected across almost the entire area, allowing for more sensitive identification of the state of the brain. On the other hand, in FIG. 17, unlike the case of FIG. 16, areas with relatively high analysis values S are clearly displayed so that they can be seen at a glance. Therefore, by using such an extraction mode, the problem of saturation of the analysis value S in the maximum value mode and the like described above can be resolved, and more sensitive information regarding the state of the brain can be presented to the diagnostician with priority. Furthermore, the presence or absence of drug accumulation in the gray matter can be more appropriately presented to the diagnostician.

[0091] 3.4. Example of brain surface image display when variable mode is used in minimum value mode Next, a display example of the brain surface image 51 when the variable mode is used in the average mode will be described. FIG. 18 is a diagram showing a display example of the brain surface image 51 of a subject who is negative for Alzheimer's disease when the variable mode is used in the average mode. FIG. 19 is a diagram showing a display example of the brain surface image 51 of a subject who is positive for Alzheimer's disease when the variable mode is used in the average mode. In this case, as shown in FIG. 18, unlike the case of FIG. 16, for the negative subject, the portion where the analysis value S is below the lower limit is not displayed as the brain surface image 51, and the external image is exposed. On the other hand, for the positive subject, the portion where the external image is exposed is larger than in FIG. 17, similar to the relationship between FIG. 16 and FIG. 18, but the area where the analysis value S is stored displayed as the brain surface image 51 is increased overall and the value is also increased in some parts, making it easier to determine the brain condition depending on whether the subject is positive or negative. Therefore, by adopting the variable mode, it is possible to suggest to the diagnostician areas that require more attention, thereby further improving the analysis support capability.

[0092] 3.5. Example of brain surface image after smoothing processing Next, a display example of the brain surface image 51 when smoothing processing is performed will be described. FIG. 20 is a diagram showing a display example of the brain surface image 51 of the subject when smoothing processing is performed in the case shown in FIG. 16. FIG. 21 is a diagram showing a display example of the brain surface image 51 of the subject when smoothing processing is performed in the case shown in FIG. 17. As shown in FIGS. 20 and 21, the brain surface image 51 has a wider distribution of the analytic value S than when smoothing processing is not performed, improving visibility and facilitating global analysis. Furthermore, by performing smoothing processing on the diagnostic value P (or the measurement value at an earlier stage), which is a stage before calculating the analytic value S, it is possible to display the brain surface image 51 with improved visibility for the diagnostician while maintaining the sensitivity of the analytic value S to information near the brain surface.

[0093] Furthermore, for example, when the user can arbitrarily set the generation conditions, the processor 33 may be configured to display on the display unit 34 the analysis support image 5 as visual information that displays the brain surface image 51 and a setting area (such as the first setting area 52) in which the display mode of the brain surface image 51, such as the generation conditions, can be set. With this configuration, the user can search for generation conditions suitable for disease analysis while changing the display mode of the brain surface image 51 without switching screens each time, compared to when the screen displaying the brain surface image 51 and the screen for changing the generation conditions are displayed on separate screens. This improves the convenience of the analysis support.

[0094] When a user such as a diagnostician sets generation conditions via a UI such as the extraction mode setting area 521 or the smoothing instruction area 522, the processor 33 may change the displayed brain surface image 51 based on the change in the generation conditions. It is preferable that the brain surface image 51 be changed in real time, for example. This can further improve the convenience of analysis support.

[0095] 4.Other The above embodiment may be modified as follows.

[0096] The extraction vector V1 does not need to coincide with the normal vector at the base point O1; it may be any direction extending from the brain surface toward the interior of the brain. FIG. 22 illustrates an example of a candidate region R1, etc., when the extraction vector V1 differs from the normal direction V2. As shown in FIG. 22, the extraction vector V1 may be inclined, for example, by an inclination angle θ relative to the normal direction V2. Note that an inclination angle θ of 0 indicates that the extraction vector V1 coincides with the normal direction V2. In this case, the candidate region R1 is identified as the region through which the extraction vector V1 passes from the base point O1, and a portion of this region is identified as the extraction region R2. Alternatively, the processor 33 may identify multiple extraction vectors V1 within a range where the inclination angle θ is equal to or less than a predetermined value, identify an extraction region R2 based on each of the multiple extraction vectors V1 for the same base point O1, calculate candidate analytic values S for each extraction region R2, and calculate the analytic value S associated with the base point O1 based on the multiple candidate analytic values S. For example, the processor 33 may calculate an average value of the multiple analytical values S as the analytical value S associated with the base point O1. With this configuration, it is possible to provide brain surface data D4 with less fluctuation in the analytical value S compared to when the analytical value S is calculated based on one extraction region R2.

[0097] The extraction region R2 may be defined to have a width in a direction different from the extraction vector V1 (for example, within a plane defined by a direction perpendicular to the extraction vector V1).

[0098] The analysis support system 1 can be applied to any disease other than Alzheimer's disease. For example, the analysis support system 1 can be applied to a drug-disease combination that tends to show specific accumulation in gray matter and non-specific accumulation in white matter.

[0099] In the above embodiment, the doctor's analysis of a disease has been described as indicating whether a specific disease is positive or negative, but this is not limited to this. For example, the analysis may include analysis for a non-deterministic diagnosis, such as the possibility of a positive or negative result for a specific disease (e.g., suspected positivity, unlikely positivity, etc.). The analysis may also include analysis of the distribution of drugs correlated with the disease, such as the degree of drug accumulation (e.g., whether drug accumulation is positive or not) of amyloid or the like. Therefore, the analysis support system 1 is not limited to support for analysis for a deterministic diagnosis, but may also include support for analysis for a non-deterministic diagnosis, such as analysis of the degree of drug accumulation.

[0100] The analysis support system 1 can also perform similar processing on parameters whose values in gray matter tend to be smaller than their values in white matter by, for example, reversing the sign of the parameter value. In other words, when the diagnostic value P tends to be smaller in gray matter than in white matter, the processor 33 may calculate the analysis value S so that the contribution of the maximum diagnostic value P within the extracted region R2 is greater than the contribution of diagnostic values P that are smaller than the average diagnostic value P within the extracted region R2.

[0101] The medical equipment included in the analysis support system 1 is not limited to the PET / CT examination device 2, but may be any device capable of constructing the subject brain data D3. The measurement results used to construct the subject brain data D3 are not limited to PET images or CT images such as the primary brain data D1.

[0102] The analysis support device 3 may be an on-premise type or a cloud type. As a cloud type analysis support device 3, the above-mentioned functions and processes may be provided in the form of, for example, SaaS (Software as a Service) or cloud computing.

[0103] In the above embodiment, the analysis support device 3 performs various storage and control operations, but multiple external devices may be used instead of the analysis support device 3. That is, various information and programs may be distributed and stored in multiple external devices using block chain technology or the like.

[0104] The above embodiment is not limited to the analysis support system 1, and may be an analysis support method or an analysis support program. The analysis support method includes each step of the analysis support system 1. The analysis support program causes at least one computer to execute each step of the analysis support system 1.

[0105] The above-described analysis support system 1 and the like may be provided in the following aspects.

[0106] (1) A disease analysis support program, which causes at least one computer to execute the following steps: a data acquisition step acquires subject brain data generated based on measurement results of the subject's brain, the measurement results including measurement values corresponding to brain positions, the subject brain data including position information regarding positions in a three-dimensional model of the brain and diagnostic values corresponding to each of the position information, the diagnostic values being parameters that can be calculated by comparing the measurement values with predetermined reference values; and a generation step generates brain surface position information representing brain surface positions corresponding to the brain surface of the subject, among the position information. and calculating an analytical value corresponding to each of the brain surface position information based on at least one of the diagnostic values inside an extraction area specified for each of the extracted regions, thereby generating brain surface data that expresses the state inside the brain on the brain surface using the analytical value corresponding to each of the brain surface position information, wherein the extraction area is defined to extend within a finite range in an inward direction toward the inside of the brain from the brain surface position as a base point, and in the generating step, the analytical value is calculated so that the contribution of the minimum value of the diagnostic value inside the extraction area is greater than the contribution of diagnostic values that are greater than the average value of the diagnostic values inside the extraction area.

[0107] With this configuration, by preferentially incorporating the contribution of relatively small diagnostic values among those included in the extracted region into the analytical value, the influence of the white matter measurement results on the analytical value can be reduced, making it easier for diagnosticians such as doctors to analyze the state of the gray matter near the brain surface.

[0108] (2) In the analysis support program described in (1) above, in the generating step, the minimum value of the diagnostic value within the extracted region is calculated as the analysis value corresponding to the brain surface position.

[0109] With this configuration, it is possible to further reduce the influence of white matter, which tends to have a greater diagnostic value than gray matter, on the analytical value.

[0110] (3) In the analysis support program according to (1) or (2) above, the extraction region is specified so as to exclude the diagnostic value that is less than a predetermined lower limit value.

[0111] With this configuration, for example, when the diagnostic value near the brain surface of the subject's brain data is complemented by a minimum value or the like, the influence of such complemented value on the analysis value can be reduced.

[0112] (4) In the analysis support program according to any one of (1) to (3) above, in the generation step, the analysis value is calculated so that the contribution of the maximum value of the diagnostic value within the extracted region is approximately zero.

[0113] This configuration can further reduce the possibility that white matter will affect the analytical values.

[0114] (5) In the analysis support program described in any one of (1) to (4) above, the analysis value is calculated based on the diagnostic value at a position included in the extraction area and a weight set based on the distance of the position from a base brain surface position, and the weight is set so that it becomes larger as the distance of the position from the base brain surface position becomes shorter.

[0115] With this configuration, the closer a location is to the base point, the more likely it is to be gray matter, thereby increasing the contribution of diagnostic values that are likely to correspond to gray matter and making analysis focusing on gray matter easier.

[0116] (6) In the analysis support program described in any one of (1) to (5) above, the subject's brain data includes a voxel corresponding to the position information and the diagnostic value stored in the voxel, the extraction region extends along a vector extending from the brain surface position serving as a base point toward the interior, and the length of the extraction region in the direction of the vector is greater than 6.75 mm.

[0117] With this configuration, the state of the gray matter near the brain surface can be more accurately incorporated into the analysis values.

[0118] (7) An analysis support program according to any one of (1) to (6) above, further comprising, in a display processing step, displaying a brain surface image that allows the analysis value for each brain surface position to be visually grasped based on the brain surface data.

[0119] According to this configuration, a diagnostician such as a doctor can visually grasp information related to the analysis values as a brain surface image, thereby providing more effective analysis support.

[0120] (8) In the analysis support program described in any one of (1) to (7) above, the designation acquisition step further acquires the designation of at least one generation condition from among a plurality of generation conditions related to the calculation of the analysis value, the plurality of generation conditions including at least one of a condition related to the extraction mode of the extraction region and a condition related to the contribution of the diagnostic value, and the generation step generates the brain surface data using the analysis value calculated based on the at least one designated generation condition.

[0121] With this configuration, appropriate brain surface data can be generated depending on the mode of analysis support.

[0122] (9) A disease analysis support program, which causes at least one computer to execute the following steps: a data acquisition step acquires subject brain data generated based on measurement results of the subject's brain, the measurement results including measurement values corresponding to brain positions, the subject brain data including position information regarding positions in a three-dimensional model of the brain and diagnostic values corresponding to each of the position information, the diagnostic values being parameters that can be calculated by comparing the measurement values with predetermined reference values; and a generation step generates brain surface position information representing brain surface positions corresponding to the brain surface of the subject, among the position information. an analysis support program that calculates an analysis value corresponding to each of the brain surface position information based on at least one of the diagnostic values inside an extraction area specified for each piece of information, thereby generating brain surface data that expresses an intracerebral state on the brain surface using the analysis value corresponding to each of the brain surface position information, wherein the extraction area is defined to extend within a finite range in an inward direction toward the inside of the brain from the brain surface position as a base point, and in the generating step, calculates the analysis value so that the contribution of the diagnostic value nearest to the brain surface position corresponding to the extraction area is greater than the contribution of the diagnostic value farthest from the brain surface position.

[0123] With this configuration, among the diagnostic values included in the extracted region, the contribution of diagnostic values that are likely to correspond to portions of gray matter closer to the brain surface than white matter is preferentially incorporated into the analytical value. This reduces the influence of the white matter measurement results on the analytical value, making it easier for diagnosticians such as doctors to analyze the state of the gray matter near the brain surface.

[0124] (10) A method for supporting analysis of a disease, comprising the steps of the analysis support program according to any one of (1) to (9) above.

[0125] (11) A disease analysis support system comprising at least one processor configured to execute a program so as to perform each step of the analysis support program described in any one of (1) to (9) above. Of course, this is not the case.

[0126] Finally, while various embodiments of the present disclosure have been described, they are presented as examples and are not intended to limit the scope of the invention. The novel embodiments may be embodied in various other forms, and various omissions, substitutions, and modifications may be made without departing from the spirit of the invention. Such embodiments and modifications are intended to be included within the scope and spirit of the invention, as well as within the scope of the inventions and their equivalents as defined in the claims. [Explanation of symbols]

[0127] 1: Analysis support system 2: PET / CT examination equipment 3:Analysis support equipment 30: Communication bus 31: Communications Department 32: Storage section 33: Processor 4: User terminal 40: Communication bus 41: Communications Department 42: Storage section 43: Processor 44:Display section 45: Input section 5: Analysis support image 51: Brain surface image 52: First setting area 521: Extraction mode setting area 522: Smoothing instruction area 523: Range setting area 53: Second setting area B1: Boundary line C1: Voxel C11: Brain surface voxel C12: interior voxel D1: Primary brain data D2: Secondary brain data D3: Subject brain data D4: Brain surface data I: Measurement value O1: base point O2: Starting point O3: End point P: diagnostic value R1: Candidate area R2:Extraction area R3: Exclusion area S: Analysis value V1: Extracted vector V2: Normal direction θ: Tilt angle

Claims

1. A disease analysis support program, causing at least one computer to perform the following steps: In the data acquisition step, subject brain data generated based on the measurement results of the subject's brain is acquired, the measurements include measurements corresponding to brain locations; the subject's brain data includes position information relating to positions in the three-dimensional model of the brain and diagnostic values corresponding to each of the position information; the diagnostic value is a parameter that can be calculated by comparing the measured value with a predetermined reference value; In the generating step, an analytical value corresponding to each piece of brain surface position information is calculated based on at least one of the diagnostic values within an extraction region identified for each piece of brain surface position information representing a brain surface position that is a position corresponding to the brain surface of the subject, among the position information, thereby generating brain surface data that represent a state within the brain on the brain surface using the analytical value corresponding to each piece of brain surface position information; the extracted region is defined so as to extend within a finite range in an inward direction toward the interior of the brain, starting from the brain surface position; an analysis support program for calculating the analytical value such that, in the generating step, a contribution of a minimum value of the diagnostic value within the extracted region is greater than a contribution of a diagnostic value greater than an average value of the diagnostic values within the extracted region.

2. 2. The analysis support program according to claim 1, In the generating step, the minimum value of the diagnostic value within the extracted region is calculated as the analytical value corresponding to the brain surface position.

3. 2. The analysis support program according to claim 1, The extraction region is specified so that the diagnostic value that is less than a predetermined lower limit value is excluded.

4. 2. The analysis support program according to claim 1, The generating step calculates the analytical value so that the contribution of the maximum value of the diagnostic value within the extracted region is approximately zero.

5. 2. The analysis support program according to claim 1, an analysis support program configured to calculate the analysis value based on the diagnostic value at a position included in the extraction region and a weight set based on the distance of the position from a base brain surface position, and the weight is set so that it becomes larger as the distance of the position from the base brain surface position becomes shorter.

6. 2. The analysis support program according to claim 1, the subject's brain data includes voxels corresponding to the position information and the diagnostic values stored in the voxels; An analysis support program, wherein the extracted region extends along a vector extending from the brain surface position serving as a base point toward the interior, and the length of the extracted region in the direction of the vector is greater than 6.75 mm.

7. 2. The analysis support program according to claim 1, Furthermore, in the display processing step, the analysis support program displays a brain surface image that allows the analysis value for each brain surface position to be visually grasped based on the brain surface data.

8. 2. The analysis support program according to claim 1, Furthermore, in the designation acquisition step, designation of at least one generation condition is acquired from among a plurality of generation conditions related to the calculation of the analytic value; the plurality of generation conditions include at least one of a condition regarding an extraction mode of the extraction region and a condition regarding a contribution of the diagnostic value; In the generating step, the brain surface data is generated using the analysis value calculated based on the at least one specified generation condition.

9. A disease analysis support program, causing at least one computer to perform the following steps: In the data acquisition step, subject brain data generated based on the measurement results of the subject's brain is acquired; the measurements include measurements corresponding to brain locations; the subject's brain data includes position information relating to positions in the three-dimensional model of the brain and diagnostic values corresponding to each of the position information; the diagnostic value is a parameter that can be calculated by comparing the measured value with a predetermined reference value; In the generating step, an analytical value corresponding to each piece of brain surface position information is calculated based on at least one of the diagnostic values within an extraction region identified for each piece of brain surface position information representing a brain surface position that is a position corresponding to the brain surface of the subject, among the position information, thereby generating brain surface data that represent a state within the brain on the brain surface using the analytical value corresponding to each piece of brain surface position information; the extracted region is defined so as to extend within a finite range in an inward direction toward the interior of the brain, starting from the brain surface position; In the generating step, the analytical value is calculated so that the contribution of the diagnostic value of the extracted region closest to the brain surface position corresponding to the extracted region is greater than the contribution of the diagnostic value of the extracted region farthest from the brain surface position.

10. A disease analysis support method, comprising: An analysis support method comprising the steps of the analysis support program according to any one of claims 1 to 9.

11. A disease analysis support system, comprising: An analysis support system comprising at least one processor configured to execute a program so as to perform each step of the analysis support program according to any one of claims 1 to 9.

Images