Computer programs and control devices

Abstract

To provide a computer program and control device capable of executing a method for providing data that quantitatively indicates the progression of fibrosis in a disease. [Solution] A computer program can cause the computer's arithmetic circuit to execute the following method. The method includes identifying a pre-identified region from a sample image and extracting a plurality of fibrotic regions. The method includes constructing a plurality of fibrotic subregions, each of which includes a divided region generated by dividing a first fibrotic region among the plurality of fibrotic regions, and a second fibrotic region other than the first fibrotic region among the plurality of fibrotic regions. The method includes setting a representative point for each of the plurality of fibrotic subregions, and grouping fibrotic subregions located within a predetermined distance based on the representative point. The method includes obtaining statistics on the grouped fibrotic subregions.

Description

Technical Field

[0001] The present disclosure relates to a computer program and a control device. In particular, the present disclosure relates to a computer program for a method of quantitatively evaluating the progression state of fibrosis of tissue related to a predetermined disease, and a control device capable of executing the method.

Background Art

[0002] Conventionally, the pathological diagnosis of a predetermined disease, such as MASH (Metabolic dysfunction-Associated Steatohepatitis), has been performed based on a local judgment by visual inspection after magnifying the lesion site under a microscope. MASH is a refractory progressive liver disease that satisfies some of the criteria for metabolic syndrome. Since the pathological diagnosis depends on the experience and subjective judgment of a pathologist, a method for quantitatively evaluating the progression state of fibrosis has been studied as a more objective evaluation method. For example, Patent Document 1 discloses a method of quantifying pathological features from a pathological tissue image and extracting a fibrotic site of tissue.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The method described in Patent Document 1 cannot evaluate the mutual relationship of a plurality of scattered fibrotic sites. In the fibrosis of MASH, fine fibrotic sites are connected to each other during progression. Therefore, it is preferable to evaluate based on objective data considering the connectivity of a plurality of scattered fibrotic sites. Therefore, a method for acquiring data for quantitatively evaluating the progression state of fibrosis of diseases such as MASH is required.

[0005] The purpose of this disclosure is to provide a computer program and control device capable of performing a method for providing data that quantitatively indicates the progression of fibrosis in a disease. [Means for solving the problem]

[0006] A computer program according to one aspect of the present disclosure causes a computer's arithmetic circuit to execute a method, the method comprising: extracting a plurality of fibrous regions from a sample image which is an image of a sample; constructing a plurality of fibrous subregions which include a divided region generated by dividing a first fibrous region which is at least a portion of the plurality of fibrous regions, and a second fibrous region which is a fibrous region other than the first fibrous region among the plurality of fibrous regions; setting a representative point for each of the plurality of fibrous subregions; grouping fibrous subregions located within a predetermined distance from the plurality of fibrous subregions based on the representative point; and obtaining statistics relating to the grouped fibrous subregions from the grouped fibrous subregions. [Effects of the Invention]

[0007] According to this disclosure, it is possible to provide a computer program and a control device that can execute a method for providing data that quantitatively indicates the progression of fibrosis in a disease. [Brief explanation of the drawing]

[0008] [Figure 1] Figure 1 shows a schematic diagram of a data acquisition system according to an embodiment of this disclosure. [Figure 2] Figure 2 is a flowchart showing an example of processing by the control device. [Figure 3] Figure 3 shows an example of a slide image captured by the imaging device. [Figure 4] Figure 4 is a magnified view showing a portion of the sample in Figure 3. [Figure 5] Figure 5 is a flowchart showing an example of a process for constructing a fibrous subregion. [Figure 6]Figure 6 shows multiple fibrotic subregions constructed in a unique space. [Figure 7] Figure 7 shows a diagram of multiple grouped fibrotic subregions. [Figure 8] Figure 8 is an enlarged view showing a portion of Figure 6. [Figure 9] Figure 9 shows a subgraph generated based on the fibrotic subregions shown in Figure 8. [Modes for carrying out the invention]

[0009] The embodiments relating to this disclosure will be described below with reference to the drawings. However, the configurations described below are merely examples of this disclosure, and this disclosure is not limited to the embodiments described below. The technology in this disclosure is not limited thereto, and various modifications, substitutions, additions, and omissions are possible in other embodiments as long as they do not depart from the technical idea relating to this disclosure, depending on the design, etc.

[0010] While this disclosure is adequately described in relation to preferred embodiments with reference to the accompanying drawings, various modifications and alterations will be obvious to those skilled in the art. Such modifications and alterations should be understood to be included within the scope of this disclosure as defined by the attached claims.

[0011] It is estimated that there are approximately 2 million patients with MASH in Japan. Traditionally, the pathological diagnosis of MASH has been made by visually inspecting the lesion under a microscope. This pathological diagnosis relies on the experience and subjective judgment of the pathologist. Therefore, even with diagnostic criteria in place, the diagnosis may differ depending on the pathologist. Consequently, there is a need for a method that can objectively evaluate the state of the disease without relying on the experience of the pathologist.

[0012] In conventional pathological diagnosis of MASH, the progression of MASH is assessed by focusing on the accumulation of collagen fibers formed in association with chronic inflammation, and there are methods to evaluate the spatial extent of fibrotic sites and their connectivity to other fibrotic sites according to predetermined diagnostic criteria. Fibrosis occurs when fibrous tissue, such as collagen fibers, proliferates. In areas where fibrosis occurs, normal tissue is replaced by fibrous components. Because the diagnostic criteria in conventional pathological diagnosis are not quantitative, the final diagnosis and the fibrotic state confirmed by actual observation may not intuitively match. Furthermore, it is known that the progression of fibrosis is uneven, and the diagnostic results may differ depending on the observed site.

[0013] The disease state is staged on a five-point scale (0, 1, 2, 3, 4) using, for example, Brandt's diagnostic criteria. The criteria for this diagnosis involve the presence of specific characteristics in various parts of the liver's histological structure. However, these criteria may yield different results depending on the selection of the observation area. Furthermore, because this diagnosis is qualitative, it lacks clarity.

[0014] Thus, since each stage in conventional pathological diagnosis is on an ordinal scale, it does not quantitatively represent the state of the disease. Therefore, it is difficult to compare diagnostic results from different samples using conventional pathological diagnosis. For this reason, there is a need to represent the state of the disease on a ratio scale, enabling comparison of the state of the disease across different samples. Furthermore, since the criteria for staging are abstract and qualitative, it is difficult to quantitatively evaluate the progression of the disease. Therefore, there is a need for a method to obtain data that can quantitatively evaluate the progression of the disease.

[0015] As a method for acquiring a pathological tissue image, a method using a system called a virtual slide system (Whole Slide Image: WSI) is known. The virtual slide system is a device capable of acquiring a pathological tissue image (virtual slide) obtained by digitally archiving a pathological specimen using a predetermined device. Conventionally, virtual slides have been used for storing pathological specimen images in medical institutions and research institutions, and for sharing and observing samples in classes or training sessions.

[0016] The present disclosure provides a computer program and a control device capable of executing a method for providing data quantitatively indicating the progression state of fibrosis of a disease. The system according to the present disclosure applies the spatial analysis methods of remote sensing and geographic information system (GIS) in geography to the analysis of pathological tissue images obtained by a system such as a virtual slide system. By using this method, the system performs a process focusing on the connectivity or fragmentation of fibers in the tissue, and generates quantitative data regarding the state of fibrosis of the tissue. A user can quantitatively analyze the state of fibrosis of the tissue using this data.

[0017] (Embodiment) [Configuration] While referring to FIG. 1, the data acquisition system 1 according to an embodiment of the present disclosure will be described. FIG. 1 is a schematic diagram of the data acquisition system 1 according to an embodiment of the present disclosure.

[0018] As shown in FIG. 1, the data acquisition system 1 includes a control device 10 and an imaging device 20. The control device 10 is, for example, a computer. The control device 10 includes an arithmetic circuit 11, a storage device 12, a communication circuit 13, an input device 14, and an output device 15.

[0019] The arithmetic circuit 11 controls the overall operation of the control device 10. The arithmetic circuit 11 may be configured in a manner in which hardware resources and software cooperate to realize a predetermined function, or in a manner in which a predetermined function is realized using a specially designed hardware circuit.

[0020] As an example of the former, the arithmetic circuit 11 includes a general-purpose processor such as a CPU or MPU that realizes predetermined processing or functions by executing a program. The arithmetic circuit 11 is configured to communicate with the storage device 12. The arithmetic circuit 11 realizes various functions in the control device 10 by reading and executing arithmetic programs etc. stored in the storage device 12. As an example of the latter, the arithmetic circuit 11 includes FPGAs and ASICs. As can be understood from the above, the arithmetic circuit 11 can be realized using semiconductor integrated circuits such as CPUs, MPUs, GPUs, FPGAs, DSPs, and ASICs.

[0021] The storage device 12 is a device equipped with a storage medium that can store various types of information in the storage medium. This information includes programs and data. For example, the storage device 12 stores calculation programs for realizing various functions according to this embodiment. The storage device 12 can also store images acquired from the imaging device 20, and data acquired or generated by performing processing on said images. The storage device 12 can be realized, for example, as a volatile or non-volatile semiconductor memory such as DRAM, SRAM, or flash memory, an SSD, an HDD, or other storage device, or a combination thereof as appropriate.

[0022] The communication circuit 13 is an interface device for connecting to other devices or systems via a communication line, either wired or wirelessly. The interface device can perform communication compliant with wired communication standards such as USB® or Ethernet®. Furthermore, the interface device can perform communication compliant with wireless communication standards such as Wi-Fi®, Bluetooth®, or cellular networks.

[0023] The input device 14 has the function of inputting information from the user to the control device 10. The input device 14 includes one or more human-machine interface devices. The human-machine interface devices include, for example, a keyboard, a pointing device (mouse, trackball, etc.), a touchpad, etc.

[0024] The output device 15 has the function of outputting information from the control device 10 to the user. The output device 15 includes one or more human-machine interface devices. The human-machine interface devices include, for example, output devices such as a display and a speaker. The human-machine interface devices may also include input / output devices such as a display equipped with an in-cell type touch panel (e.g., a liquid crystal panel or an organic EL panel).

[0025] The imaging device 20 has the function of capturing images of a sample, generating images, and transmitting the generated images to the control device 10. The imaging device 20 is, for example, a virtual slide system. In this embodiment, the imaging device 20 can capture images of one or more samples placed on a slide. The imaging device 20 can capture images of the placed slide and generate slide images. The slide images include an overall image of each sample.

[0026] [Operation] The following describes an overview of the operation of the control device 10 according to this embodiment. Figure 2 is a flowchart showing an example of processing by the control device 10.

[0027] First, the arithmetic circuit 11 of the control device 10 acquires a slide image from the imaging device 20 (S10). The imaging device 20 images a slide on which one or more samples are placed and acquires a slide image that includes an image of the entire sample. The arithmetic circuit 11 of the control device 10 receives the slide image from the imaging device 20 via the communication circuit 13 and stores it in the storage device 12. Then, as will be described later, the arithmetic circuit 11 acquires sample images for each sample from the slide image.

[0028] Figure 3 shows an example of a slide image captured by the imaging device 20. The slide image 30 shown in Figure 3 includes images of six samples 31a to 31f.

[0029] The imaging device 20 may be configured to acquire multiple slide images captured at different magnifications. The control device 10 may be configured to receive multiple slide images. For example, when the control device 10 receives multiple slide images, it may be configured to perform the processing described later using the slide image with the highest magnification among the multiple slide images. The control device 10 may also be configured to use slide images with other magnifications.

[0030] Upon acquiring the slide image, the arithmetic circuit 11 acquires the sample object (S11). In this specification, the sample object means vector data having the shape of the sample.

[0031] First, the arithmetic circuit 11 converts the slide image into a grayscale image with a predetermined number of bits, for example, 8 bits. Then, the arithmetic circuit 11 converts the brightness values ​​of the grayscale image into binary values ​​of 0 or 1 using a predetermined threshold, thereby converting the grayscale image into a binary image. A binary image is, for example, an image in which the pixel values ​​representing the sample in the slide image are converted to 1, and the values ​​of the other pixels are converted to 0. The arithmetic circuit 11 can extract the region where the pixel value is 1 as the sample region. Based on the shape of the sample region, the arithmetic circuit 11 can acquire the sample object.

[0032] If a slide image contains multiple samples, the arithmetic circuit 11 can acquire multiple sample objects corresponding to each of the multiple samples. The arithmetic circuit 11 can identify multiple sample objects, for example, by assigning a unique identifier to each sample. In this embodiment, the sample object is vector data with a planar shape. The arithmetic circuit 11 can place the sample objects in a unique space and execute the processing described later.

[0033] The arithmetic circuit 11 may be configured to remove noise from a binary image using any image processing method. For example, the arithmetic circuit 11 may be configured to remove noise from a binary image by performing morphological processing. Alternatively, the arithmetic circuit 11 may be configured to calculate the area of ​​regions with a pixel value of 1 and regions with a pixel value of 0, and remove regions with an area below a predetermined threshold as noise.

[0034] The arithmetic circuit 11 can acquire a sample image by combining the acquired slide image with the sample object. The sample image is an image representing the sample. For example, the arithmetic circuit 11 can acquire pixels in the slide image that overlap with the sample object as the sample image. If the slide image contains multiple samples, the arithmetic circuit 11 can acquire multiple sample images, similar to the sample object. Each of the multiple sample images can be associated with one of the multiple sample objects based on an identifier. In this way, the arithmetic circuit 11 of the control device 10 according to this embodiment can acquire a sample image that contains the entire sample being observed in a single image. Since processing described later is performed using this image, the arithmetic circuit 11 performs processing on the entire sample, not on a local area of ​​the sample. Because there is no need to select the range to be processed in the image representing the sample, the influence of bias in the areas where fibrosis is occurring within the sample can be eliminated.

[0035] The calculation circuit 11 assigns a grid-like mesh to the slide image. In this specification, this mesh is also referred to as a region mesh. The region mesh forms a matrix of quadrilaterals formed by four line segments. The calculation circuit 11 can combine the region mesh with the sample object to extract the mesh within the sample object. In this specification, the mesh within the sample object is also referred to as a sample mesh.

[0036] Next, the calculation circuit 11 extracts multiple fibrotic regions (S12). The calculation circuit 11 extracts multiple fibrotic regions from the sample image by identifying regions that have been previously identified using a predetermined method. In this embodiment, the sample is stained with Sirius Red. Therefore, the fibrotic tissue in the sample is stained red. Figure 4 is an enlarged view showing a part of the sample 31a in Figure 3. Part 32 in Figure 4 is an example of tissue that has been stained red. Tissue stained red with Sirius Red is an example of a region that has been previously identified.

[0037] When a sample is stained with Sirius Red, fibrous tissue within the sample may be stained red, while other tissues within the sample may be stained yellow. The staining of the sample allows for visual identification of the fibrous tissue. Before extracting multiple fibrous regions, the calculation circuit 11 acquires a sample image representing the stained sample. In the sample image, the fibrous tissue within the sample is stained a different color from other tissues within the sample due to the staining process, allowing the calculation circuit 11 to identify the fibrous tissue. Therefore, the calculation circuit 11 can extract the identified fibrous tissue in the sample image as multiple fibrous regions.

[0038] The arithmetic circuit 11 identifies the sample texture through texture analysis. In this specification, sample texture means the texture of tissue identified in the sample in a predetermined manner. In this embodiment, the tissue of the sample is identified by staining with Sirius Red. The arithmetic circuit 11 identifies the stained areas in the sample image that are stained red with Sirius Red. The arithmetic circuit 11 converts the RGB luminance values ​​of each pixel to other luminance values ​​by applying the following formula to each pixel, and obtains an image with a predetermined number of bits, for example, an 8-bit image. X = (RG) / B Equation 1 Here, X represents the brightness value after conversion according to Equation 1. R represents the red brightness value at each pixel. G represents the green brightness value at each pixel. B represents the blue brightness value at each pixel.

[0039] The arithmetic circuit 11 binarizes each pixel of the 8-bit image using a predetermined threshold to obtain a binary fibrillated image. For example, the arithmetic circuit 11 can set a statistical quantity based on the value of each pixel in the 8-bit image as the threshold. For example, the arithmetic circuit 11 can set 0.5σ of the pixel values ​​in the image as the threshold, where σ is the standard deviation. The arithmetic circuit 11 may also set any value as the threshold, such as 1σ, the mean, or the median. In this embodiment, the arithmetic circuit 11 converts the values ​​of pixels with values ​​greater than or equal to the predetermined threshold to 1, and converts the values ​​of other pixels to 0.

[0040] The arithmetic circuit 11 may, for example, perform morphological processing to convert a minute pixel having a value different from the surrounding values ​​to the same value as the surrounding values. Alternatively, if the area of ​​the region of a minute pixel having a value different from the surrounding values ​​is smaller than a predetermined threshold, the arithmetic circuit 11 may convert the value of that region to the same value as the surrounding values.

[0041] The arithmetic circuit 11 extracts multiple regions with a value of 1 from the binary fibrillation image, each as a fibrillation region. The sample texture represents these multiple regions. The sample texture may also represent pixels that represent regions identified in the sample by a predetermined method.

[0042] In this way, the calculation circuit 11 can identify pre-identified regions from the sample image and extract multiple fibrotic regions. In this embodiment, the fibrotic regions are vector data of a planar shape formed in a unique space.

[0043] Fibrosis regions can be constructed on the unique space described above. In a sample object placed on this unique space, multiple fibrous regions can each be positioned at locations corresponding to pre-identified regions in the sample image. Therefore, fibrous regions have positional information in this unique space. Thus, the data generated in this embodiment has positional information in this unique space. Therefore, data corresponding to specific tissues within a sample (e.g., fibrous regions, or fibrous sub-regions described later) can be positioned at equivalent locations in this unique space.

[0044] When the fibrotic region is extracted, the arithmetic circuit 11 constructs a fibrotic sub-region (S13). The arithmetic circuit 11 divides at least some of the multiple fibrotic regions and generates divided regions. Hereinafter, the divided fibrotic regions will be referred to as the first fibrotic region. The fibrotic regions other than the first fibrotic region will be referred to as the second fibrotic region. The arithmetic circuit 11 constructs multiple fibrotic sub-regions using the divided regions and the second fibrotic regions. Each of the multiple divided regions and the multiple second fibrotic regions corresponds to one of the multiple fibrotic sub-regions.

[0045] In this embodiment, the calculation circuit 11 constructs a fibrous sub-region by combining the sample mesh and a plurality of fibrous regions. Figure 5 is a flowchart showing an example of the process of constructing a fibrous sub-region. First, the calculation circuit 11 places a plurality of fibrous regions and the sample mesh in the same space and places the sample mesh on the plurality of fibrous regions (S20). The mesh 33 shown in Figure 4 is an example of a sample mesh. As shown in Figure 4, in this embodiment the mesh 33 is formed in a grid shape. The calculation circuit 11 divides one or more first fibrous regions, i.e., fibrous regions that overlap with the sample mesh, using the sample mesh as a boundary, and generates divided regions (S21). The calculation circuit 11 constructs each of the divided regions and the second fibrous regions as fibrous sub-regions (S22). Once the fibrous sub-regions are constructed, the calculation circuit 11 terminates the process of constructing the fibrous sub-regions (S13).

[0046] Figure 6 shows multiple fibrotic subregions constructed in a unique space. In Figure 6, regions 34a to 34c are examples of fibrotic subregions. The locations where regions 34a to 34c are constructed correspond to the area 32 stained red in Figure 4.

[0047] Next, the calculation circuit 11 sets a representative point for each of the multiple fibrillation sub-regions (S14). In this embodiment, the calculation circuit 11 calculates the centroid of each fibrillation sub-region in a unique space and determines the centroid as the position of the representative point for that fibrillation sub-region. Once the calculation circuit 11 calculates the position of the representative point, it places it in a unique space.

[0048] The calculation circuit 11 uses a set representative point to group together fibrillated sub-regions that are located within a predetermined distance from each other among a plurality of fibrillated sub-regions (S15). For example, the calculation circuit 11 uses a set representative point to connect two adjacent points among the plurality of representative points with a line segment to form a triangle formed by three line segments.

[0049] In this embodiment, the arithmetic circuit 11 groups adjacent fibrous subregions using the Delaunay triangulation algorithm. The arithmetic circuit 11 uses Delaunay triangulation to form multiple triangles, each of which has three adjacent representative points as its nodes (i.e., vertices). Hereinafter, these triangles will be referred to as Delaunay triangles. In a Delaunay triangle, two of the three nodes are connected by edges.

[0050] The arithmetic circuit 11 constructs a network using representative points as nodes of a Delaunay triangle. The arithmetic circuit 11 removes edges in this network that have a length greater than or equal to a predetermined threshold. As a result, the arithmetic circuit 11 can generate a divided network from a network in which all representative points are connected, in which representative points located within the predetermined threshold range are connected. In this specification, the divided network is an example of a subgraph of a Delaunay triangle. Each subgraph consists of nodes and edges that remain after the removal of the edges described above. A subgraph may consist of only one node, or only two nodes and edges connecting those nodes. Each subgraph is independent of the others.

[0051] The arithmetic circuit 11 assigns the same identifier to fibrillation sub-regions corresponding to one or more nodes that constitute a single subgraph. The arithmetic circuit 11 defines one or more fibrillation sub-regions assigned the same identifier as a single group. For example, the arithmetic circuit 11 groups multiple representative points located within a predetermined distance. In this way, the arithmetic circuit 11 can group multiple fibrillation sub-regions located within a predetermined distance. The arithmetic circuit 11 may define a particular fibrillation sub-region as a single group if there are no representative points of other fibrillation sub-regions located within a predetermined distance from a representative point of that particular fibrillation sub-region.

[0052] Figure 7 shows a diagram of multiple grouped fibrous subregions. As shown in Figure 7, a triangular network 35a to 35c, or multiple subgraphs 35a to 35c, representing multiple grouped fibrous subregions, is arranged within the sample object. The fibrous subregions 34a to 34c shown in Figure 6 correspond to the subgraphs 35a to 35c shown in Figure 7. The calculation circuit 11 removes edges with a length greater than a predetermined threshold to generate multiple subgraphs as shown in Figure 7.

[0053] Figure 8 is an enlarged view showing a portion of Figure 6. Figure 9 is a diagram representing a subgraph generated based on the fibrotic subregions shown in Figure 8.

[0054] Figure 8 shows a region containing multiple fibrillation sub-regions, including region 34a from the diagram shown in Figure 6. Figure 9 shows a region containing subgraph 35a from the diagram shown in Figure 7. As described above, the arithmetic circuit 11 performs representative point setting processing using each of the multiple fibrillation sub-regions and places each representative point at a position corresponding to the fibrillation sub-region in a unique space. Then, as described above, the arithmetic circuit 11 performs grouping processing and constructs a subgraph of the Delaunay triangle network using the multiple representative points. Subgraph 35a is a subgraph constructed using the representative points corresponding to fibrillation sub-region 34a. Subgraph 35a shown in Figure 9 contains 135 fibrillation sub-regions. In other words, subgraph 35a shows that 135 fibrillation sub-regions are grouped together as one group.

[0055] The calculation circuit 11 obtains statistics for grouped fibrillation subregions from multiple grouped fibrillation subregions (S16). The statistics are values ​​that represent an example of the fibrillation characteristics of the observed sample, calculated based on the grouped fibrillation subregions.

[0056] In this embodiment, the arithmetic circuit 11 calculates a statistic based on the number of groups and / or the number of fibrous subregions included in each group for a plurality of grouped fibrous subregions. For example, the arithmetic circuit 11 may calculate a statistic of at least one value selected from the group consisting of the number of groups, the median number of fibrous subregions included in each group, the standard deviation of the number of fibrous subregions included in each group, the median area of ​​the fibrous subregions included in each group, and the standard deviation of the area of ​​the fibrous subregions included in each group. The statistic is not limited to the above, and any value can be used. For example, the statistic may be the average number of fibrous subregions included in each group.

[0057] The calculation circuit 11 can display the acquired statistics to the user via the output device 15. For example, the user can use the displayed statistics to determine the degree of fibrosis progression in the sample under observation. The calculation circuit 11 may also be configured to transmit the acquired statistics to any device via the communication circuit 13 in a predetermined data format.

[0058] The calculation circuit 11 may score the disease state, i.e., the state of fibrosis, in the target sample based on the acquired statistics. The calculation circuit 11 may score the state of fibrosis in the target sample by comparing the statistics for the target sample with statistics for other samples that have been analyzed in advance. For example, data that associates statistics for other samples with the state of fibrosis determined by a pathologist for the sample in question may be stored in the storage device 12. The calculation circuit 11 can score the state of fibrosis in the target sample by comparing the acquired statistics with the statistics stored in the data.

[0059] The calculation circuit 11 may be configured to store predetermined criteria in the storage device 12 in advance, which associate statistics and scores for other samples that have been analyzed in advance, and to score the acquired statistics based on these criteria. Furthermore, the calculation circuit 11 can compare the statistics for each sample to evaluate the relationships between the samples. In this way, the statistics can serve as representative values ​​that represent the characteristics of each sample.

[0060] Scoring may mean classifying the sample into multiple stages according to the degree of fibrosis progression, and the calculation circuit 11 determining which stage to classify the sample into. Alternatively, scoring may mean that the calculation circuit 11 displays the fibrosis status on the output device 15 or transmits information about the status to other devices via the communication circuit 13.

[0061] The scoring method is not limited to the above, and any method can be used. For example, the calculation circuit 11 may score by obtaining statistics for multiple samples and calculating the similarity distance between each sample. Alternatively, the calculation circuit 11 may score by calculating principal component scores from statistics using principal component analysis.

[0062] Based on the statistics or scores obtained for multiple samples, the arithmetic circuit 11 may classify the multiple samples into multiple clusters using cluster analysis or by constructing a systematic network. For example, the arithmetic circuit 11 may classify the multiple samples into three clusters based on the standard deviation of the number of fibrotic sub-regions within each group. In this case, the three clusters may consist of cluster 1 with a large standard deviation, cluster 3 with a small standard deviation, and cluster 2 which does not belong to cluster 1 or cluster 3. In this specification, cluster 1 is an example of a cluster indicating a classification in which fibrosis is progressing. Cluster 3 is an example of a cluster indicating a classification in which fibrosis is not progressing much. Cluster 2 is an example of a cluster indicating a classification between cluster 1 and cluster 3.

[0063] The applicant used the data acquisition system 1 according to this embodiment to obtain a total of 42 samples from 21 mice, grouped each sample, and obtained statistical data. Six of the mice were administered a fibrosis-inducing diet for 14 weeks, followed by a normal diet for 4 weeks. A total of 15 samples were obtained from these six mice. The purpose of administering this diet was to induce fibrosis in the mice with the fibrosis-inducing diet and then improve the fibrosis with the subsequent normal diet.

[0064] The applicant obtained the number of groups in each sample and the standard deviation of the number of fibrotic subregions in each group as statistical measures. Comparing these statistics with the pathologist's evaluation of each sample for MASH (classification of each sample into one of four categories by the pathologist), the correlation between the number of groups and the evaluation result was negative (-0.65). The correlation between the standard deviation and the evaluation result was positive (0.5). The correlation between the number of groups and the standard deviation was negative (-0.61).

[0065] The above correlation indicates that a lower number of groups in the overall sample indicates more advanced MASH progression, and a higher number of groups indicates slower MASH progression. Furthermore, the above correlation indicates that a higher standard deviation indicates more advanced MASH progression, and a lower standard deviation indicates slower MASH progression. These results are consistent with the progression mechanism of MASH, indicating that the processing performed by the arithmetic circuit 11 is a method that corresponds to the mechanism of the disease.

[0066] Furthermore, when comparing the results of cluster analysis, which classified the 42 samples into three groups based on the statistics obtained, with the evaluation results by pathologists, the two results showed similar classification tendencies. For example, samples classified into cluster 1 by the calculation circuit 11 tended to be classified into a stage indicating progressive fibrosis in the pathologist's classification as well. Similarly, samples classified into cluster 3 by the calculation circuit 11 tended to be classified into a stage indicating minimal fibrosis in the pathologist's classification as well. Furthermore, samples classified into cluster 2 by the calculation circuit 11 tended to be classified into a similar stage in the pathologist's classification as well.

[0067] Thus, the classification based on the quantitative data acquired by the control device 10 according to this embodiment shows a similar trend to the conventional evaluation results by pathologists. Therefore, the control device 10 can acquire quantitative data that contains information about the connectivity or fragmentation of the fibrotic site, and this data more accurately represents the progression of fibrosis.

[0068] Furthermore, samples from mice administered the fibrosis-inducing diet and the normal diet were classified into cluster 3 (2 samples) and cluster 2 (13 samples). Thus, since most samples were classified into cluster 2 despite being administered the fibrosis-inducing diet, it is considered that the fibrosis in these mice improved. Additionally, the quantitative data obtained from these mouse samples may contain information about the improvement in fibrosis. Therefore, based on the quantitative data obtained by the control device 10 according to this embodiment, the user may be able to determine the trend of improvement in fibrosis.

[0069] [effect] The control device 10 and the computer program capable of executing the processing in the control device 10 according to the embodiments of this disclosure can achieve the following effects.

[0070] The computer program can cause the computer's arithmetic circuit 11 to execute the following method. The method includes extracting multiple fibrous regions from a sample image, which is an image of the sample. The method includes constructing multiple fibrous subregions, each containing a divided region generated by dividing a first fibrous region, which is at least a portion of the multiple fibrous regions, and a second fibrous region, which is a fibrous region other than the first fibrous region among the multiple fibrous regions. The method includes setting a representative point for each of the multiple fibrous subregions, and grouping fibrous subregions located within a predetermined distance from each other based on the representative point. The method includes obtaining statistics about the grouped fibrous subregions from the grouped fibrous subregions.

[0071] According to a computer program that performs this method, the arithmetic circuit 11 can obtain statistical data on the grouped fibrotic sub-regions as quantitative data. Therefore, a user of the computer program can obtain quantitative data on the fibrotic state of the sample. Because the user can use quantitative data, they can perform quantitative evaluations based on predetermined criteria. Since quantitative evaluations are performed based on predetermined criteria, intra-observer or inter-observer variability in evaluations of the same sample can be suppressed. The problem of differing diagnostic results among pathologists when diagnosing a given sample, which arises from reliance on the pathologist's experience and subjective judgment, can be resolved.

[0072] Furthermore, since representative points are set for each divided region obtained by dividing the first fibrotic region, multiple representative points may be set for a single fibrotic region. For example, if at least a portion of the first fibrotic region has a larger area compared to other fibrotic regions, multiple representative points will be set for one large fibrotic region, rather than just one. Since one first fibrotic region is grouped into multiple fibrotic sub-regions, the calculation circuit 11 can obtain statistics that more accurately reflect the progression of fibrosis compared to conventional methods. Therefore, users can analyze or evaluate the state of tissue fibrosis using more accurate quantitative data. In addition, since the method performed by the calculation circuit 11 also focuses on fibrosis fragmentation, users may be able to predict not only the worsening but also the improvement of the disease (e.g., MASH).

[0073] Because the state of fibrosis can be analyzed or evaluated using quantitative data, users can make objective judgments rather than subjective ones. Furthermore, evaluating based on the same criteria can reduce variability in evaluations among users. Therefore, the process required to evaluate the state of fibrosis in a sample can be streamlined.

[0074] Furthermore, since quantitative data on each sample is obtained without the pathologist visually inspecting each sample, the time required for the user to evaluate the fibrosis state of the sample is reduced. Also, since the above processing can be performed automatically, the calculation circuit 11 can obtain quantitative data on the sample without user involvement. The user can then use the quantitative data obtained by the calculation circuit 11 to perform evaluations, which can improve evaluation efficiency.

[0075] Furthermore, the calculation circuit 11 groups fibrotic sub-regions within a predetermined range, enabling the acquisition of statistics that consider the connectivity of the fibrotic sites. Therefore, the user can evaluate the progression of fibrosis in the sample using quantitative data that considers the connectivity of the fibrotic sites. In addition, the calculation circuit 11 can acquire statistics that consider the fragmentation of the fibrotic sites through this grouping. In recent years, the possibility of improving fibrosis has been suggested, and it is thought that fragmentation of fibers occurs during the improvement process. Therefore, in that case, the user can evaluate the improvement of fibrosis using the acquired quantitative data.

[0076] Furthermore, by comparing the acquired quantitative data with the evaluation results determined visually by pathologists based on conventional evaluation criteria, it is possible to quantitatively determine which elements pathologists focus on in their evaluations. Therefore, this quantitative data can be used to set parameters for training data for machine learning. A computer program that performs the method described herein can also be used to construct training data for machine learning.

[0077] In a computer program, when extracting multiple fibrotic regions, these regions are constructed in a unique space. Each of these fibrotic regions contains location information within that unique space.

[0078] According to such a computer program, since multiple fibrotic regions contain unique spatial positional information, the fibrotic sub-regions generated based on the fibrotic regions similarly contain unique spatial positional information. Therefore, the calculation circuit 11 can obtain statistics that take into account the connectivity of multiple fibrotic sub-regions. Thus, the user can evaluate the progression of fibrosis in the sample using quantitative data that takes into account the connectivity of the fibrotic sites.

[0079] In a computer program, constructing multiple fibrous subregions includes arranging a mesh in a unique space and dividing the first fibrous region with the mesh to generate divided regions. Constructing multiple fibrous subregions also includes constructing the divided region and the second fibrous region as fibrous subregions. The first fibrous region is a fibrous region that overlaps with the mesh among the multiple fibrous regions.

[0080] According to such a computer program, fibrotic regions overlapping with a mesh placed in a unique space are divided by the mesh. Therefore, the calculation circuit 11 can divide the fibrotic regions using a more objective method. Consequently, the calculation circuit 11 can objectively group multiple fibrotic sub-regions and obtain statistics that more accurately reflect the progression of fibrosis compared to conventional methods.

[0081] In sample images acquired by computer programs, fibrous tissue within the sample is stained a different color from other tissues in the sample due to the staining process.

[0082] According to such a computer program, the calculation circuit 11 can acquire a sample image in which the sample has been stained. Due to the staining, the fibrous tissue within the sample in the sample image is stained a different color from the other tissues in the sample, and can therefore be identified by the calculation circuit 11. Thus, the calculation circuit 11 can extract the identified fibrous tissue in the sample image as multiple fibrous regions.

[0083] In a computer program, extracting multiple fibrotic regions involves identifying sample textures from a sample image that represent the texture of tissue identified in the sample using a predetermined method, and extracting the regions indicated by these sample textures as multiple fibrotic regions.

[0084] According to such a computer program, the calculation circuit 11 can identify tissue regions in a sample image using a predetermined method. For example, by staining fibrotic tissue in a sample, the calculation circuit 11 can identify the fibrotic tissue regions from the sample image. The calculation circuit 11 can extract vector data indicating the region of the identified area as a fibrotic region. In this way, the calculation circuit 11 can acquire fibrotic regions representing the areas of fibrotic tissue using a predetermined method. Since the calculation circuit 11 can acquire quantitative data about the fibrotic state of the sample by performing a predetermined process on the fibrotic region, the user can use this quantitative data to make an objective evaluation. Furthermore, the calculation circuit 11 can extract fibrotic regions using the same method for each sample. Therefore, the calculation circuit 11 can acquire quantitative data that can be compared between multiple samples. Consequently, the user can use predetermined criteria to make an objective evaluation of the fibrotic state of each sample.

[0085] In computer programs, grouping involves grouping fibrous subregions located within a predetermined distance using Delaunay triangulation.

[0086] According to such a computer program, multiple fibrotic subregions can be grouped using quantitative methods. Therefore, the arithmetic circuit 11 can objectively group multiple fibrotic subregions and obtain statistics that more accurately reflect the progression of fibrosis compared to conventional methods.

[0087] In computer programs, grouping involves constructing a network using representative points as nodes of Delaunay triangles formed by Delaunay triangulation. Grouping also involves removing edges between two nodes in the network that have a length greater than or equal to a predetermined threshold. Grouping also involves assigning the same identifier to one or more fibrillated subregions corresponding to independent subgraphs composed of the remaining nodes and edges, and defining fibrillated subregions with the same identifier as a single group.

[0088] According to such a computer program, the arithmetic circuit 11 can divide the network constructed by Delaunay triangulation into multiple subgraphs based on a predetermined threshold distance. The arithmetic circuit 11 can group fibrotic subregions that are separated by a predetermined threshold distance or more into different groups, thus enabling grouping that takes into account the connectivity or fragmentation of the fibrotic areas. Therefore, the arithmetic circuit 11 can obtain statistics that more accurately reflect the progression of fibrosis compared to conventional methods.

[0089] In computer programs, statistics are calculated based on the number of groups and the number of fibrotic subregions included in each group.

[0090] According to such a computer program, the arithmetic circuit 11 can acquire quantitative data based on the number of groups of grouped fibrotic sub-regions and the number of fibrotic sub-regions contained in each group. For example, the arithmetic circuit 11 can calculate statistical quantities such as the standard deviation, median, and mean based on these numbers. Because the state of fibrosis can be analyzed or evaluated using quantitative data, users can make objective judgments rather than subjective ones. Furthermore, by evaluating based on the same criteria, variability in evaluations among users can be reduced.

[0091] The method executed by the arithmetic circuit 11 based on the computer program further includes scoring the disease state in the sample based on statistics.

[0092] According to such a computer program, the arithmetic circuit 11 can calculate the state of a disease as objective data based on quantitative data. For example, the arithmetic circuit 11 can classify or quantify the state of a disease in a sample by combining quantitative data with predetermined criteria. The user can objectively evaluate the state of the disease in the sample based on the score determined by the arithmetic circuit 11. Furthermore, since the arithmetic circuit 11 can determine a quantitative score based on quantitative data, the evaluation result based on that score is expressed as a ratio scale. Therefore, it can be evaluated by comparing it with the evaluation results of other samples.

[0093] The method executed by the arithmetic circuit 11 based on the computer program further includes acquiring a slide image of the slide on which the sample is placed, and the sample image is acquired from the slide image.

[0094] According to such a computer program, the calculation circuit 11 can acquire a sample image of a sample placed on a slide. The calculation circuit 11 can acquire a slide image of a slide placed on an imaging device 20, such as a virtual slide system. Therefore, the calculation circuit 11 can use images acquired from any imaging device 20 to obtain quantitative data indicating the fibrosis state of the sample.

[0095] The control device 10 includes an arithmetic circuit 11. The arithmetic circuit 11 extracts multiple fibrous regions from a sample image, which is an image of the sample. The arithmetic circuit 11 constructs multiple fibrous subregions, each containing a divided region, which is a first fibrous region that is at least a portion of the multiple fibrous regions, and a second fibrous region, which is a fibrous region other than the first fibrous region. The arithmetic circuit 11 sets a representative point for each of the multiple fibrous subregions and groups the fibrous subregions located within a predetermined distance from each other based on the representative point. The arithmetic circuit 11 obtains statistics about the grouped fibrous subregions from the grouped fibrous subregions.

[0096] With this configuration, the control device 10 can acquire statistical data on grouped fibrotic sub-regions as quantitative data. Therefore, the control device 10 can acquire quantitative data on the fibrotic state of the sample. Since users of the control device 10 can use quantitative data, they can perform quantitative evaluations based on predetermined diagnostic criteria. This can eliminate the problem of differing diagnostic results among pathologists when diagnosing a given sample, which is caused by reliance on the pathologist's experience and subjective judgment.

[0097] Furthermore, since representative points are set for each divided region obtained by dividing the first fibrotic region, multiple representative points may be set for a single fibrotic region. For example, if at least a portion of the first fibrotic region has a larger area compared to other fibrotic regions, multiple representative points will be set for one large fibrotic region, rather than just one. Since one first fibrotic region is grouped into multiple fibrotic sub-regions, the control device 10 can obtain statistics that more accurately reflect the progression of fibrosis compared to conventional methods. Therefore, users can analyze or evaluate the state of tissue fibrosis using more accurate quantitative data. Because users can analyze or evaluate the state of fibrosis using quantitative data, they can make objective judgments rather than subjective ones. Also, by evaluating based on the same criteria, variability in evaluations among users can be reduced. (modified version)

[0098] In the embodiments described above, tissue images are acquired by a virtual slide system, but the method of acquiring tissue images is not limited to a virtual slide system. Tissue images can be acquired by any method, such as using a biological microscope.

[0099] In the embodiments described above, fibrotic tissue in the sample is stained with Sirius Red, but the method for identifying the tissue is not limited to Sirius Red staining. For example, the sample may be stained with Azann Mallory staining or Masson's trichrome staining. Thus, the sample can be stained with any dye. Furthermore, formula 1 above, which is used to calculate the stained area, may be modified depending on the staining method.

[0100] In the embodiments described above, the region mesh and sample mesh are configured such that quadrilaterals are formed by vertical and horizontal line segments, but the mesh is not limited to this shape. For example, the mesh may be configured such that triangles or other polygons are formed by multiple line segments.

[0101] In the embodiment described above, the calculation circuit 11 divides the fibrous region using a sample mesh, but the process of dividing the fibrous region is not limited to a process using a sample mesh. For example, the calculation circuit 11 may be configured to divide a fibrous region having an area greater than or equal to a predetermined threshold into one or more regions to generate divided regions.

[0102] In the embodiment described above, the calculation circuit 11 sets the centroid of each fibrillated sub-region as the representative point of that fibrillated sub-region, but the representative point is not limited to the centroid of the fibrillated sub-region. The setting of the representative point is not limited to the method using the centroid, and any method can be used. For example, the calculation circuit 11 may determine the midpoint of the inscribed circle of each fibrillated sub-region, the geometric median, or the pole that is unreachable as the representative point. For example, the calculation circuit 11 can determine the geometric median of each fibrillated sub-region by calculating the median for multiple constituent points based on the X and Y coordinates of the constituent points of multiple line segments that constitute the fibrillated sub-region.

[0103] In the embodiment described above, the arithmetic circuit 11 grouped multiple fibrous subregions using Delaunay triangulation, but the grouping method is not limited to the method using Delaunay triangulation. For example, the arithmetic circuit 11 may group multiple fibrous subregions using any clustering method such as DBSCAN or OPTICS.

[0104] In the embodiment described above, the calculation circuit 11 acquires quantitative data on MASH, a disease related to the liver, but the target organ is not limited to the liver. The calculation circuit 11 can acquire quantitative data on fibrosis in organs where fibrosis can occur, such as the lungs, heart, and liver.

[0105] In the embodiment described above, the arithmetic circuit 11 acquires an image of the entire sample as the sample image, but the sample image is not limited to an image of the entire sample. The arithmetic circuit 11 may be configured to perform the above-described processing on an image of only a part of the sample.

[0106] (Summary of characteristics) As is clear from the above description, this disclosure includes the following embodiments. Hereafter, reference numerals are enclosed in parentheses solely to indicate their correspondence with the embodiments.

[0107] (Aspect 1) The computer program relating to the present disclosure is a computer program that causes a computer's arithmetic circuit (11) to execute a predetermined method, Extracting multiple fibrotic regions from a sample image, To construct multiple fibrillation subregions, each including a divided region generated by dividing a first fibrillation region, which is at least a portion of the multiple fibrillation regions, and a second fibrillation region, which is a fibrillation region other than the first fibrillation region among the multiple fibrillation regions, To set a representative point for each of the aforementioned multiple fibrotic subregions, and Based on the aforementioned representative point, the fibrotic subregions located within a predetermined distance from the plurality of fibrotic subregions are grouped together. Obtaining statistical data related to the grouped fibrosis subregions from the grouped fibrosis subregions, The computer's arithmetic circuit is made to execute a method that includes this.

[0108] (Aspect 2) In the computer program of Aspect 1, in extracting the plurality of fibrotic regions, the plurality of fibrotic regions are constructed in a unique space, and each of the plurality of fibrotic regions may include positional information in the unique space.

[0109] (Aspect 3) In the computer program of Aspect 1 or Aspect 2, constructing the plurality of fibrillated subregions is: Placing a mesh in a unique space, The first fibrous region is divided by the mesh to generate the divided region, The divided region and the second fibrosis region are constructed as the fibrosis sub-regions, Includes, The first fibrous region may be a fibrous region among the plurality of fibrous regions that overlaps with the mesh.

[0110] (Aspect 4) In any of the computer programs in aspects 1 to 3, the fibrotic tissue in the sample image may be stained a different color from the other tissues in the sample by the staining of the sample.

[0111] (Aspect 5) In any of the computer programs from Aspects 1 to 4, extracting the plurality of fibrotic regions may include identifying a sample texture from the sample image that shows the texture of the tissue identified in the sample by a predetermined method, and extracting the region shown by the sample texture as the plurality of fibrotic regions.

[0112] (Aspect 6) In any of the computer programs of aspects 1 to 5, grouping may include grouping fibrotic subregions located within a predetermined distance using Delaunay triangulation.

[0113] (Aspect 7) In the computer program of Aspect 6, grouping is Construct a network using the aforementioned representative points as nodes of the Delaunay triangles formed by the Delaunay triangulation, To remove edges between two nodes in the aforementioned network that have a length exceeding a predetermined threshold, Assigning the same identifier to one or more fibrillated subregions corresponding to independent subgraphs composed of remaining nodes and edges, The fibrotic subregions having the same identifier are defined as a single group, It may include.

[0114] (Aspect 8) In any of the computer programs from Aspect 1 to Aspect 7, the statistical quantity may be calculated based on the number of groups and the number of fibrillated subregions included in each group.

[0115] (Aspect 9) In any of the computer programs of aspects 1 to 8, the method may further include scoring the disease state in the sample based on the statistical quantity.

[0116] (Aspect 10) In any of the computer programs of aspects 1 to 9, the method further includes acquiring a slide image of the slide on which the sample is placed, and the sample image may be acquired from the slide image.

[0117] (Aspect 11) The control device (10) according to the present disclosure comprises an arithmetic circuit (11), and the arithmetic circuit is Extracting multiple fibrotic regions from a sample image, To construct multiple fibrillation subregions, each including a divided region generated by dividing a first fibrillation region, which is at least a portion of the multiple fibrillation regions, and a second fibrillation region, which is a fibrillation region other than the first fibrillation region among the multiple fibrillation regions, To set a representative point for each of the aforementioned multiple fibrotic subregions, and Based on the aforementioned representative point, the fibrotic subregions located within a predetermined distance from the plurality of fibrotic subregions are grouped together. Obtaining statistical data related to the grouped fibrosis subregions from the grouped fibrosis subregions, Execute this.

[0118] In this specification, terms such as “First,” “Second,” etc., are used for illustrative purposes only and should not be understood as expressing or implying relative importance or ranking of technical features. Features designated as “First” or “Second” express or imply that they include one or more such features.

[0119] The control device described in this disclosure is realized through the cooperation of hardware resources, such as a processor and memory, and software (computer programs). [Industrial applicability]

[0120] This disclosure provides a computer program and control device capable of performing a method for providing data that quantitatively indicates the progression of disease fibrosis, and is therefore suitably applicable in this type of industrial field. [Explanation of Symbols]

[0121] 1. Data Acquisition System 10 Control device 11 Arithmetic circuit 12 Storage device 20 Imaging device

Claims

1. Extracting multiple fibrotic regions from a sample image, To construct a plurality of fibrous subregions, which include a divided region generated by dividing a first fibrous region, which is at least a portion of the plurality of fibrous regions, and a second fibrous region, which is a fibrous region other than the first fibrous region among the plurality of fibrous regions, To set a representative point for each of the aforementioned multiple fibrotic subregions, and Based on the aforementioned representative point, the fibrotic subregions located within a predetermined distance from the plurality of fibrotic subregions are grouped together. Obtaining statistical data related to the grouped fibrosis subregions from the grouped fibrosis subregions, A computer program that causes a computer's arithmetic circuit to execute a method that includes [a specific method].

2. The computer program according to claim 1, wherein, in extracting the plurality of fibrotic regions, the plurality of fibrotic regions are constructed in a unique space, and each of the plurality of fibrotic regions includes positional information in the unique space.

3. Constructing the aforementioned multiple fibrotic subregions is Placing a mesh in a unique space, The first fibrous region is divided by the mesh to generate the divided region, The divided region and the second fibrous region are constructed as the fibrous sub-regions, Includes, The first fibrous region is a fibrous region among the plurality of fibrous regions that overlaps with the mesh. The computer program according to claim 1.

4. The computer program according to claim 1, wherein in the sample image, the fibrotic tissue within the sample is stained a different color from other tissues within the sample by the staining of the sample.

5. The computer program according to claim 1, wherein the extraction of the plurality of fibrotic regions includes identifying a sample texture from the sample image that shows the texture of tissue identified in the sample by a predetermined method, and extracting the region shown by the sample texture as the plurality of fibrotic regions.

6. The computer program according to claim 1, wherein the grouping includes grouping fibrous subregions located within a predetermined distance using Delaunay triangulation.

7. Grouping is Construct a network using the aforementioned representative points as nodes of the Delaunay triangles formed by the Delaunay triangulation, To remove edges between two nodes in the aforementioned network that have a length exceeding a predetermined threshold, Assigning the same identifier to one or more fibrillated subregions corresponding to independent subgraphs composed of remaining nodes and edges, The fibrotic subregions having the same identifier are defined as a single group, The computer program according to claim 6, including the computer program described in claim 6.

8. The computer program according to claim 1, wherein the aforementioned statistics are calculated based on the number of groups and the number of fibrillated subregions included in each group.

9. The computer program according to claim 1, further comprising scoring the disease state in the sample based on the statistical quantity.

10. The computer program according to claim 1, further comprising the method of acquiring a slide image of a slide on which the sample is placed, wherein the sample image is acquired from the slide image.

11. A control device comprising an arithmetic circuit, wherein the arithmetic circuit is Extracting multiple fibrotic regions from a sample image, To construct a plurality of fibrous subregions, which include a divided region generated by dividing a first fibrous region, which is at least a portion of the plurality of fibrous regions, and a second fibrous region, which is a fibrous region other than the first fibrous region among the plurality of fibrous regions, To set a representative point for each of the aforementioned multiple fibrotic subregions, and Based on the aforementioned representative point, the fibrotic subregions located within a predetermined distance from the plurality of fibrotic subregions are grouped together. Obtaining statistical data related to the grouped fibrosis subregions from the grouped fibrosis subregions, A control device that performs this operation.

Images