Data analysis device and data analysis method

Abstract

To provide a data analysis apparatus which can reduce the time and effort of a user when executing a plurality of scripts.SOLUTION: An image analysis apparatus 100 (data analysis apparatus) comprises a control unit 10 which performs control of receiving an operation of selecting image data 410 (analysis data) acquired by an image acquisition unit 200 (analysis device), control of receiving an operation of selecting a plurality of scripts performing analysis for the selected image data 410 and control of executing analysis in parallel with the plurality of scripts selected for the selected image data 410. The image analysis apparatus 100 comprises a display unit 20 which displays on the same screen an analysis result acquired with the analysis of the image data 410 by the control unit 10.SELECTED DRAWING: Figure 3

Description

[Technical Field]

[0001] The present invention relates to a data analysis device and a data analysis method. [Background technology]

[0002] BACKGROUND ART Conventionally, a data analysis device and a data analysis method for analyzing image data using a script are known (see, for example, Patent Document 1).

[0003] The above-mentioned Patent Document 1 discloses a cell image analysis device (data analysis device) for analyzing cell images. Here, the cell image may contain removal targets such as impurities that should be removed. The cell image analysis device described in the above-mentioned Patent Document 1 generates a labeled image that indicates the location of the removal target region in the cell image. Then, a set of the cell image and the labeled image is used as a dataset for machine learning. The above-mentioned cell image analysis device has registered therein a plurality of learning models (scripts) for identifying predetermined removal targets that have been generated by the above-mentioned machine learning. Then, processing of one learning model selected by a user from the plurality of registered learning models is executed. [Prior art documents] [Patent documents]

[0004] [Patent Document 1] International Publication No. 2019 / 180848 Summary of the Invention [Problem to be solved by the invention]

[0005] However, in the above-mentioned Patent Document 1, as described above, processing of one learning model (script) selected by the user from multiple registered learning models is executed. In this case, if it is desired to obtain analysis results for each of the multiple learning models, it is necessary to execute processing for each of the multiple learning models individually. Therefore, it is considered that the user's workload increases by the number of learning models (scripts) to be executed. Therefore, there is a need for a data analysis device and a data analysis method that can reduce the user's workload when multiple scripts are executed in parallel.

[0006] The present invention has been made to solve the above-mentioned problems, and one object of the present invention is to provide a data analysis device and a data analysis method that can reduce the user's effort when executing multiple scripts. [Means for solving the problem]

[0007] In order to achieve the above object, a data analysis device in a first aspect of the present invention includes a control unit that performs control to accept an operation to select analytical data acquired by an analytical device, control to accept an operation to select a plurality of scripts that will analyze the selected analytical data, and control to execute analysis in parallel using the selected plurality of scripts on the selected analytical data, and a display unit that displays on the same screen analysis results acquired by the analysis of the analytical data by the control unit. The control unit performs a further analysis using the analysis result of at least one of the analysis data that has been analyzed in parallel by the multiple scripts as input data, and displays the analysis result on the same screen of the display unit. Note that a script is a simple program that can be executed by a computer. Also, "performing analysis in parallel" means "performing analysis in parallel."

[0008] A data analysis method according to a second aspect of the present invention includes the steps of: accepting a selection of analytical data acquired by an analytical device; selecting a plurality of scripts for performing an analysis on the analytical data; and performing an analysis in parallel on the selected analytical data using the selected plurality of scripts. a step of displaying on the same screen the analysis results of the parallel analysis by the plurality of scripts and the analysis results of a further analysis performed using as input data the analysis results of at least one analytical data analyzed in parallel by the plurality of scripts; Equipped with. [Effects of the Invention]

[0009] In the data analysis device according to the first aspect, as described above, control is performed to execute analysis in parallel using the selected scripts for the selected analytical data. This eliminates the need to individually execute processing for each of the multiple scripts for a single piece of data, thereby reducing the user's workload. Therefore, when executing multiple scripts, the user's workload can be reduced. Furthermore, the execution time can be shortened compared to when the process is individually executed for each of the multiple scripts. Furthermore, the same effect can be achieved in the data analysis method according to the second aspect, which includes the step of executing analysis in parallel using the selected scripts for the selected analytical data. [Brief explanation of the drawings]

[0010] [Figure 1] 1 is a diagram illustrating a configuration of an image analysis device according to an embodiment. [Figure 2] FIG. 10 is a diagram showing a setting screen for an analysis recipe according to an embodiment. [Figure 3] FIG. 10 is a diagram showing a screen on which image data and a script selected on a setting screen for an analysis recipe are displayed according to an embodiment. [Figure 4] FIG. 10 is a diagram showing an example of a combination of scripts in an analysis recipe. [Figure 5] FIG. 10 is a flow diagram showing an analysis flow by a control unit according to an embodiment. [Figure 6] FIG. 10 is a diagram illustrating an example of an analysis recipe according to an embodiment. [Figure 7] FIG. 10 is a diagram showing an example of an analysis recipe according to a first modified example of an embodiment. [Figure 8] FIG. 10 is a diagram illustrating an example of an analysis recipe according to a second modified example of an embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0011] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, specific embodiments of the present invention will be described with reference to the accompanying drawings.

[0012] The configuration of an image analysis device 100 according to this embodiment will be described with reference to Figures 1 to 6. Note that the image analysis device 100 is an example of the "data analysis device" in the claims.

[0013] (Configuration of image analysis device) The image analyzing device 100 includes a control unit 10 and a display unit 20. The image analyzing device 100 also includes a memory unit 30 that stores image data and the like acquired by an external image acquiring unit 200. The memory unit 30 is configured, for example, with a hard disk or a flash memory. The display unit 20 displays the analysis results acquired by the data analysis by the control unit 10. The image data and the like acquired by the image acquiring unit 200 may be stored in a device external to the image analyzing device 100, or may be stored in a memory area (e.g., cache memory) of the image analyzing device 100 that is different from the memory unit 30. The image analyzing device 100 itself may be provided with a mechanism for acquiring image data for analysis. The image acquiring unit 200 is an example of an "analysis device" in the claims.

[0014] The control unit 10 is configured to function as an image selection receiving means 1, a script selection receiving means 2, a data analysis means 3, an analysis recipe generation means 4, a property registration means 5, and a script registration means 6. In the control unit 10, the functions of the image selection receiving means 1, the script selection receiving means 2, the data analysis means 3, the analysis recipe generation means 4, the property registration means 5, and the script registration means 6 can be realized by software such as a program.

[0015] As shown in FIG. 2, the analysis recipe generation means 4 (control unit 10) controls the generation of an analysis recipe 300, which is a serial combination of multiple analysis steps for analyzing image data 410. Specifically, the analysis recipe 300 includes an analysis processing step (Step 1) that performs analysis processing using a trained model and an analysis algorithm, and a finishing processing step (Step 2) that performs finishing processing (described later). That is, the script (described later) includes a trained model generated by machine learning for analyzing the image data 410, an analysis algorithm used to analyze the image data 410, and processing algorithms (first processing algorithm and second processing algorithm) used for finishing processing (described later). When a trained model that can exclude images inappropriate for analysis is selected, a more appropriate analysis can be performed using the analysis algorithm. The analysis recipe 300 generated by the analysis recipe generation means 4 (control unit 10) is stored (registered) in the storage unit 30. The analysis recipe 300 and the image data 410 are examples of the "analysis flow" and "analysis data" in the claims, respectively. A specific method for generating the analysis recipe 300 is described below.

[0016] The image selection receiving means 1 (controller 10) controls the reception of an operation to select image data 410 to be analyzed. Specifically, as shown in FIG. 2, when a button 311a labeled "Select Image" in the upper column 311 of the "Test Image" column 310 is selected on the setting screen of the analysis recipe 300, a window is displayed in which image data 410 to be analyzed can be selected from multiple image data stored in the storage unit 30. Furthermore, when a button 311b labeled "Register Image" in the upper column 311 is selected, a window is displayed in which image data not stored in the storage unit 30 (for example, stored on a local PC) can be selected (registered) as image data 410 to be analyzed. The image data 410 whose selection has been received is displayed in the lower column 312 of "Test Image." In the example shown in FIG. 3, image data 410 displaying a cell 400 is displayed in the lower column 312.

[0017] Furthermore, when the button 311b labeled "Register Image" is selected and the selection of image data 410 is accepted, input of accompanying information for the image data 410 is accepted. The selected image data 410 and the input accompanying information are then associated with each other and stored (registered) in the storage unit 30. As an example of registering image data 410, the registration of image data of a cell culture well will be described. In this case, for example, it may be possible to input the number of cell passages (an operation of removing the culture medium from a culture system and transferring the cells to a new culture medium) and the number of days of culture as accompanying information.

[0018] Furthermore, based on the accompanying information, the control unit 10 performs control to group the plurality of image data 410 according to conditions. For example, the control unit 10 performs control to divide the plurality of image data 410 to be analyzed into groups with the same cell passage number, groups with the same number of days in culture, groups with both the same cell passage number and the same number of days in culture, etc.

[0019] The script selection receiving means 2 (control unit 10) controls the reception of an operation to select multiple scripts to analyze the selected image data 410. For example, on the setting screen for the analysis recipe 300, when a button 321a labeled "+Add Script" in the upper column 321 of the "Analysis Process" column 320 is selected, a window is displayed in which a script to be executed can be selected from multiple scripts stored in the storage unit 30. At this time, the script to be executed can be selected from the trained models and analysis algorithms registered in the storage unit 30. The control unit 10 analyzes the image data 410 using the selected script. The analysis results of the selected script are then displayed in the lower column 322 of "Analysis Process."

[0020] In this embodiment, the script selection receiving means 2 (control unit 10) controls the execution of analysis in parallel using multiple selected scripts for the selected image data 410. Specifically, the script selection receiving means 2 (control unit 10) is configured to control the execution of analysis in parallel using multiple selected trained models for the selected image data 410, and the execution of analysis in parallel using multiple selected analysis algorithms for the selected image data 410. In particular, when script A and script B are selected in a window displayed by selecting button 321a, the selection of both script A and script B is accepted. In this case, the data analysis means 3 (control unit 10) executes analysis in parallel using script A and script B. Note that while FIG. 3 shows an example in which two scripts (A and B) are selected, only one script or three or more scripts may be selected.

[0021] Specifically, in the example analysis recipe 301 shown in Fig. 4(A), only script A is selected. In the example analysis recipe 302 shown in Fig. 4(B), scripts A1 to A4 and script B are combined in parallel, and scripts A1 to A4 are combined in series with each other.

[0022] In this embodiment, the control unit 10 is configured to control the display of the analysis results of two or more selected scripts on the same screen of the display unit 20. Specifically, the analysis results (data 420, data 430) of script A and script B selected in the analysis processing step (Step 1) are displayed in parallel in a column 322 below "Analysis Processing." The analysis results (data 450, data 460) selected in the finishing processing step (Step 2) described below are also displayed on the same screen as the analysis results of the scripts selected in the analysis processing step. That is, all analysis results of the scripts included in the analysis recipe 300 are displayed on the same screen. Also, FIG. 3 illustrates, as an example of data 420, the analysis results of a binarization processing algorithm for distinguishing cells 400 from the background. Also, FIG. 3 illustrates, as an example of data 430, the analysis results of an analysis algorithm for color-coding cells 400 based on differences in the characteristics of the cells 400 (e.g., particle diameter, area, etc.).

[0023] The control unit 10 is configured to execute another analysis algorithm that is serially combined with one analysis algorithm, using the analysis results obtained by executing one analysis algorithm as input data. Specifically, by selecting button 322a labeled "+Process Script" in the lower column 322, it becomes possible to select one or more other scripts that are serially combined with script A and perform a predetermined analysis using the analysis results of script A as input. Similarly, by selecting button 322b labeled "+Process Script" in the lower column 322, it becomes possible to select one or more other scripts that are serially combined with script B and perform a predetermined analysis using the analysis results of script B as input.

[0024] In this embodiment, the script selection receiving means 2 (controller 10) is configured to execute analysis in parallel using multiple scripts selected for the selected image data 410 in each of the multiple analysis steps of the analysis recipe 300. That is, the script selection receiving means 2 (controller 10) selects multiple scripts that can be executed in parallel not only in the analysis processing step (Step 1) but also in the finishing processing step (Step 2) described later. of Controls the selection.

[0025] Furthermore, the script registration means 6 (control unit 10) performs control to add (register) a script to the image analyzing device 100 based on a user operation.

[0026] Furthermore, the property registration means 5 (control unit 10) (see FIG. 1) controls the registration of properties including input / output type, purpose, and extension for each script. Specifically, the property registration means 5 (control unit 10) registers the script properties based on a user operation when the script registration means 6 (control unit 10) registers the script in the image analyzing device 100 (storage unit 30) (at the time of plugging in). The input / output type includes, for example, information regarding the dimensions of the input / output data of the script. The purpose includes, for example, information regarding whether the input / output data of the script is image data or table data. The extension includes information regarding the file format (for example, csv, jpg, txt, etc.) of the input / output data of the script.

[0027] Furthermore, the property registration means 5 (control unit 10) (see FIG. 1) controls the registration of properties including the input / output type, purpose, and extension for each script. Specifically, the property registration means 5 (control unit 10) registers the properties of a script when the script is registered (plugged in) in the image analysis device 100 (storage unit 30). The input / output type includes, for example, information regarding the dimensions of the input / output data of the script. The purpose includes, for example, information regarding whether the input / output data of the script is image data or table data. The extension includes information regarding the file format (for example, csv, jpg, txt, etc.) of the input / output data of the script.

[0028] In this embodiment, the control unit 10 is configured to execute a subsequent script when the properties of the subsequent script match those of the previous script in a series combination. For example, if the output data type of script A is two-dimensional data, the subsequent script is executed when the input data type of the subsequent script combined with script A in series is also two-dimensional data. The above control may be performed based on the purpose or extension rather than the type of input / output data, or may be performed based on two or more of the input / output type, purpose, and extension. If the properties of the subsequent script and the previous script do not match, a warning may be issued when the subsequent script is executed.

[0029] In this embodiment, the control unit 10 is configured to execute a script that matches the properties of the selected one or more subsequent scripts. For example, if two scripts are combined in series with script A and executed in parallel, and only one of the scripts matches the properties of script A, only the one script is executed, and the other script is not executed.

[0030] The image data 410, the selection of which is accepted by the image selection accepting means 1 (controller 10), has the above properties registered therein. A script that receives as input the image data 410, the selection of which is accepted by the image selection accepting means 1, is executed if the properties match those of the image data 410.

[0031] Each of the multiple scripts also includes information regarding whether the analysis results (output data) are subject to finishing processing, which will be described later. Data 440 that is subject to finishing processing among the analysis results of each script in the analysis processing step is temporarily saved (buffered) as a subject to finishing processing and is displayed in column 330. For simplification, data 440 is shown as a white image in FIG. 3.

[0032] In this embodiment, the script includes a first processing algorithm for executing a process of combining at least two or more analysis results from among the analysis results of the multiple scripts selected by the script selection receiving means 2 (controller 10). Specifically, the first processing algorithm is executed to combine the multiple analysis results that have been temporarily saved (buffered) as targets for finishing processing.

[0033] In this embodiment, the script also includes a second processing algorithm for executing a process for analyzing the variation of two or more analysis results. Specifically, the second processing algorithm is executed to analyze the variation of multiple analysis results that are temporarily saved (buffered) as targets for finishing processing.

[0034] In this embodiment, the control unit 10 is configured to execute the first and second processing algorithms in parallel and to accept selection of one or more of the first and second processing algorithms. Specifically, when a button 341a labeled "+Add Script" in the upper column 341 of the "Finishing Processing" column 340 on the setting screen for the analysis recipe 300 is selected, a window is displayed in which the first and second processing algorithms stored in the storage unit 30 can be selected. For example, when both the first and second processing algorithms are selected, the first and second processing algorithms are executed in parallel, and analysis results (data 450, 460) corresponding to each of the first and second processing algorithms are displayed in the lower column 342 of the "Finishing Processing" column. It is also possible to select only one of the first and second processing algorithms. For simplicity, in FIG. 3 , the data 450 and 460 are each depicted as a white image.

[0035] (Analysis flow) Next, the analysis flow by the control unit 10 of the image analyzing device 100 will be described with reference to FIG.

[0036] 5, a user's selection operation of image data 410 is accepted in step 101. Specifically, on the setting screen for the analysis recipe 300 shown in FIGS. 2 and 3, the image data 410 to be subjected to the analysis process can be selected by selecting the button 311a labeled "Select Image" or the button 311b labeled "Register Image."

[0037] Next, in step 102, the selected image data 410 are grouped by condition. Specifically, the control unit 10 performs control to divide the plurality of image data 410 selected in step 101 into a plurality of groups based on conditions such as the passage number of the cells and the number of days of culture. Each of the plurality of groups includes one or more image data.

[0038] Next, in step 103, the control unit 10 performs control to list up the image data 410 of one group among the plurality of groups formed in step 102, for which analysis processing has not yet been performed.

[0039] Next, in step 104, the control unit 10 performs control to read one of the unanalyzed image data 410 listed in step 103.

[0040] Next, in step 105, the data analysis means 3 (control unit 10) controls the execution of processing according to the analysis recipe 300 for the single image data 410 read in step 104. An example of the analysis recipe 300 will be described later with reference to Fig. 6. Note that in step 105, only the analysis processing step (Step 1, see Fig. 3) of the analysis recipe 300 is performed.

[0041] Next, in step 106, the control unit 10 determines whether or not the analysis result of each of the multiple scripts acquired in step 105 is a target for finishing processing. Note that the control unit 10 determines whether or not the analysis result is a target for finishing processing each time each of the multiple scripts is executed in step 105. That is, the control unit 10 determines whether or not the analysis result is a target for finishing processing while executing processing according to the analysis recipe 300. Analysis results determined to be a target for finishing processing are processed in step 107. Analysis results determined not to be a target for finishing processing are processed in step 108. Note that after all of the multiple scripts in the analysis recipe 300 have been executed, it may be determined whether or not each of all analysis results is a target for finishing processing.

[0042] In step 107, the analysis results determined in step 106 to be subject to finishing processing are temporarily buffered in a cache memory, etc. In step 108, the analysis results determined in step 106 not to be subject to finishing processing are saved (registered) in the storage unit 30.

[0043] Next, in step 109, it is determined whether or not processing has been completed for all of the image data 410 listed in step 103 according to the analysis recipe 300. If it is determined that processing has been completed for all of the image data 410 listed according to the analysis recipe 300, the process proceeds to step 110. If it is determined that processing has not been completed for all of the image data 410 listed according to the analysis recipe 300, the process returns to step 104.

[0044] Next, in step 110, it is determined whether or not processing has been completed according to the analysis recipes 300 for all groups grouped by condition in step 102. If it is determined that processing has been completed according to the analysis recipes 300 for all groups, the process proceeds to step 111. If it is determined that processing has not been completed according to the analysis recipes 300 for all groups, the process returns to step 103.

[0045] Next, in step 111, finishing processing is performed on the data designated for finishing processing (buffered in step 107). That is, the finishing processing step (Step 2, see FIG. 3) in the analysis recipe 300 is performed. An example of the finishing processing (first processing algorithm and second processing algorithm) will be described later.

[0046] Then, in step 112, the analysis result of the finishing process acquired in step 111 is saved (registered) in the storage unit 30.

[0047] (Analysis recipe) Next, an example of the analysis recipe 300 will be described with reference to Fig. 6. Note that steps 1051 to 1057 in Fig. 6 are included in step 105 in Fig. 5. Also, steps 1111 to 1112 in Fig. 6 are included in step 111 in Fig. 5.

[0048] First, as shown in FIG. 6, the control unit 10 applies a trained model that has learned the characteristics of in-focus and out-of-focus images to the input image data 410. As a result, the control unit 10 performs control to display in-focus images of cells 400 and out-of-focus images of cells 400 in the input image data 410, distinguishing them, for example, by color. Note that in-focus images of cells 400 and out-of-focus images of cells 400 may be distinguished by a method different from the above. Furthermore, image data in which a subject other than cells 400 is displayed may also be used.

[0049] Next, in step 1052, a focus coincidence evaluation algorithm is executed using the analysis result (output data) of step 1051 as input data, thereby calculating the ratio of the area in which the cells 400 are displayed that is in focus to the entire area in which the cells 400 are displayed in the input data.

[0050] Next, in step 1053, a trained model that has trained cell regions and background regions is applied using the analysis results (output data) of step 1052 as input data, thereby distinguishing cell regions from background regions in the input data.

[0051] Next, in step 1054, a binarization algorithm is applied to the analysis results (output data) of step 1053 as input data, thereby obtaining data in which the values ​​of the cell region and the background region in the input data are set to 1 and 0, respectively.

[0052] Next, a particle size analysis algorithm and an area fraction analysis algorithm are selected as analysis algorithms that use the analysis results (output data) of step 1054 as input data. As a result, in steps 1055 and 1056, the particle size analysis algorithm (step 1055) and the area fraction analysis algorithm of cells 400 (step 1056) are executed in parallel, using the analysis results (output data) of step 1054 as input data. By executing the particle size analysis algorithm of step 1055, the particle size of the cells 400 displayed in the input data is calculated. By executing the area fraction analysis algorithm of cells 400 of step 1056, the area fraction of the cells 400 displayed in the input data (for example, the area fraction occupied by each cell 400 with respect to the total area of ​​the region in which the cells 400 are displayed) is calculated.

[0053] Next, in step 1057, a frequency distribution graph creation algorithm is executed using the analysis processing (output data) of step 1055 as input data, thereby creating a graph showing the number of cells 400 for each particle diameter.

[0054] Next, in step 1111, a first processing algorithm for graph creation is executed using the analysis results (output data) of step 1056 and the analysis results (output data) of step 1057 as input data. Specifically, assuming that a predetermined graph is created by the processing of step 1056, the graph acquired by the processing of step 1056 and the graph acquired by the processing of step 1057 are integrated (combined into one graph) by executing the first processing algorithm. Furthermore, graphs corresponding to mutually different conditions or graphs of mutually different image data 410 may be integrated by executing the first processing algorithm.

[0055] Furthermore, in step 1112, a second processing algorithm is executed to analyze variations between data corresponding to different conditions or between different pieces of image data 410. Note that the processing contents of steps 1111 and 1112 described above are merely examples and are not limited to these.

[0056] The analysis process of step 1111 and the analysis process of step 1112 are executed in parallel.

[0057] (Effects of this embodiment) The image analyzing device 100 of this embodiment can provide the following effects.

[0058] In this embodiment, as described above, the image analyzing device 100 includes a control unit 10 that controls the execution of analysis in parallel using a plurality of selected scripts for the selected image data 410. This eliminates the need to individually execute processing for each of the plurality of scripts for one piece of data, thereby reducing the user's workload. Therefore, when executing a plurality of scripts, the user's workload can be reduced. Furthermore, compared to the case where processing is executed individually for each of the plurality of scripts, the work time can be shortened. Furthermore, when the plurality of selected scripts are executed, the user's workload can be reduced. of By displaying the analysis results of each script on the same screen, the user can compare the analysis results of each of the selected scripts on the same screen.

[0059] Furthermore, in this embodiment, further effects can be obtained by configuring as follows.

[0060] In this embodiment, as described above, the control unit 10 is configured to perform control to generate an analysis recipe 300 in which a plurality of analysis steps for analyzing image data 410 are combined in series, and to execute analysis in parallel using a plurality of scripts selected for selected image data 410 in each of the plurality of analysis steps in the generated analysis recipe 300. This allows the scripts to be executed in parallel in each of the plurality of analysis steps, thereby further reducing the user's effort.

[0061] Furthermore, in this embodiment, as described above, the script includes a trained model generated by machine learning for analyzing the image data 410, and an analysis algorithm used to analyze the image data 410. The control unit 10 is also configured to perform analysis in parallel using the selected trained models for the selected image data 410, and to perform analysis in parallel using the selected analysis algorithms for the selected image data 410. This allows the execution of multiple trained models in parallel and multiple analysis algorithms in parallel, thereby reducing the user's effort in each of the analysis using the trained model and the analysis using the analysis algorithm.

[0062] Furthermore, in this embodiment, as described above, the control unit 10 is configured to execute another analysis algorithm that can be serially combined with one analysis algorithm, using the analysis results obtained by executing one analysis algorithm as input data. This allows the analysis recipe 300 to be easily expanded by serially combining analysis algorithms. Furthermore, unlike when one algorithm and another algorithm are written as a single script, the one algorithm and the other algorithms can be used individually. As a result, the variations of the analysis recipe 300 and the variations of the processing modules consisting of multiple algorithms can be easily increased.

[0063] Furthermore, unlike when one algorithm and other algorithms are written as a single script, by writing each algorithm as a separate script, the analysis results of each algorithm can be obtained and confirmed separately. As a result, if an unexpected analysis result is obtained, each algorithm can be individually corrected. Furthermore, it is easy to analyze the variation in data based on the analysis results of each algorithm and to consolidate the data.

[0064] Furthermore, compared to when one algorithm and another algorithm are written as one script, the amount of script written is reduced, which reduces the possibility that the configuration will be partially similar to other scripts. As a result, it is possible to reduce the possibility that other similar scripts will be searched for, making it easier to search for scripts.

[0065] Furthermore, in this embodiment, as described above, the control unit 10 is configured to control the registration of properties, including input / output type, purpose, and extension, for each script, and to execute a later script of two serially combined scripts when the properties of the later script match the properties of the earlier script of the two scripts. This prevents the execution of a later script with properties different from those of the earlier script, thereby preventing the generation of inconsistent analysis results due to the later script. Furthermore, the user can select a later script based on the properties of the earlier script, making it easy to select the later script.

[0066] Furthermore, in this embodiment, as described above, the control unit 10 is configured to execute a script that matches the properties of the previous script among the one or more selected subsequent scripts. As a result, a script that does not match the properties of the previous script among the one or more selected subsequent scripts is not executed, thereby preventing an increase in the control load of the control unit 10 due to the execution of a script that does not match the properties of the previous script.

[0067] Furthermore, in this embodiment, as described above, the script includes a first processing algorithm for executing a process of combining at least two or more analysis results from the analysis results of the selected multiple scripts. This allows the first processing algorithm to generate a single analysis result based on the two or more analysis results. As a result, the user can easily understand the correlations between the two or more analysis results based on the combined single analysis result.

[0068] In this embodiment, as described above, the script further includes a second processing algorithm for executing a process for analyzing the variation of two or more analysis results, which allows the user to easily understand the correlation between the two or more analysis results based on the processing results of the second processing algorithm.

[0069] Furthermore, in this embodiment, as described above, the control unit 10 is configured to be able to execute the first processing algorithm and the second processing algorithm in parallel. This allows the first processing algorithm and the second processing algorithm to be executed in parallel, which reduces the user's workload compared to when the process of combining two or more analysis results and the analysis of variations in two or more analysis results are performed separately.

[0070] Furthermore, in this embodiment, as described above, the data analysis method includes a step of executing analysis in parallel using multiple selected scripts on selected image data 410. This allows multiple scripts to be executed in parallel on one piece of data. As a result, the task of individually executing processing for each of multiple scripts on one piece of data is no longer necessary, and an image analysis method that can reduce the user's workload can be provided.

[0071] [Variations] The embodiments disclosed herein should be considered to be illustrative and not restrictive in all respects. The scope of the present invention is defined by the claims rather than the above description of the embodiments, and further includes all modifications (variations) within the meaning and scope of the claims.

[0072] For example, in the above embodiment, an example of the analysis recipe 300 in which multiple scripts are combined in series and in parallel is shown, but the present invention is not limited to this. For example, an analysis recipe 500 (see FIG. 7) in which multiple scripts are combined only in series may be generated, or an analysis recipe 600 (see FIG. 8) in which multiple scripts are combined only in parallel may be generated.

[0073] Although the present embodiment has described an example in which analysis algorithms are combined in series and in parallel, the present invention is not limited to this. Analysis modules each consisting of a plurality of analysis algorithms may be combined in series and in parallel, or the modules and analysis algorithms (trained models) may be combined in series and in parallel.

[0074] In the above embodiment, an example was shown in which the analysis recipe 300 includes both a trained model and an analysis algorithm, but the present invention is not limited to this. For example, the analysis recipe may not include either the trained model or the analysis algorithm. Furthermore, the analysis recipe may not include a finishing process.

[0075] In addition, although the above embodiment shows an example in which an analysis processing step is provided in which both a trained model and an analysis algorithm can be selected, the present invention is not limited to this. A step in which only a trained model can be selected and a step in which only an analysis algorithm can be selected may be combined in series.

[0076] In the above embodiment, an example was shown in which the script included a first processing algorithm and a second processing algorithm as algorithms for finishing processing, but the present invention is not limited to this. The script may include, as algorithms for finishing processing, processing algorithms that perform processing on two or more processing results in addition to the first processing algorithm and the second processing algorithm.

[0077] In the above embodiment, the image analysis device 100 that analyzes the image data 410 is shown as an example of a data analysis device, but the present invention is not limited to this. The data analysis device may be an analysis device that analyzes analytical data other than image data (for example, analysis results of gas chromatography, etc.).

[0078] (Item 1) A control that accepts an operation to select analysis data acquired by the analysis device; A control that accepts an operation to select a plurality of scripts that will perform analysis on the selected analysis data; a control unit that controls execution of analyses in parallel using the selected plurality of scripts for the selected analysis data; a display unit that displays, on the same screen, analysis results obtained by the control unit analyzing the analysis data.

[0079] (Item 2) Item 1. The data analysis device according to item 1, wherein the control unit controls the generation of an analysis flow in which a plurality of analysis steps for analyzing the analysis data are combined in series, and is configured to execute analysis in parallel using a plurality of selected scripts for the selected analysis data in each of the plurality of analysis steps of the generated analysis flow.

[0080] (Item 3) The script includes a trained model generated by machine learning for analysis of the analytical data, and an analytical algorithm used to perform analysis of the analytical data; 3. The data analysis device according to item 1 or 2, wherein the control unit is configured to perform analysis in parallel using the selected plurality of trained models for the selected analytical data, and to perform analysis in parallel using the selected plurality of analytical algorithms for the selected analytical data.

[0081] (Item 4) The script includes a trained model generated by machine learning for analysis of the analytical data, and an analytical algorithm used to perform analysis of the analytical data; 4. The data analysis device according to any one of items 1 to 3, wherein the control unit is configured to execute another analysis algorithm that is combined in series with one of the analysis algorithms, using the analysis result obtained by executing one of the analysis algorithms as input data.

[0082] (Item 5) The data analysis device according to any one of items 1 to 4, wherein the control unit controls the registration of properties including input / output type, purpose, and extension for each of the scripts, and is configured to execute a later script of two serially combined scripts when the properties of the later script match the properties of the earlier script of the two scripts.

[0083] (Item 6) 6. The data analysis device according to item 5, wherein the control unit is configured to execute, from among the one or more selected subsequent scripts, a script that matches the properties of the previous script.

[0084] (Item 7) The data analysis device according to any one of items 1 to 6, wherein the script includes a first processing algorithm for executing a process of summarizing at least two or more of the analysis results of each of the selected plurality of scripts.

[0085] (Item 8) 8. The data analysis device according to item 7, wherein the script further includes a second processing algorithm for executing a process for analyzing variability in two or more of the analysis results.

[0086] (Item 9) Item 9. The data analysis device according to item 8, wherein the control unit is configured to be able to execute the first processing algorithm and the second processing algorithm in parallel.

[0087] (Item 10) accepting a selection of analytical data acquired by the analytical device; selecting a plurality of scripts for performing analysis on the analysis data; and executing analysis in parallel using the selected plurality of scripts on the selected analysis data. [Explanation of symbols]

[0088] 10 Control Unit 20 Display section 100 Image analysis device (data analysis device) 200 Image acquisition unit (analysis device) 300, 301, 302, 500, 600 analysis recipe (analysis flow) 410 Image data (analysis data)

Claims

1. A control that accepts an operation to select analysis data acquired by the analysis device; A control that accepts an operation to select a plurality of scripts that will perform analysis on the selected analysis data; a control unit that controls execution of analyses in parallel using the selected plurality of scripts for the selected analysis data; a display unit that displays, on the same screen, an analysis result obtained by the analysis of the analysis data by the control unit; The control unit of the data analysis device performs further analysis using the analysis results of at least one of the analysis data that has been analyzed in parallel by multiple scripts as input data, and displays the analysis results on the same screen of the display unit.

2. 2. The data analysis device according to claim 1, wherein the control unit controls the generation of an analysis flow in which a plurality of analysis steps for analyzing the analysis data are combined in series, and is configured to perform analysis in parallel using a plurality of selected scripts for the selected analysis data in each of the plurality of analysis steps of the generated analysis flow.

3. The script includes a trained model generated by machine learning for analysis of the analytical data, and an analytical algorithm used to perform analysis of the analytical data; 3. The data analysis device according to claim 1, wherein the control unit is configured to perform analysis in parallel using the selected plurality of trained models for the selected analytical data, and to perform analysis in parallel using the selected plurality of analytical algorithms for the selected analytical data.

4. The script includes a trained model generated by machine learning for analysis of the analytical data, and an analytical algorithm used to perform analysis of the analytical data; The data analysis device according to any one of claims 1 to 3, wherein the control unit is configured to execute another analysis algorithm that can be combined in series with one of the analysis algorithms, using the analysis result obtained by executing one of the analysis algorithms as input data.

5. The data analysis device according to any one of claims 1 to 4, wherein the control unit controls the registration of properties including input / output type, purpose, and extension for each of the scripts, and is configured to execute a later script of two serially combined scripts when the properties of the later script match the properties of the earlier script of the two scripts.

6. The data analysis apparatus according to claim 5 , wherein the control unit is configured to execute, from among the one or more selected subsequent scripts, a script that matches the properties of the previous script.

7. The data analysis device according to any one of claims 1 to 6, wherein the script includes a first processing algorithm for executing a process of consolidating at least two or more of the analysis results of each of the selected plurality of scripts.

8. The data analysis device according to claim 7 , wherein the script further includes a second processing algorithm for executing a process for analyzing variability between two or more of the analysis results.

9. The data analysis device according to claim 8 , wherein the control unit is configured to be able to execute the first processing algorithm and the second processing algorithm in parallel.

10. accepting a selection of analytical data acquired by the analytical device; selecting a plurality of scripts for performing analysis on the analysis data; executing an analysis in parallel using a plurality of the selected scripts on the selected analysis data; A data analysis method comprising a step of displaying on the same screen the analysis results of parallel analysis by the plurality of scripts and the analysis results of a further analysis performed using as input data the analysis results of at least one of the analytical data analyzed in parallel by the plurality of scripts.

Images