Method for generating quality control information, apparatus for generating quality control information, and program

Abstract

To provide a method, a device, and a program to generate accuracy management information for smear samples that corresponds to the fact that smear and staining conditions vary by testing facility or region.SOLUTION: A method to generate accuracy management information is the method to generate the accuracy management information for smear samples, comprising: acquiring image data from each of multiple smear samples (S1); obtaining feature values that reflect the staining state of the smear sample from each of the multiple image data (S2); and generating accuracy management information based on the feature values (S3).SELECTED DRAWING: Figure 1

Description

Technical Field

[0001] The present invention relates to a method for generating accuracy management information, an apparatus for generating accuracy management information, and a program.

Background Art

[0002] A method for performing accuracy management of a smear preparation apparatus based on the staining state of a smear specimen is known. For example, in Patent Document 1, the luminance value of a specific color component in the nuclear region in a blood cell image of a smear specimen is used as a characteristic value reflecting the staining state of blood cells in the smear specimen and compared with a predetermined lower limit reference value and a predetermined upper limit reference value. When the characteristic value is less than or equal to the predetermined lower limit reference value or greater than or equal to the predetermined upper limit reference value, a system for notifying the occurrence of abnormal staining is described.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] The smear and staining treatment conditions of a smear specimen include the reagents, apparatuses, and conditions of each step (e.g., smear conditions, pH of the staining solution, temperature of the staining solution, and staining time) used for preparing the smear specimen, and affect the staining state of the smear specimen. The smear and staining treatment conditions of a smear specimen may differ for each inspection facility or for each region or country (hereinafter referred to as "region, etc.").

[0005] In the accuracy management method described in Patent Document 1, a characteristic value reflecting the staining state of blood cells in a smear specimen is compared with a predetermined lower limit reference value and a predetermined upper limit reference value to notify the occurrence of abnormal staining. However, Patent Document 1 does not consider that the smear and staining conditions differ for each inspection facility or for each region, etc.

[0006] The present invention aims to provide a method for generating quality control information for smear specimens, an apparatus for generating quality control information, and a program that can accommodate differences in smear and staining conditions between testing facilities or regions. [Means for solving the problem]

[0007] As shown in Figure 5, the method for generating quality control information of the present invention is a method for generating quality control information for smear specimens, and is characterized by comprising: acquiring image data from each of a plurality of smear specimens (S1); acquiring feature values ​​that reflect the staining state of the smear specimens from each of the plurality of image data (S2); and generating quality control information based on the feature values ​​(S3).

[0008] According to the method for generating quality control information of the present invention, feature values ​​reflecting the staining state of the smear specimen are obtained from each of the acquired multiple image data, and quality control information is generated based on these feature values. Therefore, the method for generating quality control information of the present invention makes it possible to generate quality control information for smear specimens that can be adapted to differences in smear and staining conditions between testing facilities or regions.

[0009] As shown in Figures 1 and 4, the quality control information generation device (80) of the present invention is a system for generating quality control information for smear specimens, and includes a control unit (50), the control unit (50) acquires image data from each of a plurality of smear specimens, acquires feature values ​​that reflect the staining state of the smear specimen from each of the plurality of image data, and generates quality control information based on the feature values.

[0010] According to the quality control information generation device (80) of the present invention, the control unit (50) of the generation device (80) acquires feature values ​​that reflect the staining state of the smear specimen from each of the acquired multiple image data, and generates quality control information based on the feature values. Therefore, the quality control information generation device of the present invention can generate quality control information for smear specimens that corresponds to differences in smear and staining conditions for each testing facility or region.

[0011] As shown in Figures 1 and 4, the present invention is a program that causes a device for generating quality control information for smear specimens to perform the following actions: acquire image data from each of a plurality of smear specimens; acquire feature values ​​that reflect the staining state of the smear specimens from each of the plurality of image data; and generate quality control information based on the feature values.

[0012] According to the program of the present invention, feature values ​​reflecting the staining state of the smear specimen are obtained from each of the acquired multiple image data, and quality control information is generated based on these feature values. Therefore, the program of the present invention can generate quality control information for smear specimens that can accommodate differences in smear and staining conditions between testing facilities or regions. [Effects of the Invention]

[0013] According to the present invention, it is possible to generate quality control information for smear specimens that can accommodate differences in smear and staining conditions between testing facilities or regions. [Brief explanation of the drawing]

[0014] [Figure 1] Figure 1 is a schematic diagram illustrating the overview of the generation system. [Figure 2] Figure 2(A) shows an overview of the process for generating quality control information. Figure 2(B) shows an overview of the process for acquiring feature values ​​of the smear samples being managed. [Figure 3] Figure 3 shows an overview of the output screen during the process of generating quality control information and acquiring characteristic values ​​of the smear specimens under management. [Figure 4] Figure 4 is a block diagram showing the configuration of the generation apparatus. [Figure 5] Figure 5 is a flowchart showing the process of generating quality control information and the output process of feature values ​​that reflect the staining status of the smear specimens under management, as well as the generated quality control information. [Figure 6] Figure 6 is a flowchart showing the feature value acquisition process. [Figure 7]FIG. 7 is a diagram showing details of an image data acquisition process and a feature value acquisition process in a precision management information generation process. [Figure 8] FIG. 8 is a diagram showing an output screen of precision management information and feature values reflecting the staining state of a smear specimen to be managed. [Figure 9] FIG. 9 is a schematic diagram showing an overview of the generation system. [Figure 10] FIG. 10 is a diagram showing an output screen of precision management information and feature values reflecting the staining state of a smear specimen to be managed. [Figure 11] FIG. 11 is a diagram showing an output screen of precision management information and feature values reflecting the staining state of a smear specimen to be managed. [Figure 12] FIG. 12 is a diagram showing feature values reflecting the staining state of red blood cells. [Figure 13] FIG. 13 is a diagram showing an acquisition process of feature values related to red blood cells. [Figure 14] FIG. 14 is a diagram showing feature values reflecting the staining state of white blood cells. [Figure 15] FIG. 15 is a diagram showing feature values reflecting the morphology of blood cells. [Figure 16] FIG. 16 is a flowchart showing an acquisition process of a granulation index. [Figure 17] FIG. 17 is a diagram showing an acquisition process of a granulation index. [Figure 18] FIG. 18 is a diagram showing an extraction process of primary granules and secondary granules. [Figure 19] FIG. 19 is a diagram showing an output screen of precision management information and feature values reflecting the staining state of a smear specimen to be managed. [Figure 20] FIG. 20 is a diagram showing an output screen of precision management information and feature values reflecting the staining state of a smear specimen to be managed. [Figure 21] FIG. 21 is a diagram showing a screen for displaying a plurality of image data acquired from a smear specimen. [Figure 22] FIG. 22 is a diagram showing a screen for associating and displaying specific image data and a plurality of feature values associated with the specific image data. [Figure 23] Figure 23 shows a screen where you can select the feature values ​​to acquire. [Figure 24] Figure 24 shows a screen where a normal range can be set for any feature value. [Modes for carrying out the invention]

[0015] Preferred embodiments of the present invention will be described below with reference to the drawings. The same elements are denoted by the same reference numerals, and redundant descriptions are omitted. Furthermore, unless otherwise specified, positional relationships such as up, down, left, and right are based on those shown in the drawings. In addition, the dimensional ratios in the drawings are not limited to those shown. Moreover, the following embodiments are illustrative examples for explaining the present invention, and the present invention is not limited to these embodiments.

[0016] [Overview of the generation system] Referring to Figure 1, an overview of the generation system 100, which generates quality control information for blood smear specimens collected from subjects, will be described. Figure 1 is a schematic diagram showing an overview of the generation system 100.

[0017] The quality control information generation system 100 comprises an inspection system 70-1 including a smear preparation device 20, a specimen transport device 30, and a specimen image acquisition device 40, and a quality control information generation device 80.

[0018] The inspection system 70-1 is a system in which a smear slide 10 is prepared by a smear preparation device 20, the prepared smear slide 10 is transported to a specimen image acquisition device 40 by a specimen transport device 30, and the prepared smear slide 10 is imaged by the specimen image acquisition device 40. The inspection system 70-1 is installed, for example, in a single inspection facility (inspection facility A). The generation device 80 is installed in the facility of the provider of the specimen image acquisition device 40 (for example, the manufacturer of the specimen image acquisition device 40) and is connected to the smear preparation device 20, the specimen transport device 30, and the specimen image acquisition device 40 of the inspection system 70-1 via a network. The generation device 80 acquires image data of the imaged smear slide 10, acquires feature values ​​that reflect the staining state of the smear from the image data, and generates quality control information for the smear based on the acquired feature values.

[0019] Feature values ​​are numerical information that reflects the staining state of a blood smear in image data that reflects the regions of blood cells in the blood. Furthermore, feature values ​​include, for example, the color information of red blood cells. They are obtained by acquiring color indices for the intracellular regions of each cell from multiple image data sets, and quantitatively quantifying the staining state in each smear from those multiple image data sets.

[0020] Quality control information consists of the median and mean values ​​statistically calculated from multiple feature values. Furthermore, the quality control information includes at least one of an upper limit and a lower limit, for example, ±2SD (Standard Deviation) or ±3SD relative to the median and mean. In addition to the median and mean, the quality control information may also use values ​​based on, for example, a moving average of multiple feature values.

[0021] By generating quality control information based on the characteristic values ​​of multiple smear samples, it is possible to compare the characteristic values ​​of the smear sample under management with the generated quality control information to determine whether or not there are problems with the staining state of the smear sample under management. In such cases, it is possible to ensure the quality of the smear sample under management based on the quality control information.

[0022] The timing for generating quality control information is arbitrary; for example, it could be daily, every few days, weekly, or monthly, or it could be when the lot of the staining solution used in preparing the smear specimens is changed. Similarly, the timing for acquiring feature values ​​that reflect the staining status of the smear specimens under management is also arbitrary; for example, it could be daily, every few days, weekly, monthly, or every few hours.

[0023] Figures 2(A) and (B) show an overview of the process for generating quality control information and acquiring characteristic values ​​of the smear samples being managed.

[0024] As shown in Figure 2(A), the generation device 80 shown in Figure 1 acquires image data from each of several blood smears (e.g., Sample 1…Sample N) that reflects each region of blood cells in the blood, and acquires feature values ​​X1…XN that reflect the staining state of the smear from each of the multiple image data. The generation device 80 generates, for example, a target value ACI and / or a control range MW based on an upper and lower limit as quality control information. The target value ACI is calculated from the average value of the acquired feature values ​​X1…XN, and the upper and lower limits of the control range MW are calculated from ±2SD (Standard Deviation) of the target value ACI. The quality control information is generated, for example, on a daily basis from each image data of all smears imaged on that day.

[0025] As shown in Figure 2(B), the generating device 80 acquires image data from the managed smear (e.g., Sample) that reflects the region of blood cells in the blood, as part of the process of acquiring the feature value Y of the managed smear, and acquires the feature value Y that reflects the staining state of the managed smear from the acquired image data. The generating device 80 outputs the feature value Y of the managed smear to the quality control information, such as the target value ACI and the control width MW, thereby enabling the user to perform quality control on the managed smear. The feature value that reflects the staining state of the managed smear is acquired, for example, from the image data of at least one smear out of all the smears imaged on that day.

[0026] Figure 3 shows an overview of the output screen in the quality control information generation process and the acquisition of feature values ​​Y of the managed smear specimens shown in Figure 2. The example shown in Figure 3 outputs the generated quality control information and the feature values ​​Y that reflect the staining state of the acquired managed smear specimens. For example, the target value ACI and control width MW, which are quality control information, are output. In addition, a plot P corresponding to the feature value Y that reflects the staining state of the managed smear specimens is output. By visually understanding the relationship between the target value ACI and control width MW, which are quality control information, and the output plot P, the user can easily perform quality control on the managed smear specimens. Here, the horizontal axis ("Sample") in the output graph indicates the managed smear specimen, for example, that Sample 2 is a different smear specimen from Sample 4. The vertical axis shows the value of the feature value that reflects the staining state of the managed smear specimens.

[0027] The subjects are primarily humans, but may also be other animals. The testing system 100 performs clinical testing or analysis for medical research on specimens collected from patients, for example. The specimens are of biological origin. Biological specimens include, for example, liquids such as blood (whole blood, serum, or plasma), urine, or other bodily fluids collected from the subject, or liquids obtained by subjecting collected bodily fluids or blood to a predetermined pretreatment. The specimens may also be other than liquids, such as a part of the subject's tissue or cells.

[0028] The smear preparation device 20 is a device for performing a smearing process to spread the sample onto a slide, and then staining the smear slide 10 on which the sample has been smeared. The smear preparation device 20 prepares the smear slide 10 by aspirating the sample, dropping and smearing it onto the slide, and then staining it.

[0029] The specimen transport device 30 receives the smear slide 10 prepared by the smear preparation device 20 and transports it to the specimen image acquisition device 40. The specimen transport device 30 also receives the smear slide 10 after it has been imaged by the specimen image acquisition device 40 and stores it.

[0030] The specimen image acquisition device 40 captures an image of the smear specimen slide 10 that has been transported by the specimen transport device 30.

[0031] The testing system 70-1 may also include other devices. For example, the testing system 70 may include an analytical device for analyzing the sample (e.g., a hematology analyzer for classifying and counting blood cells in the sample) and a transport device for transporting containers containing the sample.

[0032] Furthermore, an inspection system 70-1 including a smear preparation device 20, a specimen transport device 30, and a specimen image acquisition device 40 is described, for example, in U.S. Patent Application Publication No. 2019 / 0049474. U.S. Patent Application Publication No. 2019 / 0049474 is incorporated herein by reference.

[0033] Figure 4 is a block diagram showing an example of the configuration of the generation device 80. As shown in Figure 4, the generation device 80 includes a control unit 50, a storage unit 71, an input unit 72, a display unit 73, and a communication unit 74.

[0034] The control unit 50 includes, for example, a CPU that performs information processing to generate quality control information for smear specimens. The control unit 50 can also communicate with the smear specimen preparation device 20, specimen transport device 30, and specimen image acquisition device 40 shown in Figure 1 via the communication unit 74. The storage unit 71 includes, for example, a memory that records information for executing the information processing of the control unit 50 and information generated by executing said information processing. The input unit 72 is, for example, a keyboard or mouse, and the display unit 73 is, for example, a liquid crystal display or an organic EL display.

[0035] Figure 5 is a flowchart showing the process of generating quality control information and the output process of feature values ​​that reflect the staining status of the smear specimens under management, as well as the generated quality control information.

[0036] As shown in the quality control information generation process in Figure 5, the control unit 50 of the generation device 80 shown in Figure 4 acquires image data from each of the multiple smear specimens (step S1). The control unit 50 acquires feature values ​​that reflect the staining state of the smear specimen from each of the multiple image data (step S2). The control unit 50 generates quality control information for the smear specimens based on the acquired feature values ​​(step S3). The control unit 50 generates quality control information by executing steps S1 to S3.

[0037] As shown in Figure 5, the output process of feature values ​​reflecting the staining state of the smear specimen under management and the generated quality control information, the control unit 50 acquires multiple image data of the under management (step S4). The control unit 50 acquires feature values ​​from the multiple image data of the under management (step S5). The control unit 50 outputs feature values ​​reflecting the staining state of the smear specimen under management and the generated quality control information on the display unit 73 (step S6). By executing steps S4 to S6, the control unit 50 outputs feature values ​​reflecting the staining state of the smear specimen under management and the generated quality control information.

[0038] Figure 6 is a flowchart showing the feature value acquisition process in step S2 of Figure 5. As shown in Figure 6, the control unit 50 shown in Figure 4 recognizes each cellular component based on the acquired multiple image data (step S11). Next, based on the recognition results in step S11, the control unit 50 identifies and extracts each region of each cell, including the nucleus and cytoplasm (step S12). For each of the multiple image data, the control unit 50 acquires the color information of the blood cells in each extracted region (feature values ​​that reflect the staining state of the smear) (step S13).

[0039] Figure 7 shows the details of the image data acquisition process and the feature value acquisition process in the quality control information generation process. As shown in Figure 7, the control unit 50 acquires, for example, several hundred or several thousand image data IDs from each of several smear specimens. More specifically, the control unit 50 acquires several image data IDs acquired by the specimen image acquisition device 40 of the inspection system 70-1 shown in Figure 1.

[0040] The control unit 50 shown in Figure 4 acquires feature values ​​that reflect the staining state of the smear specimen from each of the multiple image data IDs. As shown in Figure 7, the control unit 50 performs the following as a feature value acquisition process, corresponding to the flowchart showing the feature value acquisition process in Figure 6: (1) recognition of cellular components, (2) extraction of the nuclear and cytoplasmic regions, and (3) acquisition of blood cell color information for each image data.

[0041] More specifically, the control unit 50, as process (1), recognizes each cellular component based on the acquired multiple image data IDs. Next, as process (2), the control unit 50 identifies and extracts each region of each cell, including, for example, the nucleus and cytoplasm, based on the recognition results of process (1). As process (3), the control unit 50 associates information on the cell type (red blood cells, white blood cells, platelets, etc.) and information on structural components (nucleus, cytoplasm, granules, etc.) in each extracted region for each of the multiple image data IDs, and acquires color information of blood cells (feature values ​​that reflect the staining state of the smear).

[0042] The feature values ​​that reflect the staining state of the smear include the color information obtained from each region of the blood cells in the image data, as described above. The "feature values ​​that reflect the staining state of the smear" include, for example, the luminance, hue, saturation, and value values ​​of the color components (e.g., Red, Green, Blue) obtained from the image data of the blood cells, as well as at least one value of a combination thereof (e.g., HSV value, RGB).

[0043] "Blood cells" include, for example, red blood cells, white blood cells, and at least one of platelets. White blood cells include, for example, basophils, eosinophils, neutrophils, monocytes, and at least one of lymphocytes.

[0044] Red blood cells, which make up about 90% of the blood cells in the blood, occupy a large proportion (area) of a blood smear. Therefore, the color information of red blood cells can reflect the overall staining state of the smear, making it a more preferable characteristic value for reflecting the staining state of the smear. Compared to the color information of white blood cells, the color information of red blood cells tends to reflect changes in staining state due to differences in staining conditions such as pH.

[0045] On the other hand, leukocyte color information may be used as a feature value that reflects the staining state of the smear specimen. In this case, quality control may be performed using multiple types of feature values ​​that reflect the staining state of the smear specimen, such as red blood cells and leukocytes. For example, by using at least one erythrocyte color information and at least one leukocyte color information, it is possible to manage the overall staining quality of the smear specimen while simultaneously managing the staining properties of the leukocyte components (e.g., nuclei or granules), which are important targets of analysis.

[0046] Below, with reference to Figures 8 to 11, we will explain examples of output screens that show quality control information and feature values ​​reflecting the staining status of the smear specimens under management.

[0047] Figure 8 shows an example of an output screen for internal quality control, displaying quality control information and feature values ​​that reflect the staining status of the smear specimens under control. Internal quality control compares quality control information based on the feature values ​​of a single testing facility (e.g., control width MW) with feature values ​​that reflect the staining status of the smear specimens under control from among multiple smear specimens prepared.

[0048] For example, if a specific feature value reflecting the staining state of a controlled smear falls outside the control range (MW), it indicates a problem with the staining state of the controlled smear. Therefore, the quality of the controlled smear can be ensured through internal quality control as described above.

[0049] The generator 80 shown in Figure 1 acquires multiple feature values ​​from each of multiple smear specimens prepared at a single testing facility and outputs quality control information determined based on these multiple feature values. As shown in Figure 8, the generator 80 outputs plots P of multiple feature values ​​related to the color information of red blood cells and plots P of multiple feature values ​​related to the color information of white blood cells for each feature value index I (e.g., RBC Redness Index, RBC Color S, Granule Index), along with the control width MW. This configuration makes it possible to grasp the staining state of the entire smear specimen using the color information of red blood cells, while also performing quality control on the staining properties and morphology of white blood cells, which are important targets for analysis. Note that instead of the control width MW, at least one of the upper and lower limits may be output as a solid or dotted line.

[0050] The RBC Redness Index, which is a feature value index I, is calculated by performing principal component analysis on the hue value, saturation value, and brightness value of red blood cells, and then taking the average value of each value obtained from the principal component analysis. The RBC Color S, which is a feature value index I, is the saturation value of red blood cells. The Granule Index, which is a feature value index I, is the granule index of white blood cells. The generator 80 may output only the plot P of multiple feature values ​​related to the color information of red blood cells to the output screen, or it may output only the plot P of multiple feature values ​​related to the color information of white blood cells to the output screen.

[0051] Quality control information used for internal quality control refers to information indicating the variability in the staining properties (staining state) of smear specimens within a single testing facility. This quality control information can serve as a standard for quality control according to the staining conditions of that facility. Therefore, it becomes possible to manage the quality of smear specimens prepared daily at that facility according to standards appropriate for that facility.

[0052] Furthermore, quality control information used for internal quality control can be utilized in the following situations. For example, quality control information used for internal quality control is used to check whether there are any problems with the staining state of smear specimens obtained after maintenance of the testing equipment or after reagent replacement. Also, by checking the trend of increase or decrease in characteristic values ​​over time using quality control information used for internal quality control, it is possible to identify abnormalities in reagents or testing equipment in advance. In addition, since environmental conditions such as temperature and humidity affect the staining properties of smear specimens, quality control information used for internal quality control can also be used to optimize staining processing conditions in response to changes in environmental conditions.

[0053] Next, we will explain an example of an output screen for external quality control, which displays quality control information and feature values ​​that reflect the staining status of the smear specimens under management. External quality control compares the control range of feature values ​​for multiple testing facilities, such as those in the same region or different regions, with the feature values ​​that reflect the staining status of the smear specimens under management among the multiple smear specimens produced. For example, if a specific feature value that reflects the staining status of the smear specimens under management at a particular testing facility falls outside the control range of feature values ​​for multiple testing facilities, such as those in the same region or different regions, external quality control makes it possible to objectively understand the position of that particular testing facility among multiple testing facilities.

[0054] As shown in Figure 9, inspection systems 70-1 to 70-4 are installed in multiple inspection facilities (inspection facilities A to D). The generation device 80 is connected via a network to the smear preparation devices 20, specimen transport devices 30, and specimen image acquisition devices 40 of the multiple inspection systems 70-1 to 70-4. The generation device 80 is installed in the facility of the provider (for example, the manufacturer of the specimen image acquisition device 40) and acquires image data of the captured smear slides 10 from each of the inspection systems 70-1 to 70-4 in the multiple inspection facilities (inspection facilities A to D), acquires feature values ​​that reflect the staining state of the smear from the image data, and generates quality control information for the smear based on the acquired feature values.

[0055] As shown in the output screen in Figure 10, the plots P of multiple feature values ​​may include, for example, a plot P of feature values ​​obtained from at least one smear sample obtained at testing facility A (the first facility) and a plot P of feature values ​​obtained from at least one smear sample obtained at testing facilities B to D (the second to fourth facilities). The quality control information output on the output screen can show the quality control standards for each of the multiple testing facilities A to D (the first to fourth facilities) according to their respective staining conditions. Therefore, it becomes possible to easily grasp the differences in the staining state of smear samples adopted among the multiple testing facilities A to D (the first to fourth facilities). Furthermore, when a medical technologist belonging to one testing facility is asked to review or analyze images of smear samples obtained at another testing facility, the staining state of the smear sample often differs from the staining state handled at the testing facility to which they belong, making the analysis of the smear sample images difficult. In this regard, the above quality control information, which allows for easy identification of differences in staining status of smear specimens between multiple testing facilities, can serve as supplementary information in image analysis and reviews between multiple testing facilities.

[0056] In the example output screen in Figure 10, the differences in staining status of smear specimens between multiple different testing facilities are identified. However, for multiple different staining conditions, the quality control information obtained from the characteristic values ​​of smear specimens for each of the multiple smear and staining conditions may also be output to the output screen in an identifiable manner.

[0057] As shown in the output screen in Figure 11, the plot P of multiple feature values ​​includes the plot P of feature values ​​obtained from each of the multiple smear samples of testing facility A, and the plot P of feature values ​​obtained from each of the multiple smear samples of testing facilities B, D, ..., or X. As shown in the output screen, at least one of the control width MW and the target value ACI (quality control information) may include quality control information for testing facility A generated based on the feature values ​​P of testing facility A, and quality control information for testing facilities B, D, ..., ..., or X generated based on the feature values ​​P of testing facilities B, D, ..., ..., or X.

[0058] The quality control information output on the output screen shown in Figure 11 can indicate variations in the staining status of smear specimens across multiple testing facilities and can serve as a standard for quality control regarding the staining status of smear specimens across multiple testing facilities. Therefore, it becomes possible to easily grasp the differences in the staining status of smear specimens adopted by multiple testing facilities. Furthermore, the information on the characteristic values ​​of smear specimens at one's own testing facility and the quality control information can be used to examine the staining conditions at one's own testing facility.

[0059] Figure 12 shows an example of feature values ​​that reflect the staining properties of red blood cells. As shown in Figure 12, feature values ​​that reflect the staining properties of red blood cells include, for example, the average luminance value of each color component of the red blood cell, the average hue value of the red blood cell, the average saturation value of the red blood cell, the average lightness value of the red blood cell, and the HSV value of the red blood cell. Now, referring to Figure 13, an example of the process for obtaining the average luminance value of the red component of a red blood cell, which is an example of a feature value, will be explained.

[0060] As shown in Figure 13, the control unit 50 shown in Figure 4 acquires multiple image data IDs generated by imaging a single smear specimen (sample). Next, the control unit 50 shown in Figure 4 acquires the average value of the red (R) of the red blood cell region (redcell_r_mean) for each of the acquired multiple image data IDs. Then, based on the total number of multiple image data IDs of the single smear specimen, the control unit 50 acquires the average value of the average value of the red (R) of the red blood cell region (redcell_r_mean) for each of the multiple image data IDs (RBC Color R).

[0061] The feature values ​​obtained from leukocytes in image data include feature values ​​corresponding to color information obtained from the structural components of leukocytes, such as the nucleus, cytoplasm, and granules. These reflect the staining properties of the structural components of leukocytes. Specific examples of feature values ​​obtained from leukocytes are explained with reference to Figure 14.

[0062] Figure 14 shows an example of feature values ​​that reflect the staining properties of leukocytes. As shown in Figure 14, feature values ​​that reflect the staining properties of leukocytes include, for example, the average luminance values ​​of each color component in the cytoplasmic region of leukocytes, the average hue values ​​in the cytoplasmic region of leukocytes, the average saturation values ​​in the cytoplasmic region of leukocytes, and the average lightness values ​​in the cytoplasmic region of leukocytes. Feature values ​​that reflect the staining properties of leukocytes also include, for example, the average luminance values ​​of each color component in the nuclear region of leukocytes, the average hue values ​​in the nuclear region of leukocytes, the average saturation values ​​in the nuclear region of leukocytes, and the average lightness values ​​in the nuclear region of leukocytes. Furthermore, feature values ​​that reflect the staining properties of leukocytes may also include the standard deviation values ​​for luminance, hue, saturation, and lightness in the cytoplasmic and nuclear regions of leukocytes. Feature values ​​that reflect the staining properties of leukocytes also include, for example, the average granule index of leukocytes. Note that feature values ​​also include feature values ​​that reflect the morphology of blood cells.

[0063] In addition to the "feature values ​​that reflect the staining state of the smear" mentioned above, the feature values ​​obtained from blood cells in smear image data also include feature values ​​that reflect the morphological information of the blood cells.

[0064] Figure 15 shows an example of feature values ​​that reflect the morphology of blood cells. As shown in Figure 15, feature values ​​that reflect the morphology of blood cells obtained from blood cells include, for example, the average cell diameter of blood cells, the average nuclear / central ratio of blood cells, the average cytoplasmic area of ​​blood cells, the average roundness (circularity) of blood cells, and the average roundness of the nucleus.

[0065] Here, referring to Figures 16 and 17, we will explain an example of the process for obtaining the granule index (Granule_Index) for neutrophils, a type of white blood cell, as an example of a characteristic value.

[0066] Figure 16 is a flowchart showing the feature value acquisition process in step S2 of Figure 5, and is a flowchart of an example of the granule index acquisition process. As shown in Figure 16, the control unit 50 shown in Figure 4 acquires image data corresponding to the smear, for example, including leukocytes (step S21). The control unit 50 extracts the cytoplasmic region by performing, for example, local binarization on the acquired image data (step S22). The control unit 50 extracts granules (step S23). Specifically, the control unit 50 identifies granules (for example, granules of several pixels or more) from the extracted cytoplasmic region based on a predetermined number of pixels. For each identified granule, the control unit 50 acquires the granule size (area) and the average brightness. The control unit 50 classifies the identified granules into primary granules and secondary granules by setting predetermined thresholds for the granule size and the difference between the average granule brightness and the average cytoplasmic brightness.

[0067] Next, the control unit 50 calculates the region related to granules (step S24). Specifically, the control unit 50 calculates the total number of primary and secondary granules in the granule region recognized as granules. The control unit 50 may also calculate the total area of ​​the region recognized as granules (total granule area), or the ratio of the total granule area of ​​the region recognized as granules. The control unit 50 obtains at least one of the calculated total number of granules in the granule region, the total granule area, and the ratio of the total granule area as the granule index (step S25).

[0068] Here, referring to Figure 17, an example of the process for obtaining the granule index, for example, for neutrophils among white blood cells, as an example of a feature value, will be explained. As shown in Figure 17, (1) the control unit 50 shown in Figure 4 acquires image data containing white blood cells corresponding to the smear specimen. (2) The control unit 50 extracts the cytoplasmic region by performing, for example, local binarization on the acquired image data. Here, the "granule index" is a feature value that reflects the number of granules contained in white blood cells, and can be obtained based on the number or area of ​​granule regions identified in the cytoplasmic region of the white blood cell. The "granule region" can be identified separately from the cytoplasmic portion based on the binarized image data.

[0069] (3) The control unit 50 identifies granules (for example, granules of several pixels or more) from the extracted cytoplasmic region based on a predetermined number of pixels. For each identified granule, the control unit 50 obtains the granule size (area) and the average brightness. The control unit 50 classifies the identified granules into primary granules and secondary granules by setting predetermined thresholds for the granule size and the difference between the average granule brightness and the average cytoplasmic brightness (for example, the difference between the average brightness of the cytoplasm excluding the granule portion and the average granule brightness). Here, if the difference between the average granule brightness and the average cytoplasmic brightness is 0 (zero), it indicates that the brightness is the same as the background portion of the image data. Also, if the difference between the average granule brightness and the average cytoplasmic brightness is large in the negative direction, it indicates that the granule portion is denser (darker) than other parts. The granule classification process will be explained in more detail with reference to Figure 18.

[0070] Figure 18 shows an example of the extraction process for primary and secondary granules. For example, if the threshold for the granule size of secondary granules is set to "10" pixels and the threshold for the difference between the average granule brightness and the average cytoplasmic brightness is set to "0", the control unit 50 classifies the identified granules into primary granules G1 contained in region R1 enclosed by the dashed line in the graph in Figure 18, and secondary granules G2 contained in region R2, which is different from region R1 in the graph in Figure 18. Note that each of the above thresholds can be arbitrarily set based on, for example, the resolution of the image data, and can be changed as appropriate. The size of one pixel can also be set as appropriate, but for example, it is approximately 0.01 μm. 2 That is the case.

[0071] Returning to Figure 17, (4) the control unit 50 calculates the total number of primary and secondary granules in the granular region recognized as granules. The control unit 50 may also calculate the total area of ​​the region recognized as granules (total granule area), or the total granule area ratio of the region recognized as granules. Here, since there is a predetermined correlation between the total number of granules in the granular region and the total granule area and total granule area ratio, the total granule area and total granule area ratio can be adopted as feature values ​​in addition to, or instead of, the total number of granules in the granular region. The total granule area ratio refers to, for example, the ratio of how much of the cytoplasmic area is occupied by the granule area. (5) The control unit 50 obtains at least one of the calculated total number of granules in the granular region, the total granule area, and the total granule area ratio as the granule index. By obtaining the granule index, it is possible to objectively and quantitatively show changes in granules in blood cells. Therefore, it may be useful in diagnosing and evaluating diseases that involve increases or decreases in granules, such as infectious diseases where granules increase, and myelodysplastic syndromes (MDS) where granules decrease.

[0072] Figure 19 shows an example of an output screen displaying quality control information and feature values ​​reflecting the staining status of the smear specimens under management. When outputting quality control information and feature values ​​reflecting the staining status of the smear specimens under management, the output is designed to identify whether or not the feature values ​​reflecting the staining status of the smear specimens under management fall within a predetermined range. In the example screen shown in Figure 19, for example, since plot P1 corresponding to a specific feature value falls outside the management width MW, plot P1 corresponding to this specific feature value is output on the output screen in a different form from other plots. For example, plot P1 is output with more emphasis than other plots. More specifically, plot P1 is output in a different color, a different size, or a different shape (for example, plot P1 is round and other plots are square). With this configuration, it is possible to easily distinguish a plot corresponding to a specific feature value that does not fall within the predetermined range from other plots that do fall within the predetermined range.

[0073] Figure 20 shows an example of an output screen that includes a predetermined alert display for plots corresponding to specific feature values ​​that fall outside the control width MW. As shown in Figure 20, when the user performs a first specified operation on plot P3 corresponding to a specific feature value that falls outside the control width MW, an alert display AI containing detailed information about the specific feature value corresponding to plot P3 is output. The user's first specified operation is optional, but may include, for example, the cursor C1 corresponding to the mouse operation performed by the user stopping on plot P3 for a certain period of time. With this configuration, detailed information about the specific feature value that falls outside the control width can be appropriately presented to the user. Therefore, by interacting with this detailed information, the user can easily identify the cause of the feature value falling outside the control width. When the cursor C1 stops on plot P3 and a single mouse click operation is performed, a screen displaying multiple image data associated with the feature value of plot P3 may be displayed, as shown in Figure 21, which will be described later.

[0074] Figure 21 shows an example of a screen displaying multiple image data obtained from a single smear sample. In particular, Figure 21 is a screen that is output when a user performs a predetermined operation on the screen shown in Figure 20. For example, when a second user operation is performed on plot P5 corresponding to a specific feature value on the output screen shown in Figure 20, the screen shown in Figure 21 is output, displaying the multiple image data IDs used to obtain the specific feature value corresponding to plot P5. With this configuration, the user can easily check a list of multiple image data IDs used to obtain the specific feature value they desire.

[0075] The user's second specified operation is optional, but may include, for example, a double-click operation with the mouse when cursor C3 stops on plot P5. Furthermore, it is preferable that the user's first and second specified operations are different, but they may be the same operation.

[0076] Figure 22 shows an example of a screen that displays a specific image data and multiple feature values ​​associated with that image data. In particular, Figure 22 is a screen that is output when a user performs a predetermined operation on the screen shown in Figure 21. For example, when a third specified operation is performed by the user on a specific image data ID 1 on the output screen shown in Figure 21, a screen is output that includes a dashed area R3 that displays detailed information about the multiple feature values ​​corresponding to the specified image data ID 1. With this configuration, the user can easily check a list of multiple feature values ​​corresponding to the specific image data they desire. The user's third specified operation is optional, but includes, for example, double-clicking with the mouse when the cursor C5 shown in Figure 21 stops on image data ID 1.

[0077] Here, feature values ​​FV (e.g., granule_Index) that fall outside the pre-set normal range (management range) may be displayed in a different form from other feature values. With this configuration, it is easy to distinguish a specific feature value FV, which is an outlier, from other (normal) feature values.

[0078] Figure 23 shows an example of a screen where the feature values ​​to be acquired can be selected. The feature values ​​to be acquired, or the feature values ​​to be output on the screen, are predetermined for each testing facility or region, for example. On the other hand, by operating the feature value selection screen as shown in Figure 23, the user can select one or more feature values ​​of their choice. With this configuration, the feature values ​​desired by the user can be output on the output screen of feature values ​​and quality control information that reflect the staining state of the smear specimen under management.

[0079] Figure 24 shows an example of a screen where a normal range (control range) can be set for any feature value. In the screen shown in Figure 24, the user can set a control range for any feature value in the area R5 enclosed by the dashed line, for example, based on the results of quality control performed at each testing facility. For example, the values ​​of "feature" and "cell_s_mean" for "cell" and "5 LY", and "feature" and "cell_v_mean" for "cell" and "6 MO", indicate the control range for feature values ​​related to the staining state of the cytoplasm. Also, the values ​​of "feature" and "segment_num" for "cell" and "7 SNE" indicate the control range for feature values ​​related to the number of lobes in neutrophils. With this configuration, the user can arbitrarily set a control range for feature values ​​based on the daily quality control indicators at the testing facility, thereby suppressing subjective inter-examiner judgment errors.

[0080] Image data corresponding to abnormal feature values ​​that fall outside the control range set for each feature value shown in Figure 24 may be displayed in a different form from other image data IDs corresponding to normal feature values, for example, in a screen displaying multiple image data as shown in Figure 20, such as image data ID1, ID3, ID5, ID7, ID9, ID11, ID13. The method of displaying in a different form is arbitrary, but may include, for example, adding a colored border only to the image data corresponding to the abnormal feature value on the screen, or displaying it larger than other image data corresponding to normal feature values.

[0081] <Other Embodiments> The above embodiments are provided to facilitate understanding of the present invention and are not intended to limit it. The present invention can be modified or improved without departing from its spirit (for example, by combining the embodiments or omitting some of the components of each embodiment), and equivalents thereof are also included.

[0082] For example, although the smear slide 10 was prepared using the smear preparation device 20 included in the testing systems 70-1 to 70-4, a medical technologist belonging to the testing facility may also prepare the smear slide 10 manually. Furthermore, although the smear slide 10 was imaged using the specimen image acquisition device 40, a medical technologist belonging to the testing facility may also image the smear slide 10 manually. [Explanation of Symbols]

[0083] 10: Smear slide, 20: Smear preparation device, 30: Specimen transport device, 40: Specimen image acquisition device, 50: Control unit, 70-1~70-4: Inspection system, 71: Storage unit, 72: Input unit, 73: Display unit, 74: Communication unit, 80: Generation device, 100: Generation system

Claims

1. A method for generating quality control information for smear specimens, Obtaining image data from each of multiple smear specimens, Obtaining multiple feature values ​​from each of the multiple image data that reflect the staining state of the smear specimen, This includes generating the accuracy control information based on criteria that include at least upper and lower limits statistically calculated from the aforementioned multiple feature values, A method for generating quality control information characterized by the following features.

2. The further includes outputting multiple feature values ​​that reflect the staining status of the smear specimen under management, along with the quality control information. The method for generating accuracy control information according to feature 1.

3. The aforementioned plurality of feature values ​​include the color information values ​​obtained from red blood cells in each of the plurality of smear samples. A method for generating accuracy control information according to claim 1 or 2, characterized by the above.

4. From each of the multiple image data, a second feature value corresponding to the color information of leukocytes in each of the multiple smear specimens is obtained, The method further includes generating second quality control information corresponding to the plurality of smear samples based on the second characteristic value, A method for generating quality control information according to any one of claims 1 to 3.

5. The color information of the white blood cell includes color information related to at least one of the nucleus, cytoplasm, and granules of the white blood cell. The method for generating accuracy control information according to feature 4.

6. The aforementioned plurality of feature values ​​include values ​​obtained from at least one of the color components, hue, saturation, and brightness obtained from image data of a blood smear specimen containing blood cells, as well as combinations thereof. A method for generating accuracy control information according to any one of claims 1 to 5.

7. The aforementioned multiple feature values ​​are obtained from each of the multiple smear specimens obtained at a predetermined facility. A method for generating accuracy control information according to any one of claims 1 to 6.

8. The plurality of feature values ​​include feature values ​​obtained from at least one smear sample obtained at the first facility and feature values ​​obtained from at least one smear sample obtained at the second facility. A method for generating accuracy control information according to any one of claims 1 to 7.

9. The aforementioned multiple feature values, The feature values ​​obtained from each of the multiple smear samples of the first facility, The feature values ​​obtained from each of the multiple smear samples of the second facility and Includes, The aforementioned accuracy control information, Quality control information for the first facility generated based on the characteristic values ​​of the first facility, Quality control information for the second facility generated based on the characteristic values ​​of the second facility, including, The method for generating accuracy control information according to feature 8.

10. The accuracy control information is generated based on the feature values, which include the feature values ​​of the first facility and the feature values ​​of the second facility. The method for generating accuracy control information according to feature 8.

11. The accuracy control information includes statistical values ​​calculated from the plurality of feature values, A method for generating accuracy control information according to any one of claims 1 to 10, characterized by the present invention.

12. The accuracy control information includes upper and lower limits generated based on statistical values ​​calculated from the plurality of feature values. A method for generating accuracy control information according to any one of claims 1 to 11.

13. When outputting multiple feature values ​​that reflect the staining state of the smear specimen under management and the quality control information, the output is configured to identify whether or not the multiple feature values ​​that reflect the staining state of the smear specimen under management fall within a predetermined range. A method for generating accuracy control information according to any one of claims 1 to 12.

14. Multiple feature values ​​that reflect the staining state of the smear specimens under management are statistical values ​​calculated from the image data of each of the multiple smear specimens. The method for generating accuracy control information according to feature 13.

15. The aforementioned feature values ​​are statistical values ​​calculated from multiple image data of the smear sample. A method for generating accuracy control information according to any one of claims 1 to 14.

16. The method further includes preparing a blood smear by smearing a blood sample onto a glass slide using a smear preparation device. A method for generating accuracy control information according to any one of claims 1 to 14.

17. The aforementioned quality control information is generated daily by acquiring image data from each of the multiple smear samples prepared. The method for generating accuracy control information according to feature 16.

18. The further includes outputting a screen that allows the user to set a management range for the multiple feature values ​​based on user input, A method for generating accuracy control information according to any one of claims 1 to 17.

19. The aforementioned plurality of feature values ​​include feature values ​​that reflect the morphology of the smear specimen. A method for generating quality control information according to any one of claims 1 to 18.

20. A device for generating quality control information for smear specimens, Includes a control unit, The control unit, Image data is obtained from each of multiple smear specimens. Multiple feature values ​​reflecting the staining state of the smear specimen are obtained from each of the multiple image data, The accuracy control information is generated based on criteria that include at least an upper limit and a lower limit statistically calculated from the aforementioned multiple feature values. A device for generating accuracy control information, characterized by the following features.

21. A device that generates quality control information for smear specimens, Obtaining image data from each of multiple smear specimens, Obtaining multiple feature values ​​from each of the multiple image data that reflect the staining state of the smear specimen, The accuracy control information is generated based on criteria that include at least an upper limit and a lower limit statistically calculated from the aforementioned multiple feature values. A program to execute.

Images