Sample analyzer, sample analysis method, and program

Abstract

To provide a specimen analyzer, a specimen analysis method and a program with which it is possible to reduce the load of a computer that analyzes the data obtained by specimen measurement, using an artificial intelligence algorithm.SOLUTION: A specimen analyzer 4000 for analyzing an analyte in a specimen comprises a measurement unit 400 that includes an optical detection unit for acquiring a first optical signal from a first measurement sample and acquiring a second optical signal from a second measurement sample, and an analysis unit 300 that analyzes first data corresponding to the first optical signal and second data corresponding to the second optical signal. The analysis unit 300 executes the analysis of a first measurement item for the first measurement sample by a first analysis operation that processes the first data by an artificial intelligence algorithm, and executes the analysis of a second measurement item for the first measurement sample by the first analysis operation and / or a second analysis operation that processes a first representative value of the first data that corresponds to the feature of the analyte.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present invention relates to a sample analyzer, a sample analysis method, and a program for analyzing a sample. [Background technology]

[0002] Patent Document 1 describes a method of classifying cells according to their type by analyzing signals obtained by measuring cells using a flow cytometer using an artificial intelligence algorithm. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] International Publication No. 2018 / 203568 Summary of the Invention [Problem to be solved by the invention]

[0004] When processing data using an artificial intelligence algorithm, the greater the data volume, the greater the load on the computer processing the data. For example, if the amount of information used to classify components in a sample (e.g., cells, formed elements) is increased to improve classification accuracy, the amount of information obtained from each component increases in samples such as blood and urine that contain multiple components, resulting in an increase in the data volume per sample. The data volume also increases when the number of samples to be tested increases. Patent Document 1 does not disclose any technology that can reduce the computer load when processing data using an artificial intelligence algorithm.

[0005] In view of the above problems, the present invention aims to provide a sample analyzer, a sample analysis method, and a program that can reduce the load on a computer that analyzes data obtained by measuring samples using an artificial intelligence algorithm. [Means for solving the problem]

[0006] The sample analyzer (4000) of the present invention relates to a sample analyzer for analyzing an analyte in a sample. The sample analyzer (4000) of the present invention comprises a measurement unit (400) including a plurality of first sample preparation sections (440a) that prepare first measurement samples based on the sample and a first reagent, second sample preparation sections (440b, 440c, 440d) that prepare second measurement samples based on the sample and a second reagent, and optical detection sections (410, 470) that acquire first optical signals (80a, 80b, 80c) from the first measurement samples and acquire second optical signals (80a, 80b, 80c) from the second measurement samples, and an analysis unit (300) that analyzes first data (82a, 82b, 82c) corresponding to the first optical signals (80a, 80b, 80c) and second data (82a, 82b, 82c) corresponding to the second optical signals (80a, 80b, 80c). The analysis unit (300) performs an analysis of a first measurement item for a first measurement sample by a first analysis operation that processes first data (82a, 82b, 82c) using an artificial intelligence algorithm (60), performs an analysis of a second measurement item for the first measurement sample by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte from the first data (82a, 82b, 82c), and performs an analysis of a second measurement sample by a third analysis operation that processes a second representative value corresponding to the characteristics of the analyte from the second data (82a, 82b, 82c).

[0007] According to the sample analyzer of the present invention, the analytical processing of data corresponding to optical signals obtained from a sample is divided between a first analytical operation using an artificial intelligence algorithm and second and third analytical operations that process first and second representative values ​​corresponding to the characteristics of the analyte, thereby reducing the load on the analytical unit, which is a computer that processes the data, compared to when data corresponding to optical signals are analyzed uniformly using only an artificial intelligence algorithm.

[0008] Furthermore, the analysis of the second measurement item for the first measurement sample is performed by at least one of a first analysis operation using an artificial intelligence algorithm and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte. As a result, even if it is difficult to analyze the second measurement item using the second analysis operation, for example, it is possible to appropriately perform a highly accurate analysis using the first analysis operation.

[0009] Furthermore, since multiple first sample preparation units are provided, multiple first measurement samples can be prepared in parallel based on multiple specimens, thereby improving the throughput of specimen analysis.

[0010] The sample analysis method of the present invention relates to a sample analysis method for analyzing an analyte in a sample, and includes the steps of: preparing a first measurement sample based on the sample and a first reagent, and preparing a second measurement sample based on the sample and a second reagent (S301); acquiring first optical signals (80a, 80b, 80c) from the first measurement sample, and acquiring second optical signals (80a, 80b, 80c) from the second measurement sample (S1, S11, S121, S131, S302); and an analysis step (S2, S3, S14, S16, S71, S74, S81, S84, S91, S95, S122, S124, S132, S134, S201, S202) of analyzing first data (82a, 82b, 82c) corresponding to the signals (80a, 80b, 80c) and second data (82a, 82b, 82c) corresponding to the second optical signals (80a, 80b, 80c). In the analysis steps (S2, S3, S14, S16, S71, S74, S81, S84, S91, S95, S122, S124, S132, S134, S201, S202), analysis of the first measurement item for the first measurement sample is performed by a first analysis operation that processes first data (82a, 82b, 82c) using an artificial intelligence algorithm (60), analysis of the second measurement item for the first measurement sample is performed by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte from the first data (82a, 82b, 82c), and analysis of the second measurement sample is performed by a third analysis operation that processes a second representative value corresponding to the characteristics of the analyte from the second data (82a, 82b, 82c).

[0011] According to the sample analysis method of the present invention, the analytical processing of data corresponding to optical signals obtained from a sample is divided between a first analytical operation using an artificial intelligence algorithm and second and third analytical operations that process first and second representative values ​​corresponding to the characteristics of the analyte, thereby reducing the load on the computer that processes the data compared to when data corresponding to optical signals are analyzed uniformly using only an artificial intelligence algorithm.

[0012] Furthermore, the analysis of the second measurement item for the first measurement sample is performed by at least one of a first analysis operation using an artificial intelligence algorithm and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte. As a result, even if it is difficult to analyze the second measurement item using the second analysis operation, for example, it is possible to appropriately perform a highly accurate analysis using the first analysis operation.

[0013] The program of the present invention relates to a program that causes a computer (300, 600, 3001, 3002, 6001, 6002) to execute a process for analyzing an analyte in a sample. The program of the present invention includes a process for analyzing first data (82a, 82b, 82c) corresponding to first optical signals (80a, 80b, 80c) acquired from a first measurement sample prepared based on the sample and a first reagent, and second data (82a, 82b, 82c) corresponding to second optical signals (80a, 80b, 80c) acquired from a second measurement sample prepared based on the sample and a second reagent. In this process, an analysis of a first measurement item for a first measurement sample is performed by a first analysis operation that processes first data (82a, 82b, 82c) using an artificial intelligence algorithm (60), an analysis of a second measurement item for the first measurement sample is performed by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte from the first data (82a, 82b, 82c), and an analysis of a second measurement sample is performed by a third analysis operation that processes a second representative value corresponding to the characteristics of the analyte from the second data (82a, 82b, 82c).

[0014] According to the program of the present invention, the analytical processing of data corresponding to optical signals obtained from a sample is divided between a first analytical operation using an artificial intelligence algorithm and second and third analytical operations that process first and second representative values ​​corresponding to the characteristics of the analyte, thereby reducing the load on the computer that processes the data compared to when data corresponding to optical signals are analyzed uniformly using only an artificial intelligence algorithm.

[0015] Furthermore, the analysis of the second measurement item for the first measurement sample is performed by at least one of a first analysis operation using an artificial intelligence algorithm and a second analysis operation that processes a first representative value corresponding to the characteristics of the analyte. As a result, even if it is difficult to analyze the second measurement item using the second analysis operation, for example, it is possible to appropriately perform a highly accurate analysis using the first analysis operation. [Effects of the Invention]

[0016] According to the present invention, it is possible to reduce the load on a computer that analyzes data obtained from the measurement of a sample using an artificial intelligence algorithm. [Brief explanation of the drawings]

[0017] [Figure 1] FIG. 1 is a diagram schematically illustrating an example of the configuration of a sample analyzer according to the first embodiment. [Figure 2] FIG. 2 is a diagram showing an outline of analysis when the optical detection unit is a detection unit based on flow cytometry according to the first embodiment. [Figure 3] FIG. 3 is a diagram schematically showing waveform data and representative values ​​according to the first embodiment. [Figure 4] FIG. 4 is a diagram showing an outline of an analysis according to the first embodiment when the optical detection unit is a detection unit that detects transmitted light or scattered light from a measurement sample. [Figure 5] FIG. 5 is a flowchart showing an example of a sample analysis method according to the first embodiment. [Figure 6] FIG. 6 is a flowchart showing an example of setting an analysis operation based on a rule set in an analysis unit according to the second embodiment. [Figure 7] FIG. 7 is a flowchart showing an example of performing analysis according to measurement items according to the third embodiment. [Figure 8] FIG. 8 is an exemplary diagram illustrating a screen for setting AI analysis or computational analysis for each measurement item according to the third embodiment. [Figure 9]FIG. 9 is a flowchart showing an example in which analysis is performed in response to a measurement order according to the third embodiment. [Figure 10] FIG. 10 is an exemplary view schematically showing a screen for setting an analysis mode for a measurement order according to the third embodiment. [Figure 11] FIG. 11 is a flowchart showing an example of how analysis is performed depending on the analysis mode of the device according to the third embodiment. [Figure 12] FIG. 12 is an exemplary view schematically showing a screen for setting the analysis mode of the analysis unit according to the third embodiment. [Figure 13] FIG. 13 is a flowchart showing an example in which analysis is performed depending on the type of measurement order according to the third embodiment. [Figure 14] FIG. 14 is an exemplary view schematically showing a screen for setting an analysis mode for each type of measurement order according to the third embodiment. [Figure 15] FIG. 15 is a flowchart showing an example of an analysis being performed according to the type of measurement item and measurement order according to the third embodiment. [Figure 16] FIG. 16 is an exemplary diagram illustrating a screen for setting AI analysis or computational analysis for each type of measurement item and measurement order according to the third embodiment. [Figure 17] FIG. 17 is a flowchart showing an example of determining whether or not an AI analysis is necessary based on a flag generated by a computational analysis according to the third embodiment. [Figure 18] FIG. 18 is an exemplary diagram illustrating a screen for setting AI analysis for each flag of the analysis result according to the third embodiment. [Figure 19] FIG. 19 is a flowchart illustrating an example of performing AI analysis on specific analytes classified in a computational analysis according to the third embodiment. [Figure 20] FIG. 20 is an exemplary diagram illustrating a screen for setting whether or not to perform AI analysis for each type of analyte according to the third embodiment. [Figure 21]FIG. 21 is a diagram illustrating a classification method based on the computational analysis and AI analysis executed in the process shown in FIG. 19 according to the third embodiment. [Figure 22] FIG. 22 is a flowchart illustrating an example of performing AI analysis when a specific classification is performed in the computational analysis according to the third embodiment. [Figure 23] FIG. 23 is an exemplary diagram illustrating a screen for setting whether or not to perform AI analysis on a specific type of analyte according to the third embodiment. [Figure 24] FIG. 24 is a block diagram showing the configuration of a measurement unit according to the fourth embodiment. [Figure 25] FIG. 25 is a diagram schematically illustrating the configuration of the optical system of the FCM detection unit according to the fourth embodiment. [Figure 26] FIG. 26 is a block diagram showing the configuration of an analysis unit according to the fourth embodiment. [Figure 27] FIG. 27 is a block diagram showing the configuration of a measurement unit when the sample analyzer according to the fourth embodiment counts and classifies blood cells in a blood sample. [Figure 28] FIG. 28 is a block diagram showing the configuration of the specimen aspirating section and the sample preparing section in the measurement unit of FIG. 27 according to the fourth embodiment. [Figure 29] FIG. 29 is a block diagram showing another configuration of the sample preparation unit shown in FIG. 28 according to the fourth embodiment. [Figure 30] FIG. 30 is a flowchart illustrating an example in which analysis is performed according to a measurement channel according to the fourth embodiment. [Figure 31] FIG. 31 is an exemplary diagram illustrating a screen for setting AI analysis or computational analysis for each measurement channel according to the fourth embodiment. [Figure 32] FIG. 32 is a schematic diagram for explaining waveform data used in the analysis method according to the fourth embodiment. [Figure 33] FIG. 33 is a schematic diagram illustrating an example of a method for generating training data used to train an AI algorithm for determining the type of analyte in a sample, according to the fourth embodiment. [Figure 34] FIG. 34 is a diagram showing label values ​​corresponding to cell types according to the fourth embodiment. [Figure 35] FIG. 35 is a diagram schematically illustrating a method for analyzing waveform data of an analyte in a sample using an AI algorithm according to the fourth embodiment. [Figure 36] FIG. 36 is a flowchart illustrating an example of performing AI analysis on waveform data acquired through a WDF channel according to the fourth embodiment. [Figure 37] FIG. 37 is a flowchart showing an example of classifying nucleated red blood cells and basophils by AI analysis and classifying others by computational analysis based on waveform data acquired by the WDF channel according to the fourth embodiment. [Figure 38] FIG. 38 is a flowchart showing an example of performing AI analysis on neutrophils / basophils identified by analysis of computational processing in the WDF channel according to the fourth embodiment. [Figure 39] FIG. 39 is a block diagram schematically showing the configuration of a measurement unit according to the fifth embodiment. [Figure 40] FIG. 40 is a side view schematically showing measurement by the detection block according to the fifth embodiment. [Figure 41] FIG. 41 is a flowchart showing an analysis example according to the fifth embodiment. [Figure 42] FIG. 42 is a block diagram showing the configuration of a sample analyzer according to the sixth embodiment. [Figure 43] FIG. 43 is a block diagram showing the configuration of an analysis unit according to the sixth embodiment. [Figure 44] FIG. 44 is a block diagram showing another configuration of the sample analyzer according to the sixth embodiment. [Figure 45] FIG. 45 is a diagram illustrating an example of the configuration of a parallel processing processor according to the sixth embodiment. [Figure 46] FIG. 46 is a diagram schematically illustrating an example of mounting a parallel processor according to the sixth embodiment. [Figure 47]FIG. 47 is a diagram schematically illustrating an example of mounting a parallel processor according to the sixth embodiment. [Figure 48] FIG. 48 is a diagram schematically illustrating an example of mounting a parallel processor according to the sixth embodiment. [Figure 49] FIG. 49 is a diagram showing another example of the installation of the parallel processing processor according to the sixth embodiment. [Figure 50] FIG. 50 is a diagram illustrating an example of the configuration of a parallel processor that executes arithmetic processing according to the sixth embodiment. [Figure 51] FIG. 51 is a diagram illustrating an outline of a matrix operation executed by a parallel processor according to the sixth embodiment. [Figure 52] FIG. 52 is a conceptual diagram showing how a plurality of arithmetic processes are executed in parallel by a parallel processor according to the sixth embodiment. [Figure 53] FIG. 53 is a diagram schematically illustrating an outline of the arithmetic processing related to the convolution layer according to the sixth embodiment. [Figure 54] FIG. 54 is a flowchart showing the analysis operations of the analysis unit and the measurement unit according to the sixth embodiment. [Figure 55] FIG. 55 is a flowchart showing details of the AI ​​analysis in step S201 of FIG. 54 according to the sixth embodiment. [Figure 56] FIG. 56 is a flowchart showing details of step S2011 in FIG. 55 according to the sixth embodiment. [Figure 57] FIG. 57 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 58] FIG. 58 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 59] FIG. 59 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 60] FIG. 60 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 61] FIG. 61 is a block diagram showing another configuration of the sample analyzer according to the sixth embodiment. [Figure 62] FIG. 62 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 63] FIG. 63 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 64] FIG. 64 is a diagram illustrating a configuration example of a parallel processor that executes arithmetic processing according to the sixth embodiment. [Figure 65] FIG. 65 is a block diagram showing the configuration of a computer according to the sixth embodiment. [Figure 66] FIG. 66 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 67] FIG. 67 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 68] FIG. 68 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 69] FIG. 69 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 70] FIG. 70 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 71] FIG. 71 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 72] FIG. 72 is a block diagram showing another configuration of the measurement unit according to the sixth embodiment. [Figure 73] FIG. 73 is a block diagram showing another configuration of the analysis unit according to the sixth embodiment. [Figure 74] FIG. 74 is a diagram schematically illustrating the configuration of a waveform data analysis system according to the seventh embodiment. [Figure 75] FIG. 75 is a block diagram showing the configuration of a deep learning device according to the seventh embodiment. [Figure 76] FIG. 76 is a functional block diagram of a deep learning device according to the seventh embodiment. [Figure 77] FIG. 77 is a flowchart showing the processing performed by the deep learning device according to the seventh embodiment. [Figure 78] FIG. 78 is a schematic diagram illustrating the structure of a neural network according to the seventh embodiment, a schematic diagram illustrating the operations at each node, and a schematic diagram illustrating the operations between nodes. DETAILED DESCRIPTION OF THE INVENTION

[0018] The outline and embodiments of the present invention will be described in detail below with reference to the accompanying drawings. In the following description and drawings, the same reference numerals indicate the same or similar components, and for convenience, the description of the same or similar components will be omitted.

[0019] [Embodiment 1] This embodiment discloses a sample analyzer, a sample analysis method, and a program that can perform both analysis using an artificial intelligence (AI) algorithm and analysis without using an AI algorithm on data obtained by measuring a sample.

[0020] In analysis using AI algorithms, for example, data is analyzed using a large amount of matrix calculation processing. Hereinafter, analysis using AI algorithms will be referred to as "AI analysis" for convenience. In AI analysis, for example, convolution calculations are performed using AI algorithms.

[0021] In analysis that does not use an AI algorithm, for example, data is analyzed by computational processing of representative values ​​corresponding to the characteristics of the analyte. Hereinafter, for convenience, an analysis method that analyzes data by computational processing of representative values ​​corresponding to the characteristics of the analyte without using an AI algorithm will be referred to as "computational analysis" or "non-AI analysis." The representative values ​​processed in computational analysis have a smaller amount of data than the data input to the AI ​​algorithm in AI analysis. In computational analysis, the amount of data to be processed and the amount of calculation processing are smaller than in AI analysis, so the load on the computer performing the analysis is smaller than in AI analysis. This makes it possible to shorten the TAT (Turn Around Time) for analyzing measurement results.

[0022] According to the sample analyzer, sample analysis method, and program of embodiment 1, the analysis of data obtained by measuring samples can be divided between AI analysis and computational analysis, thereby reducing the load on the computer performing the analysis.

[0023] Figure 1 is a diagram schematically illustrating an example configuration of a sample analyzer 4000 according to embodiment 1. In Figure 1, the upper diagram illustrates the example configuration of sample analyzer 4000 according to the embodiment, and the lower diagram illustrates a modified example of the configuration of embodiment 1.

[0024] As shown in the upper diagram of Figure 1, sample analyzer 4000 of embodiment 1 includes, for example, a measurement unit 400 and an analysis unit 300. Alternatively, as shown in the lower diagram of Figure 1, sample analyzer 4000 may be configured with an integrated measurement unit 400 and analysis unit 300. Examples of sample analyzer 4000 include a blood cell analyzer, a urine analyzer, a blood coagulation analyzer, an immunoassay analyzer, a biochemistry analyzer, and a gene analyzer. Analytes to be analyzed by sample analyzer 4000 include, for example, cells, formed elements, proteins, and genes.

[0025] The measurement unit 400 measures a sample and acquires data related to the sample. The analysis unit 300 analyzes the data acquired by the measurement unit 400. The analysis unit 300 may have a function for setting measurement conditions for the measurement sample in the measurement unit 400 and controlling the execution of the measurement. The analysis unit 300 is configured as a separate device (e.g., a computer) from the measurement unit 400 and is connected to the measurement unit 400. The analysis unit 300 and the measurement unit 400 are connected by wire or wirelessly.

[0026] The measurement unit 400 includes an optical detection section for measuring a measurement sample prepared from a specimen.

[0027] The optical detection unit is, for example, a detection unit based on flow cytometry and is used to measure blood samples or urine samples. The optical detection unit acquires optical signals by irradiating light onto a measurement sample flowing through a flow cell. For example, the optical detection unit irradiates light onto a measurement sample containing an analyte (e.g., cells or formed elements) flowing through the flow cell, causing the analyte to emit forward scattered light, side scattered light, and fluorescence. A photodetector provided in the optical detection unit receives the emitted light and outputs an optical signal corresponding to the intensity of the received light. The optical signal is a waveform-shaped analog signal corresponding to the time changes in the forward scattered light, side scattered light, and fluorescence. An A / D conversion unit provided in the optical detection unit digitally converts the optical signal to obtain waveform-shaped digital data (hereinafter referred to as "waveform data") corresponding to each analyte. In this case, the waveform data is used, for example, to classify white blood cell types in a blood sample, classify the number of red blood cells and white blood cells in a blood sample, or classify formed elements in a urine sample.

[0028] The optical detection unit may be configured to irradiate a measurement sample contained in a container with light and detect light transmitted through or scattered by the measurement sample with a photodetector. In this case, the optical detection unit irradiates light onto the measurement sample containing an analyte, which is placed in a container and left stationary. The photodetector provided in the optical detection unit receives the transmitted light transmitted through the measurement sample or the scattered light generated by the measurement sample for a predetermined period of time and outputs an optical signal corresponding to the intensity of the received light. In this case, the optical signal is an analog signal with a waveform corresponding to the change over time in the transmitted light or scattered light associated with the clotting of the measurement sample. An A / D conversion unit provided in the optical detection unit digitally converts the optical signal and obtains digital waveform data (hereinafter referred to as "clotting waveform data") corresponding to the change over time in the transmitted light or scattered light. In this case, the clot waveform data is used, for example, to analyze the clotting ability of a blood sample.

[0029] Next, an example of analysis performed by the analysis unit 300 using data acquired by the measurement unit 400 will be described.

[0030] FIG. 2 is a diagram showing an outline of analysis when the optical detection unit is a detection unit based on flow cytometry.

[0031] 2, the diagram on the left shows an overview of the computational analysis, and the diagram on the right shows an overview of the AI ​​analysis. In Fig. 2, FSC, SSC, and FL represent optical signals corresponding to the forward scattered light intensity, side scattered light intensity, and fluorescence, respectively, acquired by the optical detection unit of the measurement unit 400.

[0032] The measurement unit 400 identifies a region of the digital data obtained by digitally converting the optical signal, where the region has values ​​greater than a predetermined threshold, as a region corresponding to the analyte in the sample, as shown in the upper graph of Figure 3. The region of the digital data obtained by digitally converting the optical signal, where the value is greater than the threshold, corresponds to each of the analytes in the sample. Each graph of Figure 3 schematically shows a region identified in the digital data corresponding to one analyte in the sample (e.g., the region of "waveform data" in the upper graph of Figure 3). Note that the region where the value is greater than the predetermined threshold may also be identified for the optical signal.

[0033] The measurement unit 400 acquires waveform data from digital data obtained by converting optical signals into digital data, each corresponding to an area of ​​an analyte in a sample. Waveform data is acquired corresponding to multiple analytes in the sample. The analysis unit 300 calculates representative values ​​of the waveform data corresponding to the characteristics of the analytes through computation. As shown in each graph in FIG. 3, the analysis unit 300 calculates quantities such as the peak value, width, and area of ​​the waveform data as the representative values. The peak value is the maximum value of the waveform data, the width is the width of the waveform data in the time axis direction, and the area is the area enclosed by the waveform data.

[0034] In the computational analysis, a representative value corresponding to the characteristic of the analyte is predetermined. For example, when classifying and counting blood cells as the analyte, the representative value predetermined in the algorithm of the computational analysis is a peak value. The analysis unit 300 acquires the predetermined representative value from the waveform data by a predetermined calculation and processes the acquired representative value to analyze the analyte. The analysis unit 300 acquires the predetermined representative value for each of the plurality of waveform data acquired by the measurement unit 400. In other words, the same type of representative value (e.g., peak value) is acquired from each of the plurality of waveform data by a predetermined calculation by the analysis unit 300. The predetermined representative value may be acquired by the measurement unit 400, and the acquired representative value and waveform data may be transmitted to the analysis unit 300.

[0035] On the other hand, in AI analysis, the AI ​​algorithm extracts the features of the waveform data, so the representative value is not determined in advance. The features of the waveform data extracted by the AI ​​algorithm (i.e., the features corresponding to the analyte) can change depending on the learning content of the AI ​​algorithm, so there is no need to determine the representative value in advance in AI analysis. Because the AI ​​algorithm can extract various features of the waveform data depending on the learning content, the waveform data itself, not just the representative value, is input to the AI ​​algorithm. Because the waveform data itself is input to the AI ​​algorithm, AI analysis places a higher load on the computer for calculating the data than computational analysis, and the TAT (Turn Around Time) required for the calculation is also longer.

[0036] As shown in the left diagram of FIG. 2, the analysis unit 300 acquires a representative value from the waveform data acquired corresponding to the analyte in the computational analysis, and generates, for example, a scattergram SC based on the acquired representative value. In the scattergram SC illustrated in FIG. 2, the SSCP on the horizontal axis represents the peak value of the waveform data based on side scattered light, and the FLP on the vertical axis represents the peak value of the waveform data based on fluorescence. Multiple analytes are plotted on the scattergram SC. The analysis unit 300 classifies and analyzes the analytes in the sample based on the scattergram SC.

[0037] As shown in the right diagram of Figure 2, in AI analysis, analysis unit 300 inputs waveform data corresponding to analytes into AI algorithm 60 to classify and analyze the analytes in the sample. AI algorithm 60 is a trained AI algorithm, and is generated by inputting the above-mentioned waveform data into a pre-trained AI algorithm and causing it to learn. The representative value obtained in the computational analysis has a smaller data volume than the waveform data input into the AI ​​algorithm in AI analysis.

[0038] The types of analytes classified by the computational analysis and AI analysis are, for example, the types of blood cells in a blood sample, the types of formed elements in a urine sample, etc. For example, the analysis unit 300 performs AI analysis on a measurement item that classifies the types of white blood cells in a blood sample, and performs computational analysis on other measurement items.

[0039] FIG. 4 is a diagram showing an outline of an analysis in which the optical detection unit is a detection unit that detects transmitted light or scattered light from a measurement sample.

[0040] The measurement unit 400 obtains digital data obtained by converting the optical signal into digital data as clot waveform data. In one measurement, one piece of clot waveform data is obtained from one measurement sample.

[0041] The graph in Figure 4 is an example of clot waveform data based on transmitted light detected when light is irradiated onto a measurement sample. The horizontal axis represents elapsed time, and the vertical axis represents absorbance. Absorbance is a value that indicates the degree to which light irradiated onto the measurement sample is absorbed by the measurement sample. An absorbance of 0% indicates that almost all of the light irradiated onto the measurement sample has reached the photodetector, and an absorbance of 100% indicates that almost none of the light irradiated onto the measurement sample has reached the photodetector.

[0042] Note that transmitted light intensity may be used instead of absorbance. In this case, if the vertical axis ratio (transmitted light intensity) is set to increase upward, the clot waveform data will have a shape that decreases over time, similar to Figure 4.

[0043] The clot waveform data includes at least data corresponding to the optical signal acquired from time T2, which indicates the start of clotting of the sample, to time T3, which indicates the end of clotting of the sample. The clot waveform data may also include data corresponding to the optical signal acquired from time T1, which indicates the start of photometry by the measurement unit 400, to time T4, which indicates the end of photometry.

[0044] In the computational analysis, the analysis unit 300 calculates a representative value corresponding to the characteristics of the analyte from the clot waveform data by computational processing, and performs analysis based on the calculated representative value. In the computational analysis, the analysis unit 300 identifies the clot waveform data when the detected light intensity satisfies a predetermined condition as the representative value. For example, the analysis unit 300 obtains the time (T-T2) required for the absorbance of the clot waveform data to decrease to a predetermined value (e.g., 50%) as the representative value, and provides the obtained representative value as a result indicating the time until the blood sample clots.

[0045] In the AI ​​analysis, the analysis unit 300 analyzes the clot waveform data based on an AI algorithm 60 (see FIG. 2). The analysis unit 300 acquires the presence or absence of an abnormality in the measurement based on, for example, feature amounts extracted from the clot waveform data by the AI ​​algorithm 60. The analysis unit 300 determines whether a non-specific reaction is suspected based on the presence or absence of an abnormality in the measurement.

[0046] For example, the analysis unit 300 analyzes the presence or absence of abnormalities caused by interfering substances in a blood sample. As a specific example, the analysis unit 300 analyzes the presence or absence of abnormalities using clot waveform data related to PT (prothrombin time), which is an item for measuring the coagulation ability related to prothrombin, a coagulation factor.

[0047] In addition, in AI analysis, the analysis unit 300 may input the clot waveform data into the AI ​​algorithm 60 to obtain the time until the blood sample clots. In addition, in AI analysis, the analysis unit 300 may input the clot waveform data into the AI ​​algorithm 60 to obtain the cause of the prolongation of the clotting time if the clotting time is prolonged.

[0048] FIG. 5 is a flowchart showing an example of the sample analyzing method of the first embodiment.

[0049] In step S1, the measurement unit 400 acquires an optical signal by the optical detection section, and obtains waveform data from the acquired optical signal.

[0050] In step S2, the analysis unit 300 performs AI analysis on waveform data (first data) that is the target of AI analysis from among the waveform data acquired by the measurement unit 400. For example, the analysis unit 300 identifies waveform data corresponding to the measurement item that is the target of AI analysis as the first data, and performs AI analysis on the identified first data.

[0051] In step S3, the analysis unit 300 performs a computational analysis on the waveform data (second data) that is the target of the computational analysis from the waveform data acquired by the measurement unit 400. For example, the analysis unit 300 identifies the waveform data corresponding to the measurement item that is the target of the computational analysis as the second data, and performs a computational analysis on the identified second data.

[0052] In the above steps S2 and S3, an example will be described in which a measurement for classifying the types of white blood cells in a blood sample is the subject of AI analysis. The measurement unit 400, for example, prepares a blood sample with a reagent corresponding to the measurement of white blood cell differentiation and measures the prepared measurement sample using an optical detection unit based on flow cytometry. Since the measurement related to white blood cell differentiation is the subject of AI analysis, the analysis unit 300 identifies waveform data based on the measurement sample of white blood cell differentiation as the first data. The analysis unit 300 analyzes the first data using the AI ​​algorithm 60 and classifies the white blood cells. Meanwhile, the analysis unit 300 identifies waveform data based on a measurement sample other than the white blood cell differentiation as the second data. The analysis unit 300 identifies a representative value corresponding to the characteristics of the analyte from the second data, performs computational analysis to process the identified representative value, and classifies blood cells other than white blood cells.

[0053] In step S4, the analysis unit 300 provides the analysis results acquired in steps S2 and S3. In step S4, for example, the analysis unit 300 displays the analysis results on a display unit and transmits the analysis results to another computer.

[0054] In step S1, the measurement unit 400 may acquire an optical signal from one measurement sample using an optical detection unit, and acquire waveform data from the acquired optical signal. In this case, the first data and the second data may each be a plurality of pieces of data, and some of the data may be the same as each other.

[0055] Alternatively, in step S1, an optical detection unit may acquire an optical signal from each of a plurality of measurement samples containing specimens collected from the same subject, and waveform data may be acquired from each of the acquired optical signals. In this case, the analysis unit 300 performs AI analysis on waveform data (first data) acquired from one measurement sample in step S2, and performs computational analysis on waveform data (second data) acquired from another measurement sample in step S3. The plurality of measurement samples containing specimens collected from the same subject may be prepared using the same type of reagent or different types of reagents.

[0056] Alternatively, in step S1, an optical detection unit may acquire an optical signal from each of a plurality of measurement samples containing specimens collected from different subjects, and waveform data may be acquired from each of the acquired optical signals. In this case, the analysis unit 300 performs AI analysis on waveform data (first data) acquired from one measurement sample in step S2, and performs computational analysis on waveform data (second data) acquired from another measurement sample in step S3. The plurality of measurement samples containing specimens collected from different subjects may be prepared using the same type of reagent or different types of reagent.

[0057] In the first embodiment, the analysis unit 300 identifies a representative value corresponding to the characteristic of the analyte from the second data and processes the identified representative value in step S3 as the computational analysis, but this is not limiting. For example, in step S1, the measurement unit 400 may obtain a representative value from the waveform data and output the waveform data and the representative value to the analysis unit 300, and in step S3 as the computational analysis, the analysis unit 300 may process the representative value obtained from the measurement unit 400.

[0058] [Embodiment 2] In the second embodiment, the AI ​​analysis and the computational analysis are selected based on rules set in the analysis unit 300.

[0059] The rules for selecting the analysis operation are set by the user via the analysis unit 300, for example. The user can set rules in the analysis unit 300 according to the operation policy of the laboratory, for example. This makes it possible to appropriately change the division of labor between AI analysis and computational analysis according to the operation policy of the laboratory.

[0060] The ability to set rules for analysis operations makes it possible to flexibly change the division of labor between AI analysis and computational analysis while reducing the load on the analysis unit 300. For example, if the accuracy of AI analysis is improved by having the AI ​​algorithm 60 undergo additional learning, it becomes possible to set rules so that more data is subject to AI analysis. Also, for example, if prioritizing shortening the TAT (Turn Around Time) for analyzing measurement results, it becomes possible to set rules so that more data is subject to computational analysis.

[0061] FIG. 6 is a flowchart showing an example of setting an analysis operation based on rules set in the analysis unit 300.

[0062] In step S11, the measurement unit 400 acquires an optical signal using the optical detection unit and acquires waveform data from the acquired optical signal. In step S12, the analysis unit 300 refers to rules for selecting an analysis operation and, based on the referred rules, identifies waveform data to be subjected to AI analysis and computational analysis from the waveform data acquired in step S11.

[0063] In step S13, the analysis unit 300 determines whether the waveform data identified in step S12 includes waveform data that is the target of AI analysis. If the waveform data includes waveform data that is the target of AI analysis (S12: YES), in step S14, the analysis unit 300 performs AI analysis on the waveform data that is the target of AI analysis identified in step S12.

[0064] Next, in step S15, the analysis unit 300 determines whether there is any waveform data to be subjected to computational analysis in addition to the waveform data that has been subjected to AI analysis. If there is any waveform data to be subjected to computational analysis (S15: YES), in step S16, the analysis unit 300 performs computational analysis on the waveform data to be subjected to computational analysis identified in step S12.

[0065] In step S12, the measurement unit 400 may acquire both waveform data to be subjected to AI analysis and waveform data to be subjected to computational analysis. For example, when the measurement unit 400 performs a measurement related to white blood cell differentiation and a measurement related to reticulocytes in accordance with the measurement order, it acquires waveform data for white blood cell differentiation and waveform data for reticulocyte measurement. When the white blood cell differentiation is the target of AI analysis and the reticulocyte measurement is the target of computational analysis, the analysis unit 300 determines that the waveform data includes waveform data for white blood cell differentiation, which is the target of AI analysis (S13: YES), and performs AI analysis on the waveform data. Furthermore, the analysis unit 300 determines that the waveform data includes waveform data for reticulocyte differentiation, which is the target of computational analysis (S15: YES), and performs computational analysis on the waveform data.

[0066] On the other hand, if the waveform data acquired by the measurement unit 400 does not include waveform data that is the target of AI analysis (S13: NO), in step S16, the analysis unit 300 performs a computational analysis on the waveform data that is the target of computational analysis identified in step S12. Also, if the AI ​​analysis is performed and no waveform data that is the target of computational analysis is included (S15: NO), the computational analysis is not performed and the process proceeds to step S17.

[0067] In step S17, the analysis unit 300 provides the analysis result.

[0068] Alternatively, the analysis unit 300 may determine whether or not waveform data to be subjected to computational analysis is included in step S13, and then determine whether or not waveform data to be subjected to AI analysis is included in step S15. In this case, if the analysis unit 300 determines in step S13 that waveform data to be subjected to computational analysis is included, it performs computational analysis in step S14. Furthermore, if the analysis unit 300 determines in step S15 that waveform data to be subjected to AI analysis is included, it performs AI analysis in step S16.

[0069] [Embodiment 3] In the third embodiment, various examples in which AI analysis and computational analysis are shared will be described.

[0070] For example, the division of roles between AI analysis and computational analysis is determined by a software program used by analysis unit 300 to analyze waveform data. The software program of analysis unit 300 identifies waveform data to be subjected to AI analysis and waveform data to be subjected to computational analysis, and performs the analysis. The software program is designed, for example, according to requirements related to the inspection (e.g., improving TAT, increasing analytical accuracy).

[0071] FIG. 7 is a flowchart showing an example of how analysis is performed according to measurement items.

[0072] In Fig. 7, steps S21, S22, and S23 are added instead of steps S12, S13, and S15, respectively, compared to Fig. 6. The changes from Fig. 6 will be described below.

[0073] In step S21, the analysis unit 300 refers to rules that include whether to perform AI analysis or computational analysis based on the measurement items, and based on the rules that it refers to, it identifies the waveform data of the measurement items that are the subject of AI analysis and the waveform data of the measurement items that are the subject of computational analysis for the waveform data acquired in step S11.

[0074] 8 is an exemplary diagram showing a screen for setting AI analysis or computer processing analysis for each measurement item. The measurement items shown in FIG. 8 are related to a blood cell analyzer.

[0075] The screen of FIG. 8 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of FIG. 8 has, for each measurement item, a check box for setting AI analysis and a check box for setting computational analysis. The AI ​​analysis check box and the computational analysis check box for one measurement item are configured so that only one of them can be selected. The user operates the check box to select whether to perform AI analysis or computational analysis for each measurement item, and operates the setting button. This causes the rule to be stored in the memory unit of the analysis unit 300.

[0076] Although the user selects either AI analysis or computer-generated analysis for each measurement item via the screen shown in FIG. 8, the screen may be configured to allow both AI analysis and computer-generated analysis to be selected. This allows the results of the AI ​​analysis to be compared with the results of the computer-generated analysis. The selection of analysis for each measurement item may be set in advance when the device is shipped, or may be changeable only by an administrator.

[0077] 7, in step S22, the analysis unit 300 determines whether the waveform data identified in step S21 includes waveform data of a measurement item that is the target of AI analysis. If the waveform data of a measurement item that is the target of AI analysis is included (S22: YES), in step S14, the analysis unit 300 performs AI analysis on the waveform data of the measurement item that is the target of AI analysis identified in step S21.

[0078] For example, in accordance with a measurement order to classify white blood cells (e.g., five categories: neutrophils, lymphocytes, monocytes, eosinophils, and basophils), the measurement unit 400 mixes a sample with a reagent corresponding to the classification to prepare a white blood cell measurement sample. The measurement unit 400 acquires an optical signal corresponding to the white blood cell measurement sample using an optical detection unit. The measurement unit 400 acquires waveform data corresponding to the acquired optical signal. When measurement items related to the classification of white blood cells (e.g., the count and percentage of each of neutrophils, lymphocytes, monocytes, eosinophils, and basophils) are the subject of AI analysis, the analysis unit 300 performs AI analysis on the waveform data acquired by the measurement unit 400 when measuring the white blood cell measurement sample.

[0079] Next, in step S23, the analysis unit 300 determines whether or not the waveform data identified in step S12 includes waveform data of a measurement item that is the target of computational analysis. If there is waveform data that is the target of computational analysis (S23: YES), in step S16, the analysis unit 300 performs computational analysis of the measurement item on the waveform data that is the target of computational analysis identified in step S21.

[0080] For example, in accordance with a measurement order for classifying reticulocytes, the measurement unit 400 mixes a sample with a reagent corresponding to the classification to prepare a reticulocyte measurement sample. The measurement unit 400 acquires an optical signal corresponding to the reticulocyte measurement sample using an optical detection unit. The measurement unit 400 acquires waveform data corresponding to the acquired optical signal. When a measurement item related to the classification of reticulocytes (e.g., the count and percentage of reticulocytes) is the subject of computational analysis, the analysis unit 300 performs computational analysis on the waveform data acquired by the measurement unit 400 through measurement of the reticulocyte measurement sample.

[0081] The sample analyzer 4000 is not limited to being a blood cell analyzer, but may also be a urine analyzer or a blood coagulation measuring device. For example, if the sample analyzer 4000 is a urine analyzer, the analysis unit 300 performs AI analysis on some of the measurement items and performs computer-based analysis on the remaining measurement items. If the sample analyzer 4000 is a blood coagulation measuring device, the analysis unit 300 performs computer-based analysis on all of the measurement items, and for some of the measurement items, performs AI analysis in addition to computer-based analysis to determine whether a non-specific reaction is suspected.

[0082] FIG. 9 is a flowchart showing an example in which analysis is performed in accordance with a measurement order.

[0083] In Fig. 9, steps S31 and S32 are added instead of steps S12 and S13, respectively, and step S15 is deleted, compared to Fig. 6. The changes from Fig. 6 will be described below.

[0084] In step S31, the analysis unit 300 identifies, based on the measurement order, whether the waveform data acquired in step S11 is waveform data to be subjected to AI analysis or waveform data to be subjected to computational analysis. The analysis mode for the measurement order is either the AI ​​analysis mode or the computational analysis mode, and is stored in the memory of the analysis unit 300 in association with the measurement order.

[0085] FIG. 10 is an exemplary view that schematically shows a screen for setting an analysis mode for a measurement order.

[0086] The screen of FIG. 10 is displayed, for example, on a display unit provided in the analysis unit 300. In the screen of FIG. 10, each row corresponds to a measurement order identified by a specimen number. The screen of FIG. 10 includes a check box for setting the AI ​​analysis mode and a check box for setting the computational analysis mode for each measurement order. The user operates the check box to select whether the analysis unit 300 will perform AI analysis or computational analysis for each measurement order, and operates the setting button. As a result, the analysis mode is stored in the memory unit of the analysis unit 300 in association with the measurement order.

[0087] The analysis mode associated with each measurement order is not limited to being set by the user via the analysis unit 300, but may be set in advance by a host computer or the like when setting the measurement order.

[0088] 9, in step S32, the analysis unit 300 determines whether the waveform data identified in step S31 is waveform data of the measurement order that is the target of AI analysis. If the identified waveform data is waveform data of the measurement order that is the target of AI analysis (S32: YES), in step S14, the analysis unit 300 performs AI analysis on the waveform data of the measurement order. On the other hand, if the identified waveform data is waveform data of the measurement order that is the target of computational analysis (S32: NO), in step S16, the analysis unit 300 performs computational analysis on the waveform data of the measurement order.

[0089] FIG. 11 is a flowchart showing an example of how analysis is performed depending on the analysis mode of the device.

[0090] In Fig. 11, steps S41 and S42 are added instead of steps S12 and S13, respectively, and step S15 is deleted, compared to Fig. 6. The changes from Fig. 6 will be described below.

[0091] In step S41, the analysis unit 300 refers to rules including the analysis mode of the analysis unit 300, and determines, based on the referred rules, whether the waveform data acquired in step S11 is waveform data to be subjected to AI analysis or waveform data to be subjected to computational analysis. If the AI ​​analysis mode is set in the above rules, all waveform data is subjected to AI analysis, and if the computational analysis mode is set in the above rules, all data is subjected to computational analysis.

[0092] FIG. 12 is an exemplary diagram that schematically shows a screen for setting the analysis mode of the analysis unit 300. As shown in FIG.

[0093] 12 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of FIG. 12 includes a check box for setting the AI ​​analysis mode and a check box for setting the computational analysis mode for the analysis unit 300. The user operates the check box to select whether the analysis unit 300 will perform AI analysis or computational analysis, and operates the setting button. This causes the rule to be stored in the memory unit of the analysis unit 300.

[0094] 11, in step S42, the analysis unit 300 determines whether the waveform data identified in step S41 is waveform data that is the target of AI analysis. If the identified waveform data is data that is the target of AI analysis (S42: YES), that is, if the analysis mode of the analysis unit 300 is the AI ​​analysis mode, in step S14, the analysis unit 300 performs AI analysis on the waveform data. On the other hand, if the identified waveform data is waveform data that is the target of computational analysis (S42: NO), that is, if the analysis mode of the analysis unit 300 is the computational analysis mode, in step S16, the analysis unit 300 performs computational analysis on the waveform data.

[0095] FIG. 13 is a flowchart showing an example in which analysis is performed depending on the type of measurement order.

[0096] In Fig. 13, steps S51 and S52 are added instead of steps S12 and S13, respectively, and step S15 is deleted, compared to Fig. 6. The changes from Fig. 6 will be described below.

[0097] In step S51, the analysis unit 300 refers to rules including an analysis mode corresponding to the type of measurement order, and determines whether the waveform data acquired in step S11 is waveform data to be subjected to AI analysis or computational analysis based on the type of measurement order and the referred rule. Measurement order types include "Normal," which corresponds to a normal measurement such as an initial test, "Rerun," which corresponds to a retest in which the same measurement items as the initial test are set, and "Reflex," which corresponds to a retest in which the measurement items have been changed from the initial test. The above rules set either the AI ​​analysis mode or the computational analysis mode for each type of measurement order.

[0098] FIG. 14 is an exemplary view showing a screen for setting an analysis mode for each type of measurement order.

[0099] The screen of Fig. 14 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of Fig. 14 includes check boxes for setting the AI ​​analysis mode for each type of measurement order (Normal, Rerun, Reflex) and a check box for setting the computational analysis mode. The user operates the check boxes to select whether to perform AI analysis or computational analysis for each type of measurement order, and operates the setting button. This causes the rule to be stored in the memory unit of the analysis unit 300.

[0100] The analysis mode associated with each type of measurement order is not limited to being set by the user via the analysis unit 300, but may be set in advance in a host computer or the like according to the type of measurement order.

[0101] 13, in step S52, the analysis unit 300 determines whether the waveform data identified in step S51 is waveform data to be subjected to AI analysis. If the identified waveform data is waveform data to be subjected to AI analysis (S52: YES), i.e., if the analysis mode corresponding to the type of measurement order is AI analysis mode, the analysis unit 300 performs AI analysis on the waveform data in step S14. On the other hand, if the identified waveform data is waveform data to be subjected to computational analysis (S52: NO), i.e., if the analysis mode corresponding to the type of measurement order is computational analysis mode, the analysis unit 300 performs computational analysis on the waveform data in step S16.

[0102] FIG. 15 is a flowchart showing an example in which analysis is performed according to the type of measurement item and measurement order.

[0103] In Figure 15, step S61 is added instead of step S12 compared to Figure 6. The changes from Figure 6 will be described below.

[0104] In step S61, the analysis unit 300 refers to the rules for selecting the analysis operation and, based on the measurement item and the type of measurement order, identifies the waveform data acquired in step S11 as the target of AI analysis and the waveform data as the target of computational analysis.

[0105] FIG. 16 is an exemplary diagram that schematically shows a screen for setting AI analysis or computational analysis for each measurement item and measurement order type.

[0106] The screen of FIG. 16 is displayed, for example, on a display unit provided in the analysis unit 300. As with FIG. 8, the screen of FIG. 16 includes a check box for setting either AI analysis or computational analysis for each measurement item, and a check box for setting computational analysis, and as with FIG. 14, includes a check box for setting either AI analysis or computational analysis for each measurement order type (Normal, Rerun, Reflex). The user operates the check boxes in the upper list to select whether AI analysis or computational analysis will be performed for each measurement item, operates the check boxes in the lower list to select whether AI analysis or computational analysis will be performed for each measurement order type, and then operates the setting button. This causes the rules to be stored in the memory unit of the analysis unit 300.

[0107] When the settings are made as shown in FIG. 16, if the type of measurement order is "Normal," the analysis unit 300 specifies the waveform data acquired by the measurement unit 400 as the target for computational analysis. For example, if the type of measurement order is "Normal," computational analysis is performed on all measurement items based on the measurement order, regardless of the analysis settings for each measurement item. Also, if the type of measurement order is "Rerun" or "Reflex," the analysis unit 300 sets the waveform data acquired by the measurement unit 400 as the target for AI analysis or computational analysis, depending on the analysis settings set for each measurement item. For example, if the type of measurement order is "Rerun" or "Reflex," the measurement items related to nucleated red blood cells (NRBC) and basophils (BASO) are the target of AI analysis, and the other measurement items are the target of computational analysis.

[0108] 15, in step S13, the analysis unit 300 determines whether the waveform data identified in step S61 includes data that is the target of AI analysis. If there is waveform data that is the target of AI analysis (S13: YES), in step S14, the analysis unit 300 performs AI analysis on the waveform data that is the target of AI analysis identified in step S61.

[0109] Subsequently, in step S15, the analysis unit 300 determines whether the waveform data identified in step S61 includes waveform data to be subjected to computational analysis. If there is waveform data to be subjected to computational analysis (S15: YES), in step S16, the analysis unit 300 performs computational analysis on the waveform data to be subjected to computational analysis identified in step S61.

[0110] FIG. 17 is a flowchart showing an example in which it is determined whether or not an AI analysis is necessary based on a flag generated by a computational analysis.

[0111] In Figure 17, steps S71 to S74 are added after step S11, and steps S12 to S16 are deleted, compared to Figure 6. The changes from Figure 6 will be described below.

[0112] In step S71, the analysis unit 300 performs a computational analysis on the waveform data acquired in step S11, and sets a flag indicating an abnormality in the analyte in the sample based on the results of the computational analysis. The flag may be, for example, a flag indicating that a predetermined abnormal cell has been detected, or a flag indicating that the count value of a predetermined blood cell is an abnormal value. In step S72, the analysis unit 300 refers to a rule including whether or not to perform AI analysis on the analysis result of the flag.

[0113] FIG. 18 is an exemplary diagram illustrating a screen for setting AI analysis for each flag of the analysis result.

[0114] The screen of FIG. 18 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of FIG. 18 includes a check box for setting AI analysis for each flag assigned to the analysis results by the computerized analysis. If the sample analyzer 4000 is a blood cell analyzer, flags assigned to the analysis results by the computerized analysis include a decrease or increase in blood cells, and the appearance of abnormal cells. The user operates the check box to select whether to perform AI analysis for each flag, and then operates the setting button. This causes the rules to be stored in the memory unit of the analysis unit 300.

[0115] When a check box for a flag is checked, AI analysis is performed on the waveform data corresponding to that analysis result. In the example shown in Figure 18, the check boxes for the analysis results of blasts / abnormal lymphocytes, blasts, abnormal lymphocytes, and atypical lymphocytes are checked, and if the flag indicating the presence of these blood cells is set by the computational analysis, AI analysis is performed on these blood cells.

[0116] 17, in step S73, the analysis unit 300 determines whether the sample is a target for AI analysis based on the flag assigned to the analysis result obtained in step S71 and the rules referenced in step S72. If the sample is a target for AI analysis (S73: YES), in step S74, the analysis unit 300 performs AI analysis on each waveform data. For example, if a flag indicating that blast cells have been detected in the computational analysis is issued, AI analysis is performed on each waveform data obtained in step S11 according to the rules illustrated in FIG.

[0117] On the other hand, if the sample is not a target for AI analysis (S73: NO), the analysis unit 300 skips step S74.

[0118] FIG. 19 is a flow chart illustrating an example of performing AI analysis on specific analytes classified in a computational analysis.

[0119] In Figure 19, steps S81 to S84 are added after step S11, and steps S12 to S16 are deleted, compared to Figure 6. The changes from Figure 6 will be described below.

[0120] In step S81, the analysis unit 300 performs a computational analysis on the waveform data acquired in step S11 to classify the analytes. In step S82, the analysis unit 300 refers to rules including whether to perform AI analysis for each type of analyte.

[0121] FIG. 20 is an exemplary diagram that schematically shows a screen for setting whether or not to perform AI analysis for each type of analyte.

[0122] The screen of FIG. 20 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of FIG. 20 includes check boxes for setting AI analysis for each type of analyte. If the sample analyzer 4000 is a blood cell analyzer, the types classified by computational analysis include eosinophils, neutrophils, lymphocytes, and monocytes. The user operates the check boxes to select whether or not to perform AI analysis for each type of analyte, and then operates the setting button. This causes the rules to be stored in the memory unit of the analysis unit 300.

[0123] When a check box for a type of analyte is checked, AI analysis is performed on the waveform data classified into that type. In the example shown in Figure 20, AI analysis is performed on monocytes and lymphocytes.

[0124] 19, in step S83, the analysis unit 300 identifies waveform data corresponding to the analytes classified into a specific classification (e.g., monocytes and lymphocytes in the case of the rules shown in FIG. 20) based on the type of analyte classified in step S81 and the rule referenced in step S82. In step S84, the analysis unit 300 performs AI analysis on the identified waveform data.

[0125] FIG. 21 is a diagram illustrating a classification method using computational analysis and AI analysis, which is executed in the process shown in FIG.

[0126] In the computational analysis, a scattergram is used, with two representative values ​​calculated from the waveform data as axes. For example, as shown in FIG. 20, when AI analysis is set to be performed on monocytes and lymphocytes, in step S81, the computational analysis identifies plots in the dashed boxed regions on the scattergram corresponding to the monocytes and lymphocytes. Then, in step S83, waveform data corresponding to the plots in the boxed regions is identified, and in step S84, AI analysis is performed on the identified waveform data.

[0127] 19, in step S17, the analysis unit 300 provides the analysis results obtained by the computational analysis and the AI ​​analysis. At this time, the analysis unit 300 replaces the analysis results of the type that was the target of the AI ​​analysis among the analysis results obtained by the computational analysis in step S81 with the analysis results obtained by the AI ​​analysis and provides them. Note that the analysis results obtained by the computational analysis and the analysis results obtained by the AI ​​analysis may be provided together.

[0128] FIG. 22 is a flow chart illustrating an example of performing AI analysis when a particular classification is made in the computational analysis.

[0129] In Figure 22, steps S91 to S95 are added after step S11, and steps S12 to S16 are deleted, compared to Figure 6. The changes from Figure 6 will be described below.

[0130] In step S91, the analysis unit 300 performs computational analysis on the waveform data acquired in step S11 to classify the analytes. In step S92, the analysis unit 300 refers to rules including whether to perform AI analysis on a particular type of analyte, for example, cells not present in the peripheral blood of healthy individuals.

[0131] FIG. 23 is an exemplary diagram that schematically shows a screen for setting whether or not to perform AI analysis on a specific type of analyte.

[0132] The screen of FIG. 23 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of FIG. 23 includes check boxes for setting AI analysis for each specific analyte type. If the sample analyzer 4000 is a blood cell analyzer, the types classified by computational analysis may include blast cells, abnormal lymphocytes, atypical lymphocytes, and immature granulocytes. The user operates the check boxes to select whether or not to perform AI analysis for each specific analyte type, and then operates the setting button. This causes the rules to be stored in the memory unit of the analysis unit 300.

[0133] When a checkbox for a specific analyte type is checked, AI analysis is performed on the waveform data classified into that type. In the example shown in Figure 23, AI analysis is performed on blast cells, abnormal lymphocytes, and atypical lymphocytes.

[0134] Returning to FIG. 22, in step S93, analysis unit 300 determines whether or not an analyte classified into a specific category (e.g., blast cells, abnormal lymphocytes, and atypical lymphocytes in the case of the rules shown in FIG. 23) has been detected based on the type of analyte classified in step S91 and the rules referenced in step S92. If an analyte classified into a specific category has been detected (S93: YES), in step S94 analysis unit 300 identifies waveform data corresponding to the analyte classified into the specific category. In step S95, analysis unit 300 performs AI analysis on the identified waveform data.

[0135] In step S17, the analysis unit 300 provides the analysis results obtained by the computational analysis and the AI ​​analysis. At this time, the analysis unit 300 replaces the analysis results of the type that was the target of the AI ​​analysis among the analysis results obtained by the computational analysis in step S91 with the analysis results obtained by the AI ​​analysis and provides them. Note that the analysis results obtained by the computational analysis and the analysis results obtained by the AI ​​analysis may be provided together.

[0136] [Embodiment 4] In the fourth embodiment, a detailed configuration example is shown in which a sample analyzer 4000 that analyzes samples based on flow cytometry performs computational analysis and AI analysis in a shared manner.

[0137] Examples of samples measured by the sample analyzer 4000 of embodiment 4 include biological samples collected from a subject. Samples may include, for example, peripheral blood, such as venous blood and arterial blood, urine, and body fluids other than blood and urine. Body fluids other than blood and urine may include, for example, bone marrow fluid, ascites, pleural effusion, and cerebrospinal fluid. Hereinafter, body fluids other than blood and urine may be simply referred to as "body fluids." There are no limitations on the blood sample, as long as it is in a state that allows cell counting and cell type determination. The blood is preferably peripheral blood. For example, the blood may be peripheral blood collected using an anticoagulant such as ethylenediaminetetraacetic acid (sodium salt or potassium salt) or heparin sodium. Peripheral blood may be collected from an artery or a vein.

[0138] The cell type to be determined in this embodiment is based on a morphological classification and varies depending on the type of specimen. When the specimen is blood and the blood is collected from a healthy individual, the cell type to be determined in this embodiment includes, for example, nucleated cells such as nucleated red blood cells and white blood cells, red blood cells, and platelets. Nucleated cells include, for example, neutrophils, lymphocytes, monocytes, eosinophils, and basophils. Neutrophils include, for example, segmented neutrophils and band-shaped neutrophils. When the blood is collected from a non-healthy individual, the nucleated cells may include, for example, at least one type selected from the group consisting of immature granulocytes and abnormal cells. Such cells are also included in the cell type to be determined in this embodiment. Immature granulocytes may include, for example, metamyelocytes, myelocytes, promyelocytes, myeloblasts, and other cells.

[0139] In addition to normal cells, nucleated cells may also include abnormal cells that are not found in the peripheral blood of healthy individuals. Examples of abnormal cells include cells that appear when a patient is affected with a specific disease, such as tumor cells. In the case of the hematopoietic system, the specific disease may be, for example, a disease selected from the group consisting of leukemias such as myelodysplastic syndrome, acute myeloblastic leukemia, acute promyelocytic leukemia, acute myelomonocytic leukemia, acute monocytic leukemia, erythroleukemia, acute megakaryoblastic leukemia, acute myeloid leukemia, acute lymphocytic leukemia, lymphoblastic leukemia, chronic myelogenous leukemia, or chronic lymphocytic leukemia, malignant lymphomas such as Hodgkin's lymphoma and non-Hodgkin's lymphoma, and multiple myeloma.

[0140] Furthermore, abnormal cells may include cells not normally found in the peripheral blood of healthy individuals, such as lymphoblasts, plasma cells, atypical lymphocytes, reactive lymphocytes, proerythroblasts, basophilic erythroblasts, polychromatic erythroblasts, normochromatic erythroblasts, promegaloblasts, basophilic megaloblasts, polychromatic megaloblasts, and normochromatic megaloblasts; and megakaryocytes, including micromegakaryocytes.

[0141] Furthermore, when the specimen is urine, the cell types to be determined in this embodiment may include, for example, epithelial cells such as transitional epithelium and squamous epithelium, red blood cells, white blood cells, etc. Abnormal cells may include, for example, bacteria, mold, fungi such as yeast, tumor cells, etc.

[0142] When the specimen is a body fluid that does not normally contain blood components, such as ascites, pleural effusion, or cerebrospinal fluid, the cell types may include, for example, red blood cells, white blood cells, and large cells. Large cells referred to here refer to cells that are detached from the lining of body cavities or the peritoneum of internal organs and are larger than white blood cells, such as mesothelial cells, histiocytes, and tumor cells.

[0143] When the specimen is bone marrow fluid, the cell types to be determined in this embodiment may include mature blood cells and immature blood cells as normal cells. Mature blood cells include, for example, nucleated cells such as nucleated red blood cells and white blood cells, red blood cells, and platelets. Nucleated cells such as white blood cells include, for example, neutrophils, lymphocytes, plasma cells, monocytes, eosinophils, and basophils. Neutrophils include, for example, segmented neutrophils and band-shaped neutrophils. Immature blood cells include, for example, hematopoietic stem cells, immature granulocytic cells, immature lymphocytic cells, immature monocytic cells, immature erythroid cells, megakaryocytic cells, and mesenchymal cells. Immature granulocytes may include, for example, metamyelocytes, myelocytes, promyelocytes, and myeloblasts. Immature lymphocytic cells include, for example, lymphoblasts. Immature monocytic cells include monoblasts, etc. Immature erythroid cells include nucleated erythrocytes such as proerythroblasts, basophilic erythroblasts, polychromatic erythroblasts, normochromatic erythroblasts, promegaloblasts, basophilic megaloblasts, polychromatic megaloblasts, and normochromatic megaloblasts, etc. Megakaryocytic cells include megakaryoblasts, etc.

[0144] Abnormal cells that may be contained in bone marrow include, for example, hematopoietic tumor cells selected from the group consisting of leukemias such as the above-mentioned myelodysplastic syndrome, acute myeloblastic leukemia, acute promyelocytic leukemia, acute myelomonocytic leukemia, acute monocytic leukemia, erythroleukemia, acute megakaryoblastic leukemia, acute myeloid leukemia, acute lymphocytic leukemia, lymphoblastic leukemia, chronic myelogenous leukemia, and chronic lymphocytic leukemia; malignant lymphomas such as Hodgkin's lymphoma and non-Hodgkin's lymphoma; and multiple myeloma; and metastatic tumor cells of malignant tumors that have developed in organs other than the bone marrow.

[0145] The signals obtained from the analytes (e.g., cells or formed elements) in the sample exemplified above are forward scattered light signals, side scattered light signals, and fluorescent signals, which are analog optical signals obtained by irradiating light onto cells flowing through a flow cell, but there are no particular limitations as long as the signals represent the characteristics of the analytes and can be used to classify the analytes by type.

[0146] The optical signal is preferably an optical signal obtained as an optical response when a cell is irradiated with light, and may include at least one signal selected from a signal based on light scattering, a signal based on light absorption, a signal based on transmitted light, and a signal based on fluorescence.

[0147] The signal based on light scattering may include a scattered light signal generated by light irradiation and a light loss signal generated by light irradiation. The scattered light signal represents a characteristic of the analyte in the sample depending on the angle at which the scattered light is received relative to the traveling direction of the irradiated light. The forward scattered light signal is used to calculate a representative value representing the size of the analyte. When the analyte in the sample is a cell, the side scattered light signal is used to calculate a representative value representing the complexity of the cell's nucleus.

[0148] The term "forward" in forward scattered light refers to the direction of travel of light emitted from the light source. "Forward" can include a low forward angle where the light-receiving angle is approximately 0° to 5°, and / or a high forward angle where the light-receiving angle is approximately 5° to 20°, assuming the angle of the irradiating light is 0°. "Side" is not limited as long as it does not overlap with "forward." "Side" can include a light-receiving angle of approximately 25° to 155°, preferably approximately 45° to 135°, and more preferably approximately 90°, assuming the angle of the irradiating light is 0°. The fluorescence in this embodiment is detected at the same light-receiving angle as the side scattered light.

[0149] A signal based on light scattering may include polarization or depolarization as a signal component. For example, by irradiating an analyte in a sample with light and receiving the resulting scattered light through a polarizing plate, only scattered light polarized at a specific angle can be received. Alternatively, by irradiating an analyte in a sample with light through a polarizing plate and receiving the resulting scattered light through a polarizing plate that transmits only light polarized at an angle different from the polarizing plate used for irradiation, only depolarized scattered light can be received.

[0150] The light loss signal represents the loss in the amount of received light due to the reduction in the amount of light received at the light receiving unit when light is irradiated and scattered by the analyte. The light loss signal is preferably obtained as the light loss in the optical axis direction of the irradiated light (axial light loss). The light loss signal can be expressed as the ratio of the amount of light received when the measurement sample flows through the flow cell to the amount of light received at the light receiving unit when the measurement sample is not flowing through the flow cell, which is taken as 100%. Like the forward scattered light signal, the axial light loss is used to calculate a representative value representing the size of the analyte, but the signal obtained differs depending on whether the cell is translucent or not.

[0151] The fluorescence-based signal may be fluorescence excited by irradiating a fluorescent substance-labeled analyte with light, or autofluorescence generated from an unstained analyte. When the analyte in the specimen is a cell, the fluorescent substance may be a fluorescent dye that binds to nucleic acids or membrane proteins, or a labeled antibody in which an antibody that binds to a specific protein in the cell is modified with a fluorescent dye.

[0152] The optical signal may be acquired in the form of image data obtained by irradiating light onto an analyte in a specimen and capturing an image of the irradiated analyte. The image data may be obtained by capturing images of individual analytes flowing through the flow path of a flow cell using an imaging device such as a TDI camera or a CCD camera. Alternatively, image data of cells may be acquired by applying, spraying, or spotting a specimen or measurement sample containing cells onto a glass slide and capturing an image of the glass slide using an imaging device.

[0153] The signal obtained from the analyte in the sample is not limited to an optical signal, but may also be an electrical signal obtained from the cells. For example, the electrical signal may be a change in impedance caused by applying a direct current to a flow cell and the analyte flowing through the flow cell. The electrical signal thus obtained is used to calculate a representative value reflecting the volume of the analyte. Alternatively, the electrical signal may be a change in impedance when a radio frequency is applied to the analyte flowing through the flow cell. The electrical signal thus obtained is used to calculate a representative value reflecting the conductivity of the analyte.

[0154] The signal obtained from the analyte in the sample may be a combination of at least two or more of the above-mentioned signals. Combining multiple signals allows for multifaceted analysis of the characteristics of the analyte, enabling more accurate cell classification. The combination may include a combination of at least two of multiple optical signals, such as a forward scattered light signal, a side scattered light signal, and a fluorescent signal, or a combination of scattered light signals from different angles, such as a low-angle scattered light signal and a high-angle scattered light signal. Alternatively, an optical signal and an electrical signal may be combined, and the types and number of signals to be combined are not particularly limited.

[0155] The AI ​​algorithm 60 used in the AI ​​analysis in this embodiment is, for example, a deep learning algorithm. The deep learning algorithm is one type of artificial intelligence algorithm and is composed of a neural network including multiple intermediate layers. Data input to the neural network is processed by a large number of matrix operations. Waveform data corresponding to each analyte is acquired from digital data obtained by A / D converting the analog optical signal exemplified in FIG. 2 above, and the acquired waveform data is input to the AI ​​algorithm 60 for analysis. For example, the AI ​​algorithm 60 classifies the type of analyte corresponding to the input waveform data.

[0156] In this embodiment, the type of analyte in a sample is not limited to being classified by the AI ​​algorithm 60. For each individual analyte passing through a predetermined position in the flow path, signal intensities may be acquired at multiple time points while the analyte passes through the predetermined position, and the type of each analyte may be determined based on the results of recognizing the acquired signal intensities at multiple time points for each analyte as a pattern. The pattern may be recognized as a numerical pattern of signal intensities at multiple time points, or as a shape pattern when the signal intensities at multiple time points are plotted as a graph. When recognized as a numerical pattern, the type of analyte can be determined by comparing the numerical pattern of the analyte with a numerical pattern of a known type. The comparison of the numerical pattern of the analyte with the numerical pattern of a control can be performed using, for example, Spearman's rank correlation, z-score, etc. The type of analyte can be determined by comparing the graph shape pattern of the analyte with a graph shape pattern of a known type. The comparison of the graph shape pattern of the analyte with the graph shape pattern of a known type may be performed using, for example, geometric shape pattern matching or a feature descriptor such as a SIFT Descriptor.

[0157] (Configuration example) An example of the configuration of sample analyzer 4000 will be described in which measuring unit 400 is equipped with an FCM detection section (a detection section based on flow cytometry) for measuring a sample (for example, a blood sample, a urine sample, a body fluid, or a bone marrow fluid).

[0158] FIG. 24 is a block diagram showing the configuration of the measurement unit 400.

[0159] As shown in Figure 24, the measurement unit 400 includes an FCM detection section 410 that detects an analyte in a sample, an analog processing section 420 that processes an analog optical signal output from the FCM detection section 410, an apparatus mechanism section 430, a sample preparation section 440, a sample aspirating section 450, and a measurement unit control section 460.

[0160] The specimen aspirating section 450, for example, aspirates the specimen from a specimen container and dispenses the aspirated specimen into a reaction container (e.g., a reaction chamber, a reaction cuvette). The specimen preparing section 440, for example, aspirates a reagent for preparing a measurement specimen and dispenses the reagent into a reaction container containing the specimen. The measurement specimen is prepared by mixing the specimen and the reagent in the reaction container. The device mechanism section 430 includes mechanisms within the measurement unit 400.

[0161] FIG. 25 is a diagram showing a schematic configuration of the optical system of the FCM detection unit 410. As shown in FIG.

[0162] Light emitted from light source 4111 is irradiated via illumination lens system 4112 onto an analyte in a measurement sample passing through flow cell (sheath flow cell) 4113. This causes scattered light and fluorescence to be emitted from the analyte flowing through flow cell 4113.

[0163] The wavelength of the light emitted from light source 4111 is not particularly limited, and a wavelength suitable for exciting the fluorescent dye is selected. As light source 4111, for example, a semiconductor laser light source, an argon laser light source, a gas laser light source such as a helium-neon laser, or a mercury arc lamp is used. In particular, a semiconductor laser light source is preferable because it is much cheaper than a gas laser light source.

[0164] Forward scattered light generated from the analyte in flow cell 4113 is received by light receiving element 4116 via condenser lens 4114 and pinhole portion 4115. Light receiving element 4116 is, for example, a photodiode. Side scattered light generated from the analyte in flow cell 4113 is received by light receiving element 4121 via condenser lens 4117, dichroic mirror 4118, bandpass filter 4119, and pinhole portion 4120. Light receiving element 4121 is, for example, a photodiode. Fluorescence generated from the analyte in flow cell 4113 is received by light receiving element 4122 via condenser lens 4117 and dichroic mirror 4118. Light receiving element 4122 is, for example, an avalanche photodiode. Note that photomultiplier tubes may be used as light receiving elements 4116, 4121, and 4122.

[0165] The analog light receiving signals (optical signals) output from the light receiving elements 4116, 4121, and 4122 are input to the analog processing unit 420 via amplifiers 4151, 4152, and 4153, respectively.

[0166] The analog processing unit 420 performs processes such as noise removal and smoothing on the optical signal input from the FCM detection unit 410 , and outputs the processed optical signal to the A / D conversion unit 461 .

[0167] Returning to FIG. 24, the measurement unit control section 460 includes an A / D conversion section 461, an IF (interface) section 462, a bus 463, and IF sections 464 and 465.

[0168] The A / D conversion unit 461 converts analog optical signals output from the analog processing unit 420 from the start to the end of measurement of the measurement sample into digital data. When multiple types of optical signals (for example, optical signals corresponding to forward scattered light intensity, side scattered light intensity, and fluorescence intensity) are generated from one measurement sample, the A / D conversion unit 461 converts each optical signal from the start to the end of measurement into digital data. For example, as shown in FIG. 25, three types of optical signals (forward scattered light signal, side scattered light signal, and fluorescence signal) are input to the A / D conversion unit 461 via multiple corresponding signal transmission paths 420a. The A / D conversion unit 461 converts each of the optical signals input from the multiple signal transmission paths 420a into digital data. Each signal transmission path 420a is configured to transmit, for example, an analog optical signal as a differential signal.

[0169] The A / D conversion unit 461 compares the signal level of the optical signal with a predetermined threshold and samples optical signals having a signal level greater than the threshold. The A / D conversion unit 461 samples the analog optical signal at a predetermined sampling rate (e.g., sampling at 1024 points at 10 nanosecond intervals, sampling at 128 points at 80 nanosecond intervals, or sampling at 64 points at 160 nanosecond intervals). The A / D conversion unit 461 performs sampling processing on, for example, three types of optical signals corresponding to each analyte, thereby generating digital data (waveform data) of a forward scattered light signal, digital data (waveform data) of a side scattered light signal, and digital data (waveform data) of a fluorescent light signal for each analyte. Each piece of digital data (waveform data) corresponds to one of the analytes in the specimen.

[0170] The A / D converter 461 assigns an index to each piece of generated waveform data. The generated waveform data is, for example, digital data corresponding to each of N analytes contained in one specimen. As a result, three types of waveform data are generated corresponding to three types of optical signals (forward scattered light signal, side scattered light signal, and fluorescent light signal) for each analyte.

[0171] The waveform data generated by the A / D conversion section 461 is transmitted to the analysis unit 300 via the IF sections 462, 465 and the bus 463. The device mechanism section 430, the sample preparation section 440 and the specimen aspirating section 450 are controlled by the analysis unit 300 via the IF sections 464, 465 and the bus 463.

[0172] FIG. 26 is a block diagram showing the configuration of the analysis unit 300.

[0173] The analysis unit 300 includes a processor 3001, a RAM 3017, a bus 3003, a storage unit 3004, an IF unit 3006, a display unit 3011, and an operation unit 3012. The analysis unit 300 is configured by, for example, a personal computer. The analysis unit 300 is connected to the measurement unit 400 via the IF unit 3006.

[0174] The processor 3001 is configured by, for example, a CPU. The processor 3001 executes a program loaded from the storage unit 3004 to the RAM 3017. The RAM 3017 is a so-called main memory. The processor 3001 analyzes waveform data acquired by the measurement unit 400 by executing an analysis program. The processor 3001 controls the analysis unit 300 and the measurement unit 400 by executing a control program.

[0175] The storage unit 3004 is configured, for example, by a hard disk drive (HDD) or a solid state drive (SSD). The storage unit 3004 stores waveform data received from the measurement unit 400, a program for controlling the analysis unit 300 and the measurement unit 400, and a program for analyzing the waveform data. The program for analyzing the waveform data is configured to analyze the waveform data based on the above-mentioned computational analysis and AI analysis. The storage unit 3004 also stores rules for identifying waveform data to be subjected to each of the AI ​​analysis and computational analysis, and rules for selecting analysis operations.

[0176] The display unit 3011 is configured by, for example, a liquid crystal display. The display unit 3011 is connected to the processor 3001 via the bus 3003 and the IF unit 3006. The display unit 3011 displays, for example, the analysis results obtained by the measurement unit 400.

[0177] The operation unit 3012 is composed of, for example, a keyboard, a mouse, and a pointing device including a touch panel. By operating the operation unit 3012, a user such as a doctor or a laboratory technician can input a measurement order to the sample analyzer 4000 and input a measurement instruction based on the measurement order. By operating the operation unit 3012, the user can also input an instruction to display the analysis results. The analysis results include, for example, numerical results based on the analysis, graphs, charts, and flag information assigned to the sample.

[0178] FIG. 27 is a block diagram showing the configuration of the measurement unit 400 when the sample analyzer 4000 counts and classifies blood cells in a blood sample.

[0179] The measurement unit 400 in FIG. 27 further includes an RBC / PLT detection unit 4101, an HGB detection unit 4102, analog processing units 4201 and 4202, and A / D conversion units 4611 and 4612 in addition to the configuration in FIG.

[0180] The RBC / PLT detection unit 4101 is an electrical resistance type detection unit that measures blood cells using a sheath flow DC detection method based on an RBC / PLT measurement sample. The HGB detection unit 4102 measures hemoglobin using an SLS-hemoglobin method based on a hemoglobin measurement sample. Data obtained by A / D conversion of analog signals acquired from the RBC / PLT detection unit 4101 and the HGB detection unit 4102 is subjected to computational analysis. The data from the RBC / PLT detection unit 4101 is used to count the red blood cells and platelets in the blood sample. The data from the HGB detection unit 4102 is used to obtain the amount of hemoglobin in the blood sample.

[0181] The data obtained by A / D converting the analog signals acquired from the RBC / PLT detection unit 4101 and the HGB detection unit 4102 may be subjected to AI analysis. Furthermore, AI analysis and computational analysis may be selectively used for the data based on the RBC / PLT detection unit 4101 and the HGB detection unit 4102. This reduces the load on the analysis unit 300 that processes the data.

[0182] FIG. 28 is a block diagram showing the configuration of the specimen aspirating section 450 and the specimen preparing section 440 in the measuring unit 400 of FIG.

[0183] The specimen suction unit 450 includes a nozzle 451 for aspirating a blood specimen (e.g., whole blood) from a blood collection tube TB, and a pump 452 for applying negative and positive pressure to the nozzle. The nozzle 451 is inserted into the blood collection tube TB by being moved up and down by the device mechanism unit 430 (see FIG. 27). When the pump 452 applies negative pressure while the nozzle 451 is inserted into the blood collection tube TB, the blood specimen is aspirated through the nozzle 451. The device mechanism unit 430 may also include a hand member for inverting and stirring the blood collection tube TB before aspirating blood from the blood collection tube TB.

[0184] The sample preparation unit 440 includes a WDF sample preparation unit 440a, a RET sample preparation unit 440b, a WPC sample preparation unit 440c, a PLT-F sample preparation unit 440d, and a WNR sample preparation unit 440e. Each of the sample preparation units 440a to 440e includes a reaction chamber for mixing a specimen with a reagent (e.g., a hemolytic agent and a staining solution). The sample preparation units 440a to 440e are used in the WDF channel, the RET channel, the WPC channel, the PLT-F channel, and the WNR channel, respectively.

[0185] Here, sample analyzer 4000 is equipped with multiple measurement channels corresponding to the multiple types of measurement samples to be prepared. Sample analyzer 4000 is equipped with, for example, a WDF channel, a RET channel, a WPC channel, a PLT-F channel, and a WNR channel. The WDF channel is a channel for detecting neutrophils, lymphocytes, monocytes, and eosinophils. The RET channel is a channel for detecting reticulocytes. The WPC channel is a channel for detecting blasts and abnormal lymphocyte cells. The PLT-F channel is a channel for detecting platelets. The WNR channel is a channel for detecting white blood cells other than basophils, basophils, and nucleated red blood cells.

[0186] The sample preparation units 440a to 440e are connected via flow paths to a hemolyzing agent container containing a hemolyzing agent, which is a reagent corresponding to the measurement channel, and a staining solution container containing a staining solution. For example, the WDF sample preparation unit 440a is connected via flow paths to a hemolyzing agent container containing a WDF hemolyzing agent (e.g., LyserCell WDF II; manufactured by Sysmex Corporation), which is a reagent for WDF measurement, and a staining solution container containing a WDF staining solution (e.g., FluoroCell WDF; manufactured by Sysmex Corporation). While a configuration in which one sample preparation unit is connected to a hemolyzing agent container and a staining solution container is illustrated here, one sample preparation unit does not necessarily have to be connected to both a hemolyzing agent container and a staining solution container. One reagent container may be shared by multiple sample preparation units. Furthermore, the sample preparation units and the reagent containers do not necessarily have to be connected via flow paths. Instead, a nozzle may be used to aspirate a reagent from a reagent container, and the nozzle may move and dispense the aspirated reagent from the nozzle into a reaction chamber in the sample preparation unit.

[0187] Nozzle 451, which has aspirated the blood sample, is moved horizontally and vertically by device mechanism 430 to be positioned above the reaction chamber of one of sample preparation sections 440a to 440e that corresponds to the measurement order. In this state, when pump 452 applies positive pressure, the blood sample is ejected from nozzle 451 into the corresponding reaction chamber. Sample preparation section 440 supplies a hemolytic agent and staining solution corresponding to the reaction chamber into which the blood sample has been ejected, and prepares a measurement sample by mixing the blood sample, hemolytic agent, and staining solution in the reaction chamber.

[0188] A WDF measurement sample is prepared in a WDF sample preparation unit 440a, a RET measurement sample is prepared in a RET sample preparation unit 440b, a WPC measurement sample is prepared in a WPC sample preparation unit 440c, a PLT-F measurement sample is prepared in a PLT-F sample preparation unit 440d, and a WNR measurement sample is prepared in a WNR sample preparation unit 440e. The prepared measurement samples are supplied from the reaction chamber via a flow channel to an FCM detection unit 410, where cells are measured by flow cytometry.

[0189] The above-mentioned measurement channels (WDF, RET, WPC, PLT-F, WNR) correspond to the measurement items included in the measurement order. For example, the WDF channel corresponds to the measurement item related to white blood cell classification, the RET channel corresponds to the measurement item related to reticulocytes, the PLT-F channel corresponds to the measurement item related to platelets, and the WNR channel corresponds to the measurement items related to white blood cell count and nucleated red blood cells. The measurement samples prepared in the above-mentioned measurement channels are measured by the FCM detection unit 410.

[0190] The measurement results from the RBC / PLT detection unit 4101 correspond to the measurement item related to the number of red blood cells, and the measurement results from the HGB detection unit 4102 correspond to the measurement item related to the amount of hemoglobin.

[0191] FIG. 29 is a block diagram showing another configuration of the sample preparation section 440 shown in FIG.

[0192] In the example shown in FIG. 29, the configuration of the measurement channels in the sample preparation unit 440 has been changed depending on the division of roles between AI analysis and computational analysis. Specifically, compared to the sample preparation unit 440 in FIG. 28, the sample preparation unit 440 in FIG. 29 has added a WDF sample preparation unit 440a for the WDF channel and reagents (WDF hemolyzing agent and WDF staining solution) connected to the WDF sample preparation unit 440a, instead of the WNR sample preparation unit 440e for the WNR channel and reagents (WNR hemolyzing agent and WNR staining solution) connected to the WNR sample preparation unit 440e. That is, the sample preparation unit 440 in FIG. 29 includes two sets of WDF sample preparation units 440a and reagents corresponding to the WDF channels. Note that the sample preparation unit 440 may include three or more sets of WDF sample preparation units 440a and reagents corresponding to the WDF channels.

[0193] When the sample preparation section 440 is configured as shown in FIG. 29 , the classification of basophils and nucleated red blood cells that would otherwise be performed in the WNR channel is performed in the WDF channel. The analysis unit 300 classifies neutrophils, lymphocytes, monocytes, eosinophils, basophils, and nucleated red blood cells by performing AI analysis on waveform data obtained from the measurement sample prepared in the WDF channel. In this case, for example, waveform data obtained by measurement using the WDF channel, corresponding to neutrophils, lymphocytes, monocytes, eosinophils, basophils, and nucleated red blood cells, is trained in advance as training data for the AI ​​algorithm 60. This generates an AI algorithm 60 capable of classifying neutrophils, lymphocytes, monocytes, eosinophils, basophils, and nucleated red blood cells from the waveform data of the WDF channel.

[0194] 29, for example, measurement samples of different specimens can be prepared in parallel in each reaction chamber of the multiple WDF specimen preparation units 404a, allowing measurements of different specimens in the WDF channels to be performed in parallel.

[0195] 29, the reaction chamber and reagent corresponding to the original measurement channel (WNR channel) are replaced with the reaction chamber and reagent corresponding to the subsequent measurement channel (WDF channel). In this case, the analysis of the original measurement channel must be performed by the analysis of the subsequent measurement channel.

[0196] In the configuration of Figure 29, analysis of the original measurement channel (WNR channel) is performed by AI analysis of waveform data from the subsequent measurement channel (WDF channel). This makes it possible to replace the original measurement channel (WNR channel) with the subsequent measurement channel (WDF channel). Therefore, the number of measurement channels that can be added can be increased without increasing the total number of measurement channels provided in the sample analyzer 4000. Increasing the number of WDF channels allows different samples to be measured in parallel on multiple WDF channels, improving the measurement throughput of the WDF channels. Furthermore, by sharing the AI ​​analysis and computational analysis, the computational load required for AI analysis can be reduced and sample processing throughput can be improved, resulting in significant benefits.

[0197] In the process described with reference to Figure 7, an example was shown in which analysis was performed using either AI analysis or computer-based analysis depending on the measurement item, but it may also be possible to determine whether the analysis is performed using AI analysis or computer-based analysis depending on the measurement channel.

[0198] FIG. 30 is a flow chart showing an example in which analysis is performed according to measurement channels.

[0199] In Figure 30, step S101 has been added instead of step S12 compared to Figure 6. The changes from Figure 6 will be described below.

[0200] In step S101, the analysis unit 300 refers to rules including whether to perform AI analysis or computational analysis based on the measurement channel, and based on the rules referred to, identifies waveform data to be subjected to AI analysis and waveform data to be subjected to computational analysis for the waveform data acquired in step S11.

[0201] Fig. 31 is an exemplary diagram showing a screen for setting AI analysis or computer processing analysis for each measurement channel. The measurement channel shown in Fig. 31 is related to a blood cell analyzer.

[0202] The screen of Fig. 31 is displayed, for example, on a display unit provided in the analysis unit 300. The screen of Fig. 31 includes, for each measurement channel, a check box for setting AI analysis and a check box for setting computational analysis. The user operates the check box to select whether to perform AI analysis or computational analysis for each measurement channel, and operates the setting button. This causes the rule to be stored in the memory unit of the analysis unit 300.

[0203] 30, if the waveform data identified in step S101 includes waveform data that is the target of AI analysis (S13: YES), then in step S14, the analysis unit 300 performs AI analysis on the waveform data that is the target of AI analysis identified in step S101. If the waveform data identified in step S101 includes waveform data that is the target of computational analysis (S15: YES), then in step S16, the analysis unit 300 performs computational analysis on the waveform data that is the target of computational analysis identified in step S101.

[0204] In the case of Figure 31, since the WDF channel is set as the target of AI analysis, the analysis of the measurement items associated with the WDF channel (e.g., white blood cell types) is performed by AI analysis of the waveform data obtained from the measurement sample prepared in the WDF channel. On the other hand, since the other channels are set as the target of computational analysis, the analysis of the measurement items associated with the other channels is performed by computational analysis of the waveform data obtained from the measurement sample prepared in the other channels.

[0205] When a WDF channel is set as the target of AI analysis, AI analysis may be performed on all measurement items corresponding to the WDF channel, or AI analysis may be performed on some of the measurement items corresponding to the WDF channel, and computational analysis may be performed on the remaining measurement items.

[0206] <Example of analysis method for analytes in specimens> Next, a method for generating training data 75 and a method for analyzing waveform data will be described using examples shown in FIGS.

[0207] <Waveform data> FIG. 32 is a schematic diagram for explaining waveform data used in this analysis method.

[0208] As shown in the upper diagram of Figure 32, when a measurement sample prepared from a specimen containing analyte A is flowed through flow cell 4113 and light is irradiated onto analyte A flowing through flow cell 4113, forward scattered light is generated in the forward direction relative to the direction of light propagation. Similarly, side scattered light and fluorescence are generated to the side relative to the direction of light propagation. The forward scattered light, side scattered light, and fluorescence are received by light-receiving elements 4116, 4121, and 4122, respectively, and signals corresponding to the received light intensities are output. As a result, analog optical signals representing changes in the signals over time are output from light-receiving elements 4116, 4121, and 4122, respectively. The optical signal corresponding to the forward scattered light is referred to as the "forward scattered light signal," the optical signal corresponding to the side scattered light is referred to as the "side scattered light signal," and the optical signal corresponding to the fluorescence is referred to as the "fluorescence signal." The optical signals are input to A / D conversion unit 461 and converted into digital data.

[0209] The middle diagram of Fig. 32 is a diagram showing a typical conversion to digital data by the A / D conversion unit 461. Here, an analog optical signal is directly input to the A / D conversion unit 461. The level of the optical signal may be converted directly into digital data, or may be subjected to processes such as noise removal, baseline correction, and normalization as appropriate.

[0210] As shown in the middle diagram of FIG. 32, A / D conversion unit 461 samples the forward scattered light signal, side scattered light signal, and fluorescent light signal from the analog optical signals input from light receiving elements 4116, 4121, and 4122, starting from the point when the level of the forward scattered light signal exceeds a predetermined threshold, and ending from the point when the level of the forward scattered light signal falls below the predetermined threshold. Digital waveform data corresponding to one analyte is acquired from the waveform between the starting point and the ending point. A / D conversion unit 461 samples each optical signal at a predetermined sampling rate (for example, sampling at 1024 points at 10 nanosecond intervals, sampling at 128 points at 80 nanosecond intervals, or sampling at 64 points at 160 nanosecond intervals).

[0211] Here, for convenience, a start point and an end point are set for the analog optical signal to acquire the waveform data, but as described above, after all the optical signals are converted into digital data, a start point and an end point may be set for the digital data to acquire the waveform data.

[0212] The lower diagram in FIG. 32 is a diagram schematically illustrating waveform data obtained by sampling. Through sampling, matrix data (one-dimensional array data) is obtained as waveform data corresponding to one analyte, with elements being values ​​that digitally indicate analog signal levels at multiple time points. A / D conversion unit 461 generates waveform data of forward scattered light, waveform data of side scattered light, and waveform data of fluorescence for each analyte. A / D conversion unit 461 repeats generating waveform data until the number of acquired analytes reaches a predetermined number or until a predetermined time has elapsed since the sample was flowed through flow cell 4113. This results in digital data consisting of waveform data of N analytes contained in one sample. A collection of sampling data for each analyte (for example, a collection of 1024 digital values ​​every 10 nanoseconds from t=0 ns to t=10240 ns) corresponds to the waveform data.

[0213] An index for identifying each analyte may be assigned to each piece of waveform data generated by A / D conversion unit 461. For example, integers from 1 to N are assigned as indices in the order in which the waveform data is generated, and the same index is assigned to waveform data of forward scattered light, waveform data of side scattered light, and waveform data of fluorescence obtained from the same analyte. By assigning the same index to waveform data corresponding to the same analyte, AI algorithm 60, which will be described later, can analyze the waveform data of forward scattered light, waveform data of side scattered light, and waveform data of fluorescence corresponding to each analyte as a set, and classify the type of analyte.

[0214] <Generating training data> FIG. 33 is a schematic diagram showing an example of a method for generating training data used to train an AI algorithm 50 for determining the type of analyte in a sample.

[0215] By measuring an analyte based on flow cytometry, an optical signal 70a corresponding to forward scattered light, an optical signal 70b corresponding to side scattered light, and an optical signal 70c corresponding to fluorescence are obtained from the analyte. Waveform data 72a, 72b, and 72c corresponding to the analyte are obtained based on the optical signals 70a, 70b, and 70c, respectively. For example, the training data 75 may be waveform data 72a, 72b, and 72c of analytes determined to be likely to be of a specific type as a result of a computational analysis of analytes in a sample measured based on flow cytometry.

[0216] An example in which the sample analyzer 4000 is used as a blood cell counter for analyzing blood samples will be described below.

[0217] The operator measures a blood sample using the FCM detection unit 410 and accumulates waveform data of forward scattered light, side scattered light, and fluorescence for each analyte contained in the sample. Next, the operator classifies the analytes (cells) in the sample into groups of neutrophils, lymphocytes, monocytes, eosinophils, basophils, immature granulocytes, and abnormal cells, for example, based on the peak values ​​of the waveform data based on the side scattered light and the peak values ​​of the waveform data based on the fluorescence. The operator obtains training data 75 by assigning label values ​​77 corresponding to the classified cell types to the waveform data of the cells. Because the training data 75 is generated for each cell type, the label values ​​77 differ depending on the cell type, as shown in FIG. 34 .

[0218] At this time, the operator calculates the mode, average, or median of the peak values ​​of waveform data based on the side scattered light and fluorescence of cells contained in the neutrophil population, identifies a representative cell based on these values, and assigns a label value of "1" corresponding to a neutrophil to the waveform data of the identified cell.

[0219] The method of generating training data 75 is not limited to this. For example, the operator may collect only specific cells using a cell sorter, measure those cells based on flow cytometry, and assign cell label values ​​to the obtained waveform data to obtain training data 75.

[0220] Waveform data 72a, 72b, and 72c are combined with label values ​​77 that represent the cell types from which the data originated. Training data 75 includes three associated waveform data (waveform data based on optical signals 70a, 70b, and 70c) corresponding to each cell. Training data 75 is then input to AI algorithm 50.

[0221] <Deep Learning Overview> Using Figure 33 as an example, we will explain the outline of neural network training.

[0222] The AI ​​algorithm 50 is configured by a neural network including multiple hidden layers. In this case, the neural network is, for example, a convolutional neural network having a convolutional layer. The number of nodes in the input layer 50a of the neural network corresponds to the number of elements in the arrays contained in the waveform data 72a, 72b, and 72c of the input training data 75. The number of elements in the arrays is equal to the sum of the number of elements in the waveform data 72a, 72b, and 72c of forward scattered light, side scattered light, and fluorescence corresponding to one analyte.

[0223] In the example of Figure 33, each of the waveform data 72a, 72b, and 72c contains 1024 elements, so the number of nodes in the input layer 50a is 1024 x 3 = 3072. The waveform data 72a, 72b, and 72c are input to the input layer 50a of the neural network. The label values ​​77 of each waveform data in the training data 75 are input to the output layer 50b of the neural network to train the neural network. An intermediate layer 50c is positioned between the input layer 50a and the output layer 50b.

[0224] <Waveform data analysis method> FIG. 35 is a diagram showing a schematic diagram of a method for analyzing waveform data of an analyte in a sample using an AI algorithm 60.

[0225] 35, an analyte is measured using flow cytometry to obtain an optical signal 80a corresponding to forward scattered light, an optical signal 80b corresponding to side scattered light, and an optical signal 80c corresponding to fluorescence from the analyte. Waveform data 82a, 82b, and 82c corresponding to the analyte are obtained based on the optical signals 80a, 80b, and 80c, respectively. Analysis data 85 consisting of the waveform data 82a, 82b, and 82c is then generated.

[0226] It is preferable that the analytical data 85 and the training data 75 have at least the same acquisition conditions. The acquisition conditions include conditions for measuring an analyte in a specimen based on flow cytometry, such as preparation conditions for the measurement sample, the flow rate when the measurement sample is passed through the flow cell, the intensity of light irradiated onto the flow cell, and the amplification factor of the light receiving unit that receives scattered light and fluorescent light. The acquisition conditions also include the sampling rate when analog optical signals are A / D converted.

[0227] The analytical data 85 includes three pieces of waveform data (waveform data based on optical signals 80a, 80b, and 80c) corresponding to each analyte in an associated state. The analytical data 85 is then input to the trained AI algorithm 60. The AI ​​algorithm 60 is configured by a neural network including multiple hidden layers.

[0228] When analytical data 85 is input to input layer 60a of the neural network that constitutes AI algorithm 60, classification information 82 regarding the type of analyte corresponding to analytical data 85 is output from output layer 60b. Intermediate layer 60c is positioned between input layer 60a and output layer 60b. Classification information 82 includes the probability that the analyte falls into each of multiple types. Furthermore, the type with the highest probability is determined to be the type to which the analyte belongs, and a label value 83, which is an identifier representing that type, and an analysis result 84, which is a character string representing that type, are output.

[0229] 35, the type of analyte corresponding to analytical data 85 has the highest probability of being a neutrophil, so "1" is output as label value 83, and the text data "neutrophil" is output as analytical result 84. The label value 83 and analytical result 84 may be output by AI algorithm 60, or another computer program may output the most favorable label value 83 and analytical result 84 based on the probability calculated by AI algorithm 60.

[0230] The method of analyzing the waveform data in the examples shown in the above-mentioned FIGS. 19 to 21 will be described with reference to the above-mentioned FIGS.

[0231] 19 to 21, first, the analysis unit 300 analyzes the acquired waveform data by computational processing. Then, the analysis unit 300 performs AI analysis on the waveform data corresponding to the predetermined types of cells classified by the computational analysis (monocytes and lymphocytes in the examples of FIGS. 19 to 21).

[0232] When a cell is classified into a predetermined cell in the computational analysis, the cell is identified by, for example, the index of the waveform data in FIG. 32. As a result, the waveform data classified into monocytes and lymphocytes in the computational analysis is specified by the index in the AI ​​analysis. Analysis unit 300 performs AI analysis on the waveform data specified based on the index, according to the example of FIG. 35. Analysis unit 300 inputs, for example, the waveform data specified by the index into AI algorithm 60 that has been trained to be able to classify monocytes and lymphocytes in more detail.

[0233] The method of analyzing waveform data described above with reference to FIG. 29 will now be described with reference to FIGS. 32 and 35.

[0234] In the analysis method described with reference to Fig. 29, the analysis unit 300 performs AI analysis of waveform data obtained by measurement using the WDF channel, for example, to classify and count nucleated red blood cells (NRBCs) and basophils (BASOs), as well as eosinophils, neutrophils, lymphocytes, and monocytes. In this example, the AI ​​algorithm 60 is trained to be able to classify nucleated red blood cells, basophils, eosinophils, neutrophils, lymphocytes, and monocytes using waveform data. Using such an AI algorithm 60 makes it possible to replace the WNR channel with the WDF channel.

[0235] FIG. 36 is a flowchart showing an example of performing AI analysis on waveform data acquired through a WDF channel.

[0236] In step S111, the measurement unit 400 acquires an optical signal from the measurement sample prepared in the WDF channel and acquires waveform data from the acquired optical signal. In step S112, the analysis unit 300 performs AI analysis on the waveform data acquired in step S111. In step S113, the analysis unit 300 provides both the analysis results for the waveform data of the WDF channel and the analysis results for the waveform data of the other channels. How the analysis of the waveform data of the other channels is divided between AI analysis and computational analysis is determined based on, for example, any of the rules exemplified in the above-mentioned embodiments.

[0237] 29 , the analysis unit 300 classifies and counts nucleated red blood cells and basophils, for example, by AI analysis of waveform data obtained from the WDF channel. The analysis unit 300 performs computational analysis on waveform data corresponding to cells that are not classified as either nucleated red blood cells or basophils, and classifies and counts eosinophils, neutrophils, lymphocytes, and monocytes. In this example, the AI ​​algorithm 60 is trained, for example, to classify analytes into nucleated red blood cells, basophils, and other analytes from the waveform data.

[0238] The analysis unit 300 performs computational analysis on waveform data corresponding to cells that have not been classified as either nucleated red blood cells or basophils. For example, peak values ​​of waveform data corresponding to cells that have not been classified as either nucleated red blood cells or basophils are extracted, and the cell type is classified based on a two-dimensional graph (scattergram) generated from peak values ​​corresponding to side scattered light and peak values ​​corresponding to fluorescence. For example, based on the two-dimensional graph, the cells are classified as eosinophils, neutrophils, lymphocytes, monocytes, or others. Cells classified as other than eosinophils, neutrophils, lymphocytes, and monocytes in the analysis based on the two-dimensional graph are classified as, for example, debris.

[0239] FIG. 37 is a flowchart showing an example of classifying nucleated red blood cells and basophils by AI analysis and classifying the rest by computational analysis based on waveform data acquired by the WDF channel.

[0240] In step S121, the measurement unit 400 acquires an optical signal from the measurement sample prepared in the WDF channel and acquires waveform data from the acquired optical signal. In step S122, the analysis unit 300 performs AI analysis on the waveform data acquired in step S121. This allows nucleated red blood cells and basophils to be classified. In step S123, the analysis unit 300 identifies waveform data corresponding to cells that are not classified as either nucleated red blood cells or basophils.

[0241] In step S124, the analysis unit 300 performs a computational analysis on the waveform data identified in step S123, thereby classifying lymphocytes, monocytes, eosinophils, and neutrophils. In step S125, the analysis unit 300 provides the analysis results of the waveform data of the WDF channel and the analysis results of the waveform data of the other channels together.

[0242] In another analysis method based on the configuration shown in FIG. 29 , the analysis unit 300 performs, for example, computational analysis of waveform data obtained from the WDF channel to classify and count lymphocytes, monocytes, eosinophils, and neutrophils or basophils. In classifying and counting neutrophils or basophils, for example, cells classified as either neutrophils or basophils are counted. The analysis unit 300 then performs AI analysis on waveform data corresponding to lymphocytes, monocytes, eosinophils, and cells not classified as either neutrophils or basophils, as well as cells classified as either neutrophils or basophils. This classifies the analytes into nucleated red blood cells, basophils, and other cells.

[0243] For example, the count results of cells classified as basophils by AI analysis are subtracted from the count results of cells classified as either neutrophils or basophils by computational analysis to calculate the count results of neutrophils and basophils, respectively. Cells that are not classified as either nucleated red blood cells or basophils by AI analysis are classified as, for example, debris.

[0244] FIG. 38 is a flowchart showing an example of performing AI analysis on neutrophils / basophils identified by analysis of computational processing in the WDF channel.

[0245] In step S131, the measurement unit 400 acquires an optical signal from the measurement sample prepared in the WDF channel and obtains waveform data from the acquired optical signal. In step S132, the analysis unit 300 performs a computational analysis on the waveform data acquired in step S131. This results in classification into groups consisting of lymphocytes, monocytes, eosinophils, and neutrophils and basophils. In step S133, the analysis unit 300 identifies waveform data corresponding to (1) cells not classified as lymphocytes, monocytes, eosinophils, neutrophils, or basophils, and (2) cells classified as neutrophils or basophils.

[0246] In step S134, the analysis unit 300 performs AI analysis on the waveform data identified in step S133, thereby classifying neutrophils and basophils. In step S135, the analysis unit 300 provides the analysis results of the waveform data of the WDF channel and the analysis results of the waveform data of the other channels.

[0247] [Embodiment 5] In the fifth embodiment, a detailed configuration example is shown in which a sample analyzer 4000 that analyzes the coagulation ability of a blood sample performs computational analysis and AI analysis in a shared manner.

[0248] The specimen to be measured by the specimen analyzer 4000 of the fifth embodiment may be a biological sample collected from a subject. The specimen may include, for example, whole blood or plasma. The specimen analyzer 4000 of the fifth embodiment analyzes the specimen for abnormalities caused by interfering substances using a coagulation method, a synthetic substrate method, a turbidimetric immunoassay, an agglutination method, or a chemiluminescent enzyme immunoassay (CLEIA) method. The specimen analyzer 4000 of the fifth embodiment includes, for example, a measurement unit 400 and an analysis unit 300, similar to the configuration example of the first embodiment shown in FIG. 1.

[0249] (Configuration example) FIG. 39 is a block diagram schematically showing the configuration of a measurement unit 400 according to the fifth embodiment.

[0250] 24, the measurement unit 400 in FIG. 39 includes a detection unit 470 instead of the FCM detection unit 410, and further includes a control unit 466.

[0251] The detection unit 470 includes a light source unit 471 and a detection block 476. The light source unit 471 includes, for example, a halogen lamp. The light source unit 471 is configured to be able to emit, for example, light with a wavelength of 660 nm for measuring blood clotting time, light with a wavelength of 405 nm for measuring synthetic substrates, and light with a wavelength of 800 nm for immunoturbidimetric measurement. The sample preparation unit 440 prepares a measurement sample by mixing a blood clotting reagent with a specimen. The detection unit 470 irradiates the measurement sample consisting of the blood clotting reagent and the specimen with light from the light source unit 471 and detects the light transmitted through the specimen. Alternatively, the detection unit 470 may irradiate the measurement sample with light from the light source unit 471 and detect light scattered by the specimen.

[0252] The control unit 466 is configured by, for example, an FPGA. The control unit 466 is connected to the analysis unit 300 via the bus 463 and the IF unit 465. The control unit 466 controls each part of the measurement unit 400 based on instructions from the analysis unit 300.

[0253] FIG. 40 is a side view that schematically illustrates measurement by the detection block 476.

[0254] The detection block 476 includes a holder 472 , an optical fiber 473 , a condenser lens 474 , and a light receiving unit 475 .

[0255] The holder 472 holds a reaction vessel C1 that contains a measurement sample prepared from a specimen and a sample corresponding to the measurement item. This allows the measurement sample to stand still. Light emitted from the light source 471 (see FIG. 39) is guided to a condenser lens 474 by an optical fiber 473. The condenser lens 474 condenses the light from the optical fiber 473 into the reaction vessel C1. The transmitted light that is condensed into the reaction vessel C1 and passes through the measurement sample in the reaction vessel C1 is received by a light receiving unit 475. The light receiving unit 475 is, for example, a photodiode. The light receiving unit 475 outputs an optical signal based on the intensity of the received transmitted light.

[0256] Returning to FIG. 39, the analog processing unit 420 processes the analog optical signal output from the light receiving unit 475 (see FIG. 40) and outputs it to the A / D conversion unit 461. The A / D conversion unit 461 converts the analog optical signal into a digital signal. As described above, the digital data obtained by converting the optical signal into digital data is the clot waveform data as shown in FIG. 4. The control unit 466 transmits the acquired clot waveform data to the analysis unit 300.

[0257] 41 is a flowchart showing an example of analysis according to embodiment 5. In embodiment 5, the processor 3001 (see FIG. 26) of the analysis unit 300 performs computational analysis and AI analysis on the clot waveform data.

[0258] In step S141, the measurement unit 400 acquires an optical signal in the detection section 470 and obtains clot waveform data from the acquired optical signal. In step S142, the analysis unit 300 performs a computational analysis on the clot waveform data acquired in step S141. For example, as described with reference to FIG. 4, the analysis unit 300 acquires the time (T-T2) required for the absorbance of the clot waveform data to decrease to 50% as a result indicating the time until the blood sample clots.

[0259] In step S143, the analysis unit 300 performs AI analysis on the clot waveform data acquired in step S141. As a result, the analysis unit 300 acquires the presence or absence of measurement abnormalities based on the feature amounts extracted from the clot waveform data by the AI ​​algorithm 60. The analysis unit 300 determines whether a nonspecific reaction is suspected based on the presence or absence of measurement abnormalities.

[0260] In step S144, the analysis unit 300 provides the results obtained in step S142 indicating the time it takes for the blood sample to clot, and the results obtained in step S143 indicating the presence or absence of an abnormality in the measurement.

[0261] In FIG. 41, the AI ​​analysis is always performed in step S143, but the analysis unit 300 may perform the process of step S143 based on a preset rule indicating whether or not the AI ​​analysis is necessary.

[0262] Although sample analyzer 4000 of embodiment 5 is a blood coagulation measurement device that optically measures changes in turbidity of a measurement sample due to coagulation of the blood sample, the present invention is not limited to this. For example, the present invention may be a blood coagulation measurement device that measures changes in the amplitude motion of a steel ball in a measurement sample due to changes in viscosity of the measurement sample due to coagulation of the blood sample, using a high-frequency receiving frequency transmitted from a high-frequency transmitting coil. Furthermore, although sample analyzer 4000 of embodiment 5 is a blood coagulation measurement device, it may also be an immunoassay device, a biochemistry measurement device, or a gene measurement device.

[0263] [Embodiment 6] In the sixth embodiment, a configuration example of a sample analyzer 4000 including a host processor and a parallel processor is shown. In the sixth embodiment, parallel processing is performed on waveform data in the parallel processor 3002, and information regarding each type of analyte is generated based on the results of the parallel processing.

[0264] According to the sixth embodiment, even when analyzing a huge amount of data, ranging from several hundred megabytes to several gigabytes per sample, a parallel processor provided separately from the host processor can execute processing related to the waveform data in parallel. Therefore, even when processing a huge amount of data using the AI ​​algorithm 60, for example, data processing can be completed within the sample analyzer 4000. This eliminates the need to transmit data to an analysis server storing the AI ​​algorithm 60 via the Internet or an intranet. Therefore, according to the sixth embodiment, it is not necessary to transmit a large amount of data from the sample analyzer 4000 to the analysis server and obtain the analysis results returned from the analysis server. This allows the sample analyzer 4000 to maintain a high processing capacity while improving the classification accuracy of analytes in samples.

[0265] The configuration of a sample analyzer 4000 of embodiment 6 will be described with reference to Figures 42 and 43. In the configuration example shown in Figures 42 and 43, a measurement unit 400 includes an FCM detection section 410 for measuring a sample (e.g., a blood sample, a urine sample, a body fluid, or a bone marrow fluid).

[0266] FIG. 42 is a block diagram showing the configuration of a sample analyzer 4000 according to the sixth embodiment.

[0267] The sample analyzer 4000 of the sixth embodiment includes a measurement unit 400 and an analysis unit 300 provided inside the measurement unit 400. Compared to the measurement unit 400 of the fourth embodiment shown in Figure 24, the measurement unit 400 of the sixth embodiment does not include IF units 462, 464, 465 and bus 463. The analysis unit 300 of the sixth embodiment is connected to an A / D conversion unit 461, an apparatus mechanism unit 430, a sample preparation unit 440, and a specimen aspirating unit 450 within the measurement unit 400, and to a computer 301 arranged outside the measurement unit 400.

[0268] FIG. 43 is a block diagram showing the configuration of the analysis unit 300 according to the sixth embodiment.

[0269] The analysis unit 300 of the sixth embodiment includes a parallel processing processor 3002, a bus controller 3005, and IF units 462 and 464, as compared with the analysis unit 300 of the fourth embodiment shown in FIG.

[0270] The parallel processing processor 3002 is configured to be able to process the calculations performed by the AI ​​algorithm 60 in place of the master processor. Using the parallel processing processor 3002, which is suitable for processing the matrix operations executed by the AI ​​algorithm 60, makes it possible to improve the TAT required for AI analysis. However, while the parallel processing processor 3002 improves TAT, the computer load required for AI analysis increases as the amount of data to be analyzed increases. In contrast, as described above, by sharing the data analysis between calculation processing analysis and AI analysis, the computer load can be reduced and inspection efficiency can be improved.

[0271] Processor 3001 uses parallel processing processor 3002 to perform analysis processing of waveform data using AI algorithm 60. That is, processor 3001 executes analysis software 3100 to perform AI analysis of waveform data based on AI algorithm 60. Analysis software 3100 is used to analyze waveform data corresponding to an analyte in a sample based on AI algorithm 60.

[0272] The analysis software 3100 may be stored in the storage unit 3004. In this case, the processor 3001 executes the analysis software 3100 stored in the storage unit 3004 to perform AI analysis of the waveform data based on the AI ​​algorithm 60.

[0273] In this embodiment, for example, AI analysis is performed by processor 3001 and parallel processing processor 3002, and computational analysis is performed by processor 3001 without using parallel processing processor 3002.

[0274] The processor 3001 is, for example, a central processing unit (CPU). The processor 3001 may be, for example, Intel's Core i9, Core i7, or Core i5, or AMD's Ryzen 9, Ryzen 7, Ryzen 5, or Ryzen 3.

[0275] The processor 3001 controls the parallel processing processor 3002. The parallel processing processor 3002 executes parallel processing related to, for example, matrix operations in accordance with the control of the processor 3001. In other words, the processor 3001 is the master processor of the parallel processing processor 3002, and the parallel processing processor 3002 is the slave processor of the processor 3001. The processor 3001 is also called a host processor or a main processor. The processor 3001 executes matrix operations according to the AI ​​algorithm 60 in parallel processing by the parallel processing processor 3002.

[0276] The parallel processing processor 3002 executes multiple arithmetic processes in parallel, which are at least a part of the processing related to the analysis of waveform data. The parallel processing processor 3002 is, for example, a graphics processing unit (GPU), a field programmable gate array (FPGA), or an application specific integrated circuit (ASIC). When the parallel processing processor 3002 is an FPGA, the parallel processing processor 3002 may be pre-programmed with arithmetic processes related to the trained AI algorithm 60, for example. When the parallel processing processor 3002 is an ASIC, the parallel processing processor 3002 may be pre-programmed with a circuit for executing arithmetic processes related to the trained AI algorithm 60, for example, or may have a programmable module built in addition to such a built-in circuit.

[0277] For example, NVIDIA GeForce, Quadro, TITAN, Jetson, etc. may be used as the parallel processing processor 3002. In the Jetson series, for example, Jetson Nano, Jetson Tx2, Jetson Xavier, or Jetson AGX Xavier may be used.

[0278] The processor 3001 executes, for example, calculations related to the control of the measurement unit 400. The processor 3001 executes, for example, calculations related to control signals transmitted and received between the device mechanism section 430, the sample preparation section 440, and the specimen aspirating section 450. The processor 3001 also executes, for example, calculations related to the transmission and reception of information to and from the computer 301.

[0279] The computer 301 has a function of displaying, for example, analysis results sent from the analysis unit 300 based on processing by the processor 3001. The computer 301, for example, sends a measurement order to the analysis unit 300. The measurement order is sent to the computer 301, for example, from a host computer. A user can also input the measurement order via an input device of the computer 301.

[0280] The processor 3001 executes processes related to, for example, reading program data from the storage unit 3004, loading the program into the RAM 3017, and transmitting and receiving data to and from the RAM 3017. The above-described processes executed by the processor 3001 are required to be executed in a predetermined order, for example. For example, if the processes required to control the device mechanism unit 430, the sample preparation unit 440, and the specimen aspirating unit 450 are designated as A, B, and C, respectively, they may be required to be executed in the order B, A, and C. Because the processor 3001 often executes sequential processes that depend on such an order, increasing the number of arithmetic units (sometimes referred to as "processor cores," "cores," etc.) does not necessarily increase the processing speed.

[0281] On the other hand, the parallel processing processor 3002 executes routine, large-volume calculations, such as operations on matrix data containing a large number of elements. In this embodiment, the parallel processing processor 3002 executes parallel processing in which at least a portion of the processing for analyzing waveform data according to the AI ​​algorithm 60 is parallelized. The AI ​​algorithm 60 includes, for example, a large number of matrix operations. The AI ​​algorithm 60 may include, for example, at least 100 matrix operations, or may include at least 1,000 matrix operations.

[0282] The parallel processing processor 3002 has multiple arithmetic units, each of which can simultaneously execute a matrix operation. In other words, the parallel processing processor 3002 can execute matrix operations in parallel by each of the multiple arithmetic units as parallel processing. For example, the matrix operation included in the AI ​​algorithm 60 can be divided into multiple arithmetic operations that are not order-dependent. The arithmetic operations divided in this way can be executed in parallel by each of the multiple arithmetic units. These arithmetic units are sometimes called "processor cores," "cores," etc.

[0283] By performing such parallel processing, it is possible to speed up the calculation processing of the entire sample analyzer 4000. Processing such as matrix calculations included in the AI ​​algorithm 60 is sometimes called "Single Instruction Multiple Data" (SIMD). The parallel processing processor 3002 is suitable for such SIMD calculations, for example. Such a parallel processing processor 3002 is sometimes called a vector processor.

[0284] As described above, the processor 3001 is suitable for executing a variety of complex processes. On the other hand, the parallel processing processor 3002 is suitable for executing a large amount of standardized processes in parallel. By executing a large amount of standardized processes in parallel, the TAT required for calculation processing is shortened.

[0285] The parallel processing performed by the parallel processing processor 3002 is not limited to matrix operations. For example, when the parallel processing processor 3002 performs learning processing on the AI ​​algorithm 50, differential operations related to the learning processing may be parallel processed.

[0286] The number of arithmetic units of the processor 3001 is, for example, a dual-core (number of cores: 2), a quad-core (number of cores: 4), or an octa-core (number of cores: 8). On the other hand, the parallel processing processor 3002 has, for example, at least 10 arithmetic units (number of cores: 10) and can execute 10 matrix operations in parallel. Some parallel processing processors 3002 have tens of arithmetic units. Other parallel processing processors 3002 have, for example, at least 100 arithmetic units (number of cores: 100) and can execute 100 matrix operations in parallel. Some parallel processing processors 3002 have hundreds of arithmetic units. Other parallel processing processors 3002 have, for example, at least 1,000 arithmetic units (number of cores: 1,000) and can execute 1,000 matrix operations in parallel. Some parallel processing processors 3002 have thousands of arithmetic units.

[0287] Figure 44 is a block diagram showing another configuration of sample analyzer 4000 according to embodiment 6. Sample analyzer 4000 of Figure 44 counts and classifies blood cells in a blood sample.

[0288] Compared to sample analyzer 4000 in Figure 42, sample analyzer 4000 in Figure 44 includes an RBC / PLT detection unit 4101, an HGB detection unit 4102, analog processing units 4201 and 4202, and A / D conversion units 4611 and 4612 similar to those in Figure 27. Sample preparation unit 440 in Figure 44 has the same configuration as sample preparation unit 440 shown in Figure 28 or 29.

[0289] FIG. 45 is a diagram showing an example of the configuration of the parallel processing processor 3002.

[0290] The parallel processing processor 3002 includes a plurality of arithmetic units 3200 and a RAM 3201. Each of the arithmetic units 3200 performs arithmetic processing of matrix data in parallel. The RAM 3201 stores data related to the arithmetic processing performed by the arithmetic units 3200. The RAM 3201 is a memory having a capacity of at least 1 gigabyte. The RAM 3201 may also have a capacity of 2 gigabytes, 4 gigabytes, 6 gigabytes, 8 gigabytes, 10 gigabytes, or more. The arithmetic units 3200 obtain data from the RAM 3201 and perform arithmetic processing. The arithmetic units 3200 may be referred to as a "processor core," a "core," or the like.

[0291] 46 to 48 are diagrams showing schematic examples of how the parallel processing processor 3002 is installed.

[0292] In the example shown in FIG. 46, the processor 3001 is mounted on a substrate 3301. The parallel processing processor 3002 is mounted on a graphic board 3300, and the graphic board 3300 is connected to the substrate 3301 via a connector 3310. The processor 3001 is connected to the parallel processing processor 3002 via a bus 3003. In the example shown in FIG. 47, the parallel processing processor 3002 is mounted directly on the substrate 3301 and connected to the processor 3001 via the bus 3003. In the example shown in FIG. 48, the processor 3001 and the parallel processing processor 3002 are provided as an integrated unit. In this case, the parallel processing processor 3002 is built into the processor 3001 mounted on the substrate 3301.

[0293] FIG. 49 is a diagram showing another example of the parallel processing processor 3002.

[0294] In the example shown in FIG. 49, a parallel processing processor 3002 is mounted on the measurement unit 400 by an external device 3400 connected to the measurement unit 400. The parallel processing processor 3002 is implemented in the external device 3400, which is, for example, a USB device. The external device 3400 is connected to the bus 3003 via the IF unit 467, whereby the parallel processing processor 3002 is mounted on the sample analyzer 4000. The USB device may be, for example, a small device such as a USB dongle. The IF unit 467 is, for example, a USB interface having a transfer speed of several hundred Mbps, and more preferably a USB interface having a transfer speed of several Gbps to several tens of Gbps or more. For example, a Neural Compute Stick 2 manufactured by Intel Corporation may be used as the external device 3400 on which the parallel processing processor 3002 is mounted.

[0295] Multiple parallel processors 3002 may be installed in the sample analyzer 4000 by connecting multiple USB devices each equipped with a parallel processor 3002 to the IF section 467. Since the parallel processor 3002 installed in one USB device may have a smaller number of arithmetic units 3200 than a GPU or the like, it is possible to scale up the number of cores by adding multiple USB devices to be connected to the measurement unit 400.

[0296] Next, an overview of the arithmetic processing executed by the parallel processing processor 3002 based on the control of the analysis software 3100 running on the processor 3001 will be described with reference to FIGS.

[0297] FIG. 50 is a diagram showing an example of the configuration of a parallel processor 3002 that executes arithmetic processing.

[0298] The parallel processing processor 3002 has a plurality of arithmetic units 3200 and a RAM 3201. The processor 3001, which executes the analysis software 3100, instructs the parallel processing processor 3002 to execute at least a portion of the arithmetic processing required when analyzing waveform data using the AI ​​algorithm 60. The processor 3001 instructs the parallel processing processor 3002 to execute arithmetic processing related to the analysis of waveform data based on the AI ​​algorithm 60.

[0299] All or at least a portion of the waveform data is stored in RAM 3017. The data stored in RAM 3017 is transferred to RAM 3201 of the parallel processing processor 3002, for example, by DMA (Direct Memory Access). Each of the multiple arithmetic units 3200 of the parallel processing processor 3002 executes arithmetic processing in parallel on the data stored in RAM 3201. Each of the multiple arithmetic units 3200 obtains necessary data from RAM 3201 and executes arithmetic processing. Data corresponding to the arithmetic results is stored in RAM 3201 of the parallel processing processor 3002. The data corresponding to the arithmetic results is transferred from RAM 3201 to RAM 3017, for example, by DMA.

[0300] FIG. 51 is a diagram showing an outline of the matrix operations executed by the parallel processor 3002.

[0301] Matrix multiplication calculations (matrix operations) are performed when analyzing waveform data according to the AI ​​algorithm 60. The parallel processing processor 3002 executes, for example, a plurality of arithmetic processes related to matrix operations in parallel.

[0302] The top diagram in Figure 51 shows a formula for matrix multiplication. In this formula, matrix c is obtained by multiplying matrix a, which has n rows and n columns, by matrix b, which has n rows and n columns. As shown in the top diagram in Figure 51, the formula is written using a multi-level loop syntax. The bottom diagram in Figure 51 shows an example of arithmetic processing executed in parallel by the parallel processor 3002. The formula shown in the bottom diagram in Figure 51 can be divided into n x n arithmetic processing operations, which is the number of combinations of loop variable i in the first level and loop variable j in the second level. Each of the divided arithmetic processing operations is independent of each other and can therefore be executed in parallel.

[0303] FIG. 52 is a conceptual diagram showing that the multiple arithmetic processes illustrated in the lower diagram of FIG. 51 are executed in parallel by a parallel processor 3002.

[0304] 52, each of the multiple arithmetic processes is assigned to one of the multiple arithmetic units 3200 included in the parallel processor 3002. Each of the arithmetic units 3200 executes the assigned arithmetic processes in parallel with each other. In other words, each of the arithmetic units 3200 executes the divided arithmetic processes simultaneously.

[0305] 51 and 52, information regarding the probability that a cell corresponding to waveform data belongs to each of a plurality of cell types is obtained. Processor 3001 executing analysis software 3100 performs an analysis of the cell type of the cell corresponding to the waveform data based on the results of the calculation.

[0306] The calculation of the probability that an analyte in a sample belongs to each of the multiple classification types may be performed by a processor other than the parallel processor 3002. For example, the calculation results by the parallel processor 3002 may be transferred from RAM 3201 to RAM 3017, and the processor 3001 may calculate information regarding the probability that an analyte corresponding to each piece of waveform data belongs to each of the multiple classification types based on the calculation results read from RAM 3017. Alternatively, the calculation results by the parallel processor 3002 may be transferred from RAM 3201 to the analysis unit 300, and a processor mounted on the analysis unit 300 may calculate information regarding the probability that an analyte corresponding to each piece of waveform data belongs to each of the multiple classification types.

[0307] The processing shown in Figures 51 and 52 is applied, for example, to calculation processing (also called filtering processing) related to a convolution layer in the AI ​​algorithm 60.

[0308] FIG. 53 is a diagram schematically illustrating an overview of the computational processing related to the convolutional layer.

[0309] The upper diagram in Figure 53 shows waveform data obtained based on forward scattered light as waveform data input to the AI ​​algorithm 60. The waveform data in this embodiment is one-dimensional matrix data as shown in Figure 32. More simply, waveform data is array data in which elements are arranged in a row. For ease of explanation, the number of elements in the waveform data is assumed to be n (n is an integer equal to or greater than 1). The upper diagram in Figure 53 shows multiple filters. The filters are generated by the learning process of the AI ​​algorithm 50. Each of the multiple filters is one-dimensional matrix data that represents the characteristics of the waveform data. The filter shown in the upper diagram in Figure 53 is matrix data with one row and three columns, but the number of columns is not limited to three. By performing a matrix operation on the waveform data input to the AI ​​algorithm 60 and each filter, characteristics related to the waveform data that correspond to the cell type are calculated.

[0310] The bottom diagram in Figure 53 shows an overview of the matrix operation of waveform data and filters. The matrix operation is performed while shifting each filter by one for each element of the waveform data. The matrix operation is calculated using the following (Equation 1).

[0311]

number

[0312] In (Equation 1), the subscripts of x are variables indicating the row and column numbers of the waveform data. The subscripts of h are variables indicating the row and column numbers of the filter. In the example shown in FIG. 53, the waveform data is one-dimensional matrix data, and the filter is matrix data with 1 row and 3 columns, so L=1, M=3, p=0, q=0, 1, 2, i=0, j=0, 1, ..., n-1.

[0313] The parallel processing processor 3002 executes the matrix operation expressed by (Equation 1) in parallel using each of the multiple calculation units 3200. Classification information regarding the type of each analyte in the sample is generated based on the calculation process executed by the parallel processing processor 3002. The generated classification information is used to generate and display test results for the sample based on the classification information.

[0314] 42 and 43, the computer 301 is connected to the processor 3001 via the IF unit 3006 and the bus 3003, and can receive the analysis results from the processor 3001 and the parallel processing processor 3002. The IF unit 3006 is, for example, a USB interface. The computer 301 receives the analysis results from the analysis unit 300 via the IF unit 3006 and displays the analysis results on a display device of the computer 301.

[0315] The computer 301 may be equipped with an operation unit consisting of a keyboard, a mouse, or a pointing device including a touch panel. By operating the operation unit, a user such as a doctor or laboratory technician can input a measurement order to the sample analyzer 4000 and input measurement instructions according to the measurement order. The user can input instructions to display test results to the computer 301 via the operation unit. The user can operate the operation unit to view various information related to the test results, such as numerical results based on analysis, graphs, charts, and flag information assigned to samples.

[0316] <Sample analyzer operation> The sample analysis operation performed by sample analyzer 4000 will now be described with reference to Figures 54 to 56.

[0317] FIG. 54 is a flowchart showing the analysis operations of the analysis unit 300 and the measurement unit 400.

[0318] In step S200, upon receiving a measurement order, the processor 3001 of the analysis unit 300 instructs the measurement unit 400 to perform measurement. For example, the analysis unit 300 controls the operation of each detection unit (FCM detection unit 410, RBC / PLT detection unit 4101, HGB detection unit 4102), specimen aspirating unit 450, and sample preparing unit 440 of the measurement unit 400 by issuing instructions to the measurement unit 400. The measurement unit 400 starts measuring the specimen in response to instructions from the analysis unit 300.

[0319] In step S300, the sample aspirating unit 450 aspirates the sample from the blood collection tube and dispenses the aspirated sample into the reaction chamber in response to a measurement instruction from the analysis unit 300. The measurement instruction from the analysis unit 300 includes information on the measurement channel for which measurement is requested by the measurement order. Based on the measurement channel information included in the measurement instruction, the sample aspirating unit 450 dispenses the sample into the reaction chamber of the corresponding measurement channel.

[0320] In step S301, the sample preparation section 440 prepares a measurement sample in response to a measurement instruction from the analysis unit 300. Specifically, the sample preparation section 440 supplies reagents (hemolytic agent and staining solution) to the reaction chamber into which the sample has been ejected, based on the measurement channel information included in the measurement instruction, and mixes the sample and reagents. This prepares a measurement sample (for example, a WDF measurement sample, a RET measurement sample, a WPC measurement sample, a PLT-F measurement sample, or a WNR measurement sample). The sample preparation unit 440 also supplies a reagent to the reaction chamber into which the sample has been discharged and mixes the sample and reagent to prepare an RBC / PLT measurement sample.The sample preparation unit 440 supplies a reagent to the reaction chamber into which the sample has been discharged and mixes the sample and reagent to prepare a hemoglobin measurement sample.

[0321] In step S302, the FCM detection unit 410 measures the prepared measurement sample in response to a measurement instruction from the analysis unit 300. Specifically, the device mechanism unit 430 sends the measurement sample in the reaction chamber of the sample preparation unit 440 to the FCM detection unit 410 in response to a measurement instruction from the analysis unit 300. The measurement sample sent from the reaction chamber is made to flow into the flow cell 4113, and is irradiated with laser light by the light source 4111 (see FIG. 25). When an analyte contained in the measurement sample passes through the flow cell 4113, the analyte is irradiated with light, and forward scattered light, side scattered light, and fluorescence generated from the analyte are detected by the light receiving elements 4116, 4121, and 4122, respectively, and an analog optical signal corresponding to the received light intensity is output. The optical signal is processed by the analog processing unit 420 and then output to the A / D conversion unit 461.

[0322] Furthermore, the RBC / PLT detection unit 4101 measures blood cells using a sheath flow DC detection method based on an RBC / PLT measurement sample. The HGB detection unit 4102 measures hemoglobin using an SLS-hemoglobin method based on a hemoglobin measurement sample. The analog signal detected by the RBC / PLT detection unit 4101 is processed by an analog processing unit 4201 and then output to an A / D conversion unit 4611, and the analog signal detected by the HGB detection unit 4102 is processed by an analog processing unit 4202 and then output to an A / D conversion unit 4612 (see FIG. 27). In step S303, the A / D conversion unit 461 generates digital data by sampling the analog optical signal at a predetermined rate, as described above, and generates waveform data corresponding to each analyte based on the digital data. The waveform data generated by the A / D conversion unit 461 is transferred directly to RAM by, for example, DMA transfer without going through the processor 3001 of the analysis unit 300. As a result, waveform data based on the forward scattered light signal acquired from the analyte, waveform data corresponding to the side scattered light, and waveform data corresponding to the fluorescence are loaded into RAM 3017.

[0323] Furthermore, the A / D conversion unit 4611 generates digital data by sampling at a predetermined rate the analog signal from the RBC / PLT detection unit 4101. The A / D conversion unit 4612 generates digital data by sampling at a predetermined rate the analog signal from the HGB detection unit 4102. These digital data may also be stored in the RAM 3017.

[0324] In step S201, the processor 3001 of the analysis unit 300 performs AI analysis on the waveform data using the AI ​​algorithm 60, and performs computational analysis on representative values ​​of the waveform data that correspond to the characteristics of the analytes. The AI ​​analysis and computational analysis are divided as described above. This classifies the analytes in the sample. The AI ​​analysis processing in step S201 will be described later, but as a result of processing using the parallel processing processor 3002, the processor 3001 obtains, for example, classification information 82 for each analyte in the sample, and obtains label values ​​83 and analysis results 84 (see FIG. 35).

[0325] In step S202, processor 3001 analyzes label values ​​83 and analysis results 84 using a program stored in memory 3004 to generate a test result for the specimen. In step S202, for example, the number of analytes for each type of analyte is counted based on the label values ​​83 and analysis results 84 of the individual analytes.

[0326] For example, in the case of testing blood cells in a blood sample, if one sample has N pieces of classification information assigned with a label value of "1" indicating neutrophils, a count result is obtained as the test result for the sample, with the number of neutrophils equal to N. The processor 3001 obtains the count result for the measurement item corresponding to the measurement channel based on the analysis result 84, and stores it in the memory unit 3004 together with the identification information of the sample.

[0327] Here, measurement items according to the measurement channel are items for which count results are requested by the measurement order. For example, measurement items according to the WDF channel include measurement items for the five white blood cell differentials, namely, the number of monocytes, neutrophils, lymphocytes, eosinophils, and basophils. Measurement items according to the RET channel include measurement items for the number of reticulocytes. Measurement items according to the PLT-F channel include measurement items for the number of platelets. Measurement items according to the WPC channel include measurement items for the number of hematopoietic progenitor cells. Measurement items according to the WNR channel include measurement items for the number of white blood cells and nucleated red blood cells.

[0328] The count results are not limited to the items for which measurement is required (also called reportable items) as listed above, but may also include the count results of other cells that can be measured using the same measurement channel. For example, in the case of the WDF channel, as shown in Figure 34, in addition to the five types of white blood cells, the count results also include immature granulocytes (IG) and abnormal cells.

[0329] Furthermore, the processor 3001 generates test results for the specimen by analyzing the obtained counting results and stores them in the memory unit 3004. Analysis of the counting results includes determining, for example, whether the counting results are within the normal range, whether abnormal cells have been detected, whether the deviation from the previous test results is within an acceptable range, and so on.

[0330] In step S203, the computer 301 displays the test results generated by the analysis unit 300 on the display unit.

[0331] FIG. 55 is a flowchart showing the details of the AI ​​analysis in step S201 of FIG.

[0332] Step S201 is executed by the processor 3001 in response to the operation of the analysis software 3100.

[0333] In step S2010, the processor 3001 transfers the waveform data taken into the RAM 3017 in step S303 to the parallel processing processor 3002. The waveform data is DMA transferred from the RAM 3017 to the RAM 3201 by DMA transfer, as shown in Fig. 50. At this time, the processor 3001 controls, for example, the bus controller 3005 to cause the waveform data to be DMA transferred from the RAM 3017 to the RAM 3201.

[0334] In step S2011, the processor 3001 instructs the parallel processor 3002 to execute parallel processing on the waveform data. The processor 3001 instructs the parallel processor 3002 to execute parallel processing, for example, by calling a kernel function of the parallel processor 3002. The processing executed by the parallel processor 3002 will be described later with reference to FIG. 56. The processor 3001 instructs the parallel processor 3002 to execute a matrix operation related to the AI ​​algorithm 60, for example. Waveform data corresponding to each analyte in the sample is input to the AI ​​algorithm 60. The waveform data input to the AI ​​algorithm 60 is operated on by the parallel processor 3002.

[0335] In step S2012, processor 3001 receives the results of the calculations performed by parallel processor 3002. As shown in Fig. 50, the results of the calculations are DMA transferred from RAM 3201 to RAM 3017. In step S2013, processor 3001 generates analysis results for each analyte type based on the results of the calculations performed by parallel processor 3002.

[0336] FIG. 56 is a flowchart showing the details of step S2011 in FIG.

[0337] Step S2011 is executed by the parallel processing processor 3002 based on an instruction from the processor 3001.

[0338] In step S2100, processor 3001 executing analysis software 3100 causes parallel processor 3002 to assign arithmetic operations to arithmetic units 3200. Processor 3001, for example, calls a kernel function of parallel processor 3002, causing parallel processor 3002 to assign arithmetic operations to arithmetic units 3200. As shown in FIG. 52 , for example, a matrix operation related to AI algorithm 60 is divided into multiple arithmetic operations, and each divided arithmetic operation is assigned to arithmetic units 3200. Waveform data corresponding to each analyte in a sample is input to AI algorithm 60. A matrix operation corresponding to the waveform data is divided into multiple arithmetic operations, and each divided arithmetic operation is assigned to arithmetic units 3200.

[0339] In step S2101, each arithmetic process is performed in parallel by multiple arithmetic units 3200. The arithmetic process is performed on multiple waveform data. In step S2102, the arithmetic results generated by the parallel processing by the multiple arithmetic units 3200 are transferred from RAM 3201 to RAM 3017. The arithmetic results are DMA transferred from RAM 3201 to RAM 3017, as shown in FIG.

[0340] 54, the processor 3001 of the analysis unit 300 may acquire the analysis results of the measurement items corresponding to the RBC / PLT channel (e.g., red blood cell count, hematocrit value, etc.) by applying the AI ​​algorithm 60 to the digital data based on the analog signal from the RBC / PLT detection section 4101. The processor 3001 may also acquire the analysis results of the measurement items for the HGB channel (e.g., hemoglobin level, etc.) by applying the AI ​​algorithm 60 to the digital data based on the analog signal from the HGB detection section 4102.

[0341] Next, with reference to Figures 57 and 58, another example of the configuration of a sample analyzer 4000 configured by a measurement unit 400 and an analysis unit 300 will be described.

[0342] FIG. 57 is a block diagram showing another configuration of the measurement unit 400.

[0343] In the example shown in Fig. 57, an analog optical signal processed by an analog processing unit 420 is transmitted to the analysis unit 300 via a connection port 421. A connection cable 4210 is connected to the connection port 421. The other components shown in Fig. 57 have the same configurations and functions as the measurement unit 400 of the above-described embodiment.

[0344] FIG. 58 is a block diagram showing another configuration of the analysis unit 300.

[0345] In the example shown in FIG. 58, the analysis unit 300 is connected to the measurement unit 400 via the IF unit 3006. The RAM 3017 and the bus 3003 are transmission paths having a data transfer rate of, for example, several hundred MB / s or more. The bus 3003 may also be a transmission path having a data transfer rate of 1 GB / s or more. The bus 3003 transfers data based on, for example, the PCI-Express or PCI-X standard. The configurations of the processor 3001, the parallel processing processor 3002, the storage unit 3004, and the RAM 3017, and the processes executed therein, are the same as those described above.

[0346] The analysis unit 300 includes a connection port 3007 , an A / D conversion unit 3008 , and an IF unit 3009 .

[0347] The connection port 3007 is connected to the connection port 421 of the measuring unit 400 (see FIG. 57) via a connection cable 4210. The connection cable 4210 has, for example, a number of transmission paths corresponding to the types of analog signals transmitted from the measuring unit 400 to the analysis unit 300. For example, the connection cable 4210 is configured as a twisted pair cable, and has a number of pairs of wiring corresponding to the types of analog signals transmitted to the analysis unit 300. The connection cable 4210 is preferably, for example, one meter or less in length to reduce noise during signal transmission.

[0348] The A / D conversion unit 3008 is connected to the connection port 3007. As described above, the A / D conversion unit 3008 samples the analog optical signal output from the measurement unit 400 and generates waveform data corresponding to each analyte in the specimen. The generated waveform data is stored in the storage unit 3004 or RAM 3017 via the IF unit 3009 and the bus 3003. The transmission path from the connection port 3007 to the A / D conversion unit 3008 may also have a number of wires corresponding to the types of optical signals transmitted to the analysis unit 300.

[0349] The processor 3001 and the parallel processing processor 3002 perform arithmetic processing on the waveform data stored in the storage unit 3004 or the RAM 3017. The analysis software 3100 that runs on the processor 3001 is the same as the analysis software 3100 shown in Fig. 50. By executing the analysis software 3100, the processor 3001 generates classification information regarding the types of analytes in the sample through the same operations as those described above.

[0350] Next, with reference to Figures 59 and 60, another example of the configuration of a sample analyzer 4000 configured by a measurement unit 400 and an analysis unit 300 will be described.

[0351] FIG. 59 is a block diagram showing another configuration of the measurement unit 400.

[0352] 59 includes an IF unit 4631 for transmitting waveform data generated by the A / D conversion unit 461 to the analysis unit 300. A transmission path 4632 is connected to the IF unit 4631. Other configurations and functions are the same as those of the measurement unit 400 described above.

[0353] The IF unit 4631 is, for example, an interface serving as a dedicated line with a communication bandwidth of 1 gigabit / second or more. For example, the IF unit 4631 is an interface compliant with Gigabit Ethernet, USB 3.0, or Thunderbolt 3. If the IF unit 4631 is Gigabit Ethernet, the transmission path 4632 is a LAN cable. If the IF unit 4631 is USB 3.0, the transmission path 4632 is a USB cable compliant with USB 3.0. The transmission path 4632 is, for example, a dedicated transmission path for transmitting digital data between the measurement unit 400 and the analysis unit 300.

[0354] FIG. 60 is a block diagram showing another configuration of the analysis unit 300.

[0355] 60 includes an IF unit 3010. Other configurations and functions are similar to those of the above-described analysis unit 300. The analysis unit 300 may be connected to a plurality of measurement units 400 via a plurality of IF units 3010 and a plurality of IF units 3006.

[0356] The analysis software 3100 running on the processor 3001 has the same functions as the above-described analysis software 3100. The analysis software 3100 analyzes the type of analyte in the sample by performing the same operations as those described above.

[0357] 59 and 60, an A / D conversion section 461 in the measurement unit 400 generates digital waveform data based on the analog optical signal generated in the FCM detection section 410. The waveform data is sent to the analysis unit 300 via an IF section 462, a bus 463, an IF section 4631, and a transmission path 4632.

[0358] The measuring unit 400 and the analyzing unit 300 are connected one-to-one via, for example, a transmission path 4632. In this case, the transmission path 4632 is a transmission path that does not involve the transmission of data related to devices other than the components that make up the sample analyzer 4000 (for example, the measuring unit 400 and the analyzing unit 300). The transmission path 4632 is a transmission path that is separate from, for example, an intranet or the Internet. This makes it possible to avoid bottlenecks in the communication speed of digital data transmission even when waveform data generated within the measuring unit 400 is sent to the analyzing unit 300.

[0359] Next, other configuration examples of the sample analyzer 4000 will be described with reference to Figures 61 to 65.

[0360] FIG. 61 is a block diagram showing another configuration of the sample analyzer 4000.

[0361] In this configuration example, an analysis unit 600 is provided between the measurement unit 400 and the computer 301. That is, in the configuration of Figures 61 to 65, the sample analyzer 4000 comprises the measurement unit 400, the computer 301, and the analysis unit 600. The analysis unit 600 analyzes the type of measured cells. As will be described later, the parallel processing processor 6002 in this configuration example is mounted on the sample analyzer 4000 by being incorporated into the analysis unit 600.

[0362] FIG. 62 is a block diagram showing another configuration of the measurement unit 400. As shown in FIG.

[0363] 62 differs from the configuration of FIG. 59 in that a computer 301 is connected to an IF unit 465, and an analysis unit 600 is provided between the IF unit 4631 and the computer 301. The analysis unit 600 is communicatively connected to the IF unit 4631 and the computer 301. The analysis unit 600 may be connected to a plurality of measurement units 400. The analysis unit 600 may be connected to a plurality of computers 301.

[0364] FIG. 63 is a block diagram showing the configuration of the analysis unit 600.

[0365] The analysis unit 600 includes a processor 6001, a parallel processing processor 6002, a bus 6003, a storage unit 6004, a RAM 6005, and IF units 6006 and 6007. Each unit of the analysis unit 600 is connected to the bus 6003.

[0366] The bus 6003 is, for example, a transmission line having a data transfer rate of several hundred MB / s or more. The bus 3003 may be a transmission line having a data transfer rate of 1 GB / s or more. The bus 3003 performs data transfer based on, for example, the PCI-Express or PCI-X standard. The analysis unit 600 may be connected to a plurality of measurement units 400 via a plurality of IF sections 6006. When a plurality of measurement units 400 are provided, the analysis unit 600 may be connected to each of the measurement units 400. In this case, for example, the plurality of measurement units 400 and the plurality of analysis units 600 are connected one-to-one.

[0367] FIG. 64 is a diagram showing an example of the configuration of a parallel processor 6002 that executes arithmetic processing.

[0368] The processor 6001 and the parallel processing processor 6002 have the same configurations and functions as the above-described processor 3001 and parallel processing processor 3002, respectively. The parallel processing processor 6002 includes a plurality of arithmetic units 6200 and a RAM 6201. Analysis software 6100 that analyzes the type of analyte in a sample runs on the processor 6001. The analysis software 6100 running on the processor 6001 has the same functions as the analysis software 3100 shown in FIG. 50. The analysis software 6100 analyzes the type of analyte in a sample in the same manner as the operation described in FIG. 50. The analysis software 6100 transmits classification information of the analyte in the sample to the computer 301 via the IF unit 6007.

[0369] FIG. 65 is a block diagram showing the configuration of the computer 301.

[0370] 65 has the same configuration as the analysis unit 600 in Fig. 63, except that the parallel processing processor 6002 is omitted. The computer 301 includes a processor 3501, a bus 3503, a storage unit 3504, a RAM 3505, and an IF unit 3506.

[0371] The analysis software 3100 does not have to be running on the processor 3501. The computer 301 receives the analysis results from the analysis unit 600 via the IF unit 3506. The IF unit 3506 is, for example, Ethernet or USB. The IF unit 3506 may also be an interface capable of wireless communication.

[0372] 62 to 65, the analog optical signal of the cells generated in the FCM detection section 410 is converted into digital waveform data in the A / D conversion section 461 in the measurement unit 400. The waveform data is sent to the analysis unit 600 via the IF section 462, the bus 463, the IF section 4631, and the transmission path 4632.

[0373] As described above, IF unit 4631 is a dedicated interface for connecting measuring unit 400 and analyzing unit 600, and connects measuring unit 400 and analyzing unit 600 one-to-one. In other words, transmission path 4632 is a transmission path that does not involve the transmission of data related to devices other than the components that make up sample analyzer 4000 (e.g., measuring unit 400 and analyzing unit 300). Transmission path 4632 is a transmission path separate from an intranet or the Internet. This makes it possible to avoid bottlenecks in the communication speed of waveform data transmission even when waveform data generated within measuring unit 400 is sent to analyzing unit 600.

[0374] In this case, steps S200 to S202 in FIG. 54 are executed by the analysis unit 600, and step S203 is executed by the computer 301.

[0375] Next, another example of the configuration of the sample analyzer 4000 of Figure 61 will be described with reference to Figures 66 and 67. The sample analyzer 4000 of this example comprises a measurement unit 400, a computer 301, and an analysis unit 600.

[0376] 57 , the measuring unit 400 in FIG. 66 has a computer 301 connected to the IF section 465, and an analyzing unit 600 provided between the connection port 421 and the computer 301. The analyzing unit 600 is communicatively connected to the connection port 421 and the computer 301. The measuring unit 400 transmits an analog optical signal to the analyzing unit 600 via a connection cable 4210.

[0377] Compared to the configuration of FIG. 63, the analysis unit 600 of FIG. 67 includes a connection port 6008 and an A / D conversion unit 6009 instead of the IF unit 6006.

[0378] An analog optical signal transmitted from the analysis unit 600 via the connection cable 4210 is input to the A / D conversion unit 6009 via the connection port 6008. The A / D conversion unit 6009 generates waveform data from the optical signal through processing similar to that of the A / D conversion unit 461.

[0379] The analysis unit 600 may be connected to a plurality of measurement units 400 via a plurality of connection ports 6008. When a plurality of measurement units 400 are provided, the analysis unit 600 may be connected to each of the measurement units 400. In this case, for example, the plurality of measurement units 400 and the plurality of analysis units 600 are connected one-to-one.

[0380] 66 and 67, in step S303 of FIG. 54, the analysis unit 600 generates waveform data based on the analog optical signal transmitted from the measurement unit 400. Steps S200 to S202 of FIG. 54 are executed by the analysis unit 600, and step S203 is executed by the computer 301.

[0381] Next, with reference to Figures 68 and 69, another example of the configuration of the measuring unit 400 and the analyzing unit 300 provided in the sample analyzer 4000 will be described.

[0382] 27, the measuring unit 400 in FIG. 68 has connection ports 421, 4211, 4212 instead of the A / D conversion units 461, 4611, 4612 and the IF unit 462. The analog optical signals acquired by each detection unit are transmitted to the analysis unit 300 via connection cables 4210.

[0383] Compared to the configuration in Figure 58, the analysis unit 300 in Figure 69 has three sets of connection ports 3007, A / D conversion units 3008, and IF units 3009. The three connection ports 3007 are connected to the connection ports 421, 4211, and 4212 in Figure 68, respectively.

[0384] In the configurations of FIGS. 68 and 69, in step S303 of FIG. 54, the analysis unit 300 generates waveform data based on the analog optical signal transmitted from the measurement unit 400.

[0385] Next, other configuration examples of the measuring unit 400 and the analyzing unit 300 included in the sample analyzer 4000 will be described with reference to Figures 70 and 71.

[0386] 27, measuring unit 400 in FIG. 70 includes IF unit 4631. A / D conversion units 461, 4611, and 4612 each generate waveform data based on an analog optical signal acquired by a corresponding detection unit. The waveform data corresponding to each detection unit is transmitted to analysis unit 300 via transmission path 4632.

[0387] 71 includes three IF units 3010, as compared with the configuration in FIG. 60. The three IF units 3010 are each connected to the transmission path 4632 in FIG.

[0388] Next, with reference to Figures 72 and 73, other configuration examples of the measuring unit 400 and analyzing unit 300 included in the sample analyzer 4000 will be described.

[0389] Compared to the configuration in Figure 68, the measuring unit 400 in Figure 72 has a computer 301 connected to the IF section 465, and an analysis unit 600 is disposed between the connection ports 421, 4211, 4212 and the computer 301. The analysis unit 600 is communicably connected to the connection ports 421, 4211, 4212 and the computer 301. The analysis unit 600 and the computer 301 are connected so as to be able to send and receive digital data.

[0390] Compared to the configuration in Figure 67, the analysis unit 300 in Figure 73 has three pairs of connection ports 6008 and A / D conversion units 6009. The three connection ports 6008 are connected to the connection ports 421, 4211, and 4212 in Figure 72, respectively.

[0391] Next, the data sizes of the waveform data and digital data will be described.

[0392] In this embodiment, for example, sampling is performed for one analyte in the specimen for each of an analog optical signal based on forward scattered light (FSC), an analog optical signal based on side scattered light (SSC), and an analog optical signal based on fluorescence (FL).

[0393] Examples of sampling rates include 1024 points at 10 nanosecond intervals, 128 points at 80 nanosecond intervals, or 64 points at 160 nanosecond intervals. The amount of data is, for example, 2 bytes per sampling. For each of FSC, SSC, and FL, the amount of data acquired depends on the sampling rate (for a 1024-point rate, 2 bytes x 1024 = 2048 bytes). This amount of data is per analyte in the sample.

[0394] In one measurement, for example, FSC, SSC, and FL are measured for at least 100 analytes. Alternatively, in one measurement, for example, FSC, SSC, and FL may be measured for at least 1,000 analytes. Alternatively, in one measurement, for example, FSC, SSC, and FL may be measured for approximately 10,000 to approximately 140,000 analytes. Therefore, if the number of analytes measured in one measurement is 100,000 and the sampling rate is 1024, the amount of digital data for each of FSC, SSC, and FL is 2 bytes × 1,024 × 100,000 = 204,800,000 bytes, and the total amount of FSC, SSC, and FL is 614,400,000 bytes.

[0395] Furthermore, FSC, SSC, and FL are measured for each measurement channel. Therefore, if the number of analytes measured in one measurement is 100,000, the sampling rate is 1024, and the number of measurement channels is 5, the data volume for each of FSC, SSC, and FL is 2 bytes x 1024 x 100,000 x 5 = 1,024,000,000 bytes, and the total data volume for FSC, SSC, and FL is 3,072,000,000 bytes.

[0396] As described above, the capacity of digital data per sample can range from several hundred megabytes to several gigabytes, and can reach at least 1 gigabyte depending on the number of analytes, sampling rate, and number of measurement channels.

[0397] According to this embodiment, when analyzing a huge amount of digital data ranging from several hundred megabytes to several gigabytes per sample, the analysis process using AI algorithm 60 is completed within sample analyzer 4000 as described above, and the digital data is not transmitted via the Internet or an intranet to an analysis server installed outside sample analyzer 4000. This prevents a decrease in processing capacity due to an increase in the communication load that occurs when digital data is transmitted from sample analyzer 4000 to the analysis server.

[0398] [Embodiment 7] <Configuration of waveform data analysis system> FIG. 74 is a diagram schematically showing the configuration of a waveform data analysis system according to this embodiment.

[0399] The configuration of the measurement unit 400a is the same as that of the measurement unit 400 described above. The measurement unit 400a sends a measurement sample prepared based on a specimen to a flow cell 4113. A light source 4111 (see FIG. 25) irradiates the measurement sample supplied to the flow cell 4113 with light, and light-receiving elements 4116, 4121, and 4122 (see FIG. 25) detect forward scattered light, side scattered light, and fluorescence generated from an analyte in the measurement sample. The measurement unit 400a generates waveform data from optical signals based on the forward scattered light, side scattered light, and fluorescence output from the light-receiving elements 4116, 4121, and 4122, and transmits the generated waveform data to the deep learning device 100.

[0400] The deep learning device 100 is a vendor-side device. The deep learning device 100 receives training waveform data acquired by the measurement unit 400a. The method for generating the training waveform data is as described above. The AI ​​algorithm 50 stored in the deep learning device 100 is a deep learning algorithm. The deep learning device 100 trains the AI ​​algorithm 50, which is composed of an untrained neural network, using training data, and provides the AI ​​algorithm 60, which has been trained using the training data, to the user. The AI ​​algorithm 60, which is composed of a trained neural network, is provided from the deep learning device 100 to the sample analyzer 4000 via a recording medium 98 or a communication network 99. The recording medium 98 is a computer-readable, non-transitory, tangible recording medium, such as a DVD-ROM or USB memory.

[0401] The deep learning device 100 is configured, for example, by a general-purpose computer, and performs deep learning processing based on the flowchart described below.

[0402] The sample analyzer 4000 uses an AI algorithm 60, which is composed of a trained neural network, to perform AI analysis on waveform data corresponding to an analyte.

[0403] <Hardware configuration of deep learning device> FIG. 75 is a block diagram showing the configuration of the deep learning device 100.

[0404] The deep learning device 100 includes a processing unit 10, an input unit 16, and an output unit 17.

[0405] The input unit 16 and the output unit 17 are connected to the processing unit 10 via the IF unit 15. The input unit 16 is, for example, an input device such as a keyboard or a mouse. The output unit 17 is, for example, a display device such as a liquid crystal display.

[0406] The processing unit 10 includes a CPU 11 , a memory 12 , a storage unit 13 , a bus 14 , an IF unit 15 , and a GPU 19 .

[0407] The CPU 11 performs data processing, which will be described later. The memory 12 is used as a working area for data processing. The storage unit 13 stores programs and processing data, which will be described later. The bus 14 transmits data between the various units. The IF unit 15 inputs and outputs data to and from external devices. The GPU 19 functions as an accelerator that assists the arithmetic processing (e.g., parallel arithmetic processing) performed by the CPU 11. In other words, in the following description, the processing performed by the CPU 11 also includes processing performed by the CPU 11 using the GPU 19 as an accelerator. The GPU 19 has functions equivalent to the parallel processing processors 3002 and 6002 described above. Note that instead of the GPU 19, a chip suitable for neural network calculations may be used. Examples of such chips include FPGAs, ASICs, and Myriad X (Intel).

[0408] In order to perform the processing of each step described below with reference to FIG. 77, the processing unit 10 pre-records the program according to this embodiment and the AI ​​algorithm 50 composed of the pre-trained neural network in, for example, an executable format in the storage unit 13. The executable format is, for example, a format generated by conversion from a programming language by a compiler. The processing unit 10 uses the program recorded in the storage unit 13 to perform training processing of the pre-trained AI algorithm 50.

[0409] In the following description, unless otherwise specified, the processing performed by the processing unit 10 refers to the processing performed by the CPU 11 based on the program and AI algorithm 50 stored in the storage unit 13 or memory 12. The CPU 11 uses the memory 12 as a working area to temporarily store necessary data (such as intermediate data during processing) and records data to be stored long-term, such as calculation results, in the storage unit 13 as appropriate.

[0410] <Hardware configuration of the analyzer> Sample analyzer 4000 (see FIG. 74) has the same configuration as described above and processes waveform data based on the algorithm provided by deep learning device 100. Sample analyzer 4000 may also have the functions of deep learning device 100 and use training data to train AI algorithm 50. In this case, deep learning device 100 is not required.

[0411] In order to perform the processing of each step described below in the waveform data analysis process, the sample analyzer 4000 has a program according to this embodiment and an AI algorithm 60 consisting of a trained neural network pre-stored, for example in executable form, in the memory unit 3004 (see, for example, FIG. 26) or the memory unit 6004 (see, for example, FIG. 63). The sample analyzer 4000 performs processing using the program and AI algorithm 60 stored in the memory unit 3004.

[0412] The AI ​​algorithm 60 stored in the memory units 3004, 6004 may be updated via a communication network. The deep learning device 100 transmits the AI ​​algorithm 60 to the sample analyzer 4000 via a communication network (e.g., the Internet or an intranet). The sample analyzer 4000 updates the AI ​​algorithm 60 already stored in the memory units 3004, 6004 with the received AI algorithm 60.

[0413] <Function blocks and processing procedures> (Deep learning processing) Figure 76 is a functional block diagram of the deep learning device 100.

[0414] Processing unit 10A of deep learning device 100 includes training data generation unit 101, training data input unit 102, and algorithm update unit 103. A program that causes a computer to execute deep learning processing is installed in storage unit 13 or memory 12 of processing unit 10 shown in FIG. 75, and this program is executed by CPU 11 and GPU 19, thereby realizing each functional block of processing unit 10A.

[0415] The training data database (DB) 104 and the algorithm database (DB) 105 are recorded in the storage unit 13 or memory 12 of the processing unit 10 shown in FIG. 75. The training waveform data 72a, 72b, and 72c are acquired in advance by, for example, the measurement unit 400a and stored in advance in the training data database 104. The AI ​​algorithm 50 is stored in the algorithm database 105.

[0416] FIG. 77 is a flowchart showing the processing performed by the deep learning device 100.

[0417] 77 is executed by the training data generation unit 101. The processing of step S402 is executed by the training data input unit 102. The processing of steps S403 and S405 is executed by the algorithm update unit 103.

[0418] First, the processing unit 10A acquires training waveform data 72a, 72b, and 72c. The training waveform data 72a, 72b, and 72c are waveform data based on forward scattered light, side scattered light, and fluorescence, respectively. The training waveform data 72a, 72b, and 72c may be acquired, for example, by an operator's operation, from the measurement unit 400a, from the recording medium 98, or via the communication network 99. When acquiring the training waveform data 72a, 72b, and 72c, information regarding which cell type the training waveform data 72a, 72b, and 72c represent is also acquired. The information regarding the cell type may be linked to the training waveform data 72a, 72b, and 72c, or may be input by the operator via the input unit 16.

[0419] 33, the processing unit 10A generates training data 75 from the training waveform data 72a, 72b, and 72c and the label value 77. In step S402, the processing unit 10A inputs the training data 75 to the AI ​​algorithm 50 and obtains trial results. The trial results are accumulated each time multiple pieces of training data 75 are input to the AI ​​algorithm 50.

[0420] The cell type analysis method according to this embodiment uses a convolutional neural network and employs stochastic gradient descent, so in step S403, the processing unit 10A determines whether training results for a predetermined number of trials have been accumulated. If the predetermined number of training results have been accumulated (S403: YES), the processing unit 10A proceeds to step S404. On the other hand, if the predetermined number of training results have not been accumulated (S403: NO), the processing unit 10A skips step S404.

[0421] If a predetermined number of training results have been accumulated (S403: YES), in step S404, the processing unit 10A uses the training results accumulated in step S402 to update the connection weights w of the neural network that constitutes the AI ​​algorithm 50. In the cell-type analysis method according to this embodiment, the stochastic gradient descent method is used, and therefore the connection weights w of the neural network are updated when a predetermined number of training results have been accumulated. Specifically, the process of updating the connection weights w is a process of performing calculations using the gradient descent method shown in (Equation 12) and (Equation 13) described below.

[0422] In step S405, the processing unit 10A determines whether the AI ​​algorithm 50 has been trained with a specified number of training data 75. If the AI ​​algorithm 50 has been trained with the specified number of training data 75 (S405: YES), the deep learning process ends. On the other hand, if the AI ​​algorithm 50 has not been trained with the specified number of training data 75 (S405: NO), in step S406, the processing unit 10A retrieves other training waveform data 72a, 72b, 72c, and returns the process to step S401.

[0423] Through the above-described processing, the processing unit 10A trains the AI ​​algorithm 50 and obtains the AI ​​algorithm 60.

[0424] (Neural network structure) The upper part of Figure 78 is a schematic diagram illustrating the structure of a neural network that constitutes the AI ​​algorithm 50. As described above, a convolutional neural network is used in this embodiment. The neural network of the AI ​​algorithm 50 includes an input layer 50a, an output layer 50b, and a middle layer 50c between the input layer 50a and the output layer 50b, and the middle layer 50c is composed of multiple layers. The number of layers that make up the middle layer 50c is, for example, 5 or more, preferably 50 or more, and more preferably 100 or more.

[0425] In the neural network of the AI ​​algorithm 50, multiple nodes 89 are arranged in layers and connected between layers, allowing information to propagate in only one direction, as indicated by arrow D in the figure, from the input layer 50a to the output layer 50b.

[0426] (operation at each node) The middle part of Figure 78 is a schematic diagram showing the calculations at each node 89. Each node 89 receives multiple inputs and calculates one output (z). In the example shown in the middle part of Figure 78, node 89 receives four inputs. The total input (u) received by node 89 is expressed, for example, by the following (Equation 2). Here, in this embodiment, one-dimensional matrix data is used as training data 75 and analysis data 85, so if the variables of the calculation formula correspond to two-dimensional matrix data, processing is performed to convert the variables so that they correspond to one-dimensional matrix data.

[0427]

number

[0428] Each input is multiplied by a different weight. In (Equation 2), b is a value called the bias. The output (z) of the node is the output of a predetermined function f for the total input (u) expressed in (Equation 2), and is expressed in the following (Equation 3). The function f is called the activation function.

[0429]

number

[0430] The lower part of Figure 78 is a schematic diagram showing the operations between nodes. In a neural network, nodes 89 are arranged in layers, and each node 89 outputs a result (z) expressed by (Equation 3) for the total input (u) of each node 89 expressed by (Equation 2). The output of node 89 in the previous layer becomes the input to node 89 in the next layer. In the example shown in the lower part of Figure 78, the output of node 89a in the layer on the left side of the figure becomes the input to node 89b in the layer on the right side of the figure. Each node 89b receives the output from node 89a. A different weight is applied to each connection between each node 89a and each node 89b. If the outputs of each of the multiple nodes 89a are x1 to x4, the inputs to each of the three nodes 89b are expressed by the following (Equation 4-1) to (Equation 4-3).

[0431]

number

[0432] Generalizing these (Equation 4-1) to (Equation 4-3), we obtain the following (Equation 4-4), where i = 1, ..., I and j = 1, ..., J. I is the total number of inputs, and J is the total number of outputs.

[0433]

number

[0434] Applying (Equation 4-4) to the activation function gives the output expressed by (Equation 5) below.

[0435]

number

[0436] (activation function) In the cell type analysis method according to the embodiment, a rectified linear unit function is used as the activation function. The rectified linear unit function is expressed by the following (Equation 6).

[0437]

number

[0438] (Equation 6) is a linear function of z=u, where the part of u<0 is set to u=0. In the example shown in the lower part of Figure 78, the output of the node j=1 is expressed by the following (Equation 6).

[0439]

number

[0440] (Neural network training) If a function expressed using a neural network is denoted by y(x:w), the function y(x:w) changes when the parameter w of the neural network is changed. Adjusting the function y(x:w) so that the neural network selects a parameter w that is more suitable for the input x is called training or learning of the neural network. Assume that multiple pairs of input and output of a function expressed using a neural network are given. If the desired output for a certain input x is d, the input-output pairs are given as {(x1, d1), (x2, d2), ..., (xn, dn)}. The set of pairs represented by (x, d) is called training data. Specifically, as shown in Figure 33, the set of waveform data 72a, 72b, and 72c is training data 75.

[0441] Learning a neural network means adjusting the weight w so that, for any input / output pair (xn, dn), when an input xn is given, the output y(xn:w) of the neural network becomes as close as possible to the output dn, as shown in the following equation.

[0442]

number

[0443] An error function is a measure of the closeness between a function expressed using a neural network and training data. The error function is also called a loss function. The error function E(w) used in the cell type analysis method according to the embodiment is expressed by the following formula (7). Formula (7) is called cross entropy.

[0444]

number

[0445] A method for calculating the cross entropy of (Equation 7) will be explained. In the output layer 50b of the neural network used in the cell type analysis method according to the embodiment, i.e., in the final layer of the neural network, an activation function is used to classify the input x into a finite number of classes according to its content. The activation function is called a softmax function, and is expressed by the following (Equation 8). It is assumed that the output layer 50b has the same number of nodes as the number of classes k. The total input u of each node k (k=1, ..., K) of the output layer L is calculated by dividing u by uk from the output of the previous layer L-1. (L) As a result, the output of the k-th node in the output layer is expressed as follows (Equation 8).

[0446]

number

[0447] (Equation 8) is the softmax function. The sum of the outputs y1, ..., yK determined by (Equation 8) is always 1.

[0448] If each class is represented as C1,...,CK, the output yK of node k in the output layer L (i.e., uk (L) ) represents the probability that a given input x belongs to class CK. Input x is classified into the class that maximizes the probability expressed by the following (Equation 9).

[0449]

number

[0450] In neural network training, the function represented by the neural network is regarded as a model of the posterior probability of each class, and under such a probability model, the likelihood of the weight w for the training data is evaluated, and the weight w that maximizes the likelihood is selected.

[0451] The target output by the softmax function in (Equation 8) is set to 1 only if the output is the correct class, and 0 otherwise. If the target output is expressed in vector form as dn = [dn1, ..., dnK], for example, if the correct class of input xn is C3, only the target output dn3 will be 1, and the other target outputs will be 0. When encoded in this way, the posterior distribution is expressed by the following (Equation 10).

[0452]

number

[0453] The likelihood L(w) of weight w for training data {(xn, dn)} (n = 1, ..., N) is expressed as follows (Equation 11): Taking the logarithm of the likelihood L(w) and inverting the sign yields the error function of (Equation 7).

[0454]

number

[0455] Learning refers to minimizing the error function E(w) calculated based on training data with respect to the parameter w of the neural network. In the cell-type analysis method according to the embodiment, the error function E(w) is expressed by Equation 7.

[0456] Minimizing the error function E(w) with respect to the parameter w is the same as finding a local minimum of the error function E(w). The parameter w is the weight of the connection between nodes. The minimum of the weight w is found by iterative calculations that start with an arbitrary initial value and repeatedly update the parameter w. An example of such calculations is the gradient descent method.

[0457] The gradient descent method uses a vector expressed by the following (Equation 12).

[0458]

number

[0459] Gradient descent involves repeatedly moving the current value of the parameter w in the negative gradient direction (i.e., -∇E). (t) The weight after the movement is w (t+1) Then, the calculation by the gradient descent method is expressed by the following (Equation 13): The value t indicates the number of times the parameter w is moved.

[0460]

number

[0461] The symbol shown in the following (Equation 14) used in (Equation 13) is a constant that determines the magnitude of the update amount of the parameter w, and is called a learning coefficient.

[0462]

number

[0463] By repeating the calculation expressed by (Equation 13), the error function E(w (t) ) decreases and the parameter w reaches a minimum.

[0464] The calculation using (Equation 13) may be performed on all training data (n=1, ..., N) or only on a portion of the training data. A gradient descent method performed on only a portion of the training data is called a stochastic gradient descent method. The cell type analysis method according to the embodiment uses the stochastic gradient descent method.

[0465] [Effects of the embodiment] The measurement unit 400 includes a plurality of first sample preparation units that prepare first measurement samples based on the specimen and a first reagent, a second sample preparation unit that prepares second measurement samples based on the specimen and a second reagent, and an FCM detection unit 410 (optical detection unit) that acquires a first optical signal from the first measurement samples and a second optical signal from the second measurement samples. The analysis unit 300 analyzes first data corresponding to the first optical signals and second data corresponding to the second optical signals. The analysis unit 300 performs analysis of a first measurement item for a first measurement sample by AI analysis (first analysis operation) that processes first data using an artificial intelligence algorithm, performs analysis of a second measurement item for the first measurement sample by at least one of AI analysis (first analysis operation) and computational processing analysis (second analysis operation) that processes a first representative value from the first data that corresponds to the characteristics of the analyte, and performs analysis of a second measurement sample by computational processing analysis (third analysis operation) that processes a second representative value from the second data that corresponds to the characteristics of the analyte.

[0466] When the measurement unit 400 is configured as shown in FIG. 29, the first reagent is used, for example, to classify white blood cells in a sample, and more specifically, to classify cells in a sample into neutrophils, lymphocytes, monocytes, and eosinophils. The first reagent is, for example, a WDF hemolyzing agent and a WDF staining solution. The first measurement sample is, for example, a WDF measurement sample. The first sample preparation unit is, for example, reaction chamber 440a of the WDF channel. The second reagent is, for example, a hemolyzing agent and a staining solution for a measurement channel other than the WDF channel. The second measurement sample is, for example, a measurement sample for a measurement channel other than the WDF channel. The second sample preparation unit is, for example, reaction chambers 440b, 440c, and 440d of a measurement channel other than the WDF channel.

[0467] The first and second optical signals are analog signals output from the light-receiving element based on forward scattered light, side scattered light, and fluorescence. The first and second optical signals have regions corresponding to each analyte (e.g., cell) in the sample and are signals reflecting the presence of the analyte in the sample. The first and second data are digital data corresponding to the intensities of the first and second optical signals, respectively, based on light emitted from each analyte (e.g., cell). The first and second data are waveform data generated corresponding to the regions of the first and second optical signals, respectively. In other words, the first and second data correspond to the first and second optical signals, respectively, acquired while the analyte passes through the position illuminated by light from the light source 4111.

[0468] The first measurement item is, for example, an item analyzed based on measurement of the WDF channel. An example of the first measurement item is an item related to white blood cells. The second measurement item is, for example, an item analyzed based on a measurement channel other than the WDF channel. An example of the second measurement item is an item related to platelets or reticulocytes.

[0469] The first, second, and third analysis operations are operations for determining the type of analyte. The first representative value corresponding to the characteristic of the analyte is, for example, a value such as a peak value, area, or width obtained from the first data (waveform data) corresponding to the analyte (see FIG. 3), and the second representative value corresponding to the characteristic of the analyte is, for example, a value such as a peak value, area, or width obtained from the second data (waveform data) corresponding to the analyte (see FIG. 3).

[0470] According to this configuration, the analysis processing of data corresponding to optical signals obtained from a sample is divided between a first analysis operation (AI analysis) using an artificial intelligence algorithm and second and third analysis operations (computational analysis) that process first and second representative values ​​corresponding to the characteristics of the analyte, thereby reducing the load on the analysis unit 300, which is a computer that processes data, compared to when data corresponding to optical signals are analyzed uniformly using only an artificial intelligence algorithm.

[0471] Furthermore, the analysis of the second measurement item for the first measurement sample is performed by at least one of a first analysis operation (AI analysis) that processes using an artificial intelligence algorithm, and a second analysis operation (computational analysis) that processes a first representative value corresponding to the characteristics of the analyte. As a result, even if it is difficult to analyze the second measurement item using the second analysis operation, for example, it is possible to appropriately perform a highly accurate analysis using the first analysis operation.

[0472] Furthermore, since multiple first sample preparation units are provided, multiple first measurement samples can be prepared in parallel based on multiple specimens, thereby improving the throughput of specimen analysis.

[0473] The representative values ​​of the waveform data (first and second data) that can be processed in the computational analysis are identified based on the magnitude of the waveform data (first and second data). Specifically, the representative values ​​of the waveform data, such as peak value, area, and width, are identified based on the magnitude of the waveform data. This allows for smooth identification of the representative values.

[0474] The first and second optical signals have regions corresponding to each of the analytes in the sample. The analysis unit 300 identifies first and second representative values ​​that can be subjected to computational analysis based on waveform data (first and second data) corresponding to each of the regions of the first and second optical signals. In this way, because the optical signals include regions corresponding to each of the analytes, representative values ​​such as peak values, areas, and widths corresponding to each analyte can be smoothly identified based on the waveform data corresponding to each region of the first and second optical signals.

[0475] The first optical signal has regions corresponding to each of the analytes in the sample. The analysis unit 300 inputs waveform data (first data) corresponding to each of the regions of the first optical signal into an artificial intelligence algorithm. In this way, since the first optical signal includes regions corresponding to each of the analytes, inputting waveform data corresponding to each region of the first optical signal into the artificial intelligence algorithm allows for smooth execution of AI analysis.

[0476] As described above, when the first and second optical signals have regions corresponding to the respective analytes in the sample, the measurement unit 400 acquires waveform data (first and second data) based on signals greater than a predetermined threshold corresponding to the intensities of the first and second optical signals, as shown in the upper diagram of Figure 3. This configuration makes it possible to accurately acquire waveform data corresponding to the respective analytes.

[0477] As shown in Figure 37, in analyzing a WDF measurement sample (first measurement sample) using an artificial intelligence algorithm, the analysis unit 300 classifies cells in the sample as at least either nucleated red blood cells or basophils. Normally, classification of nucleated red blood cells and basophils must be performed in the WNR channel. In contrast, AI analysis using an artificial intelligence algorithm (first analysis operation) allows accurate classification of nucleated red blood cells and basophils based on a WDF measurement sample (first measurement sample) prepared for classifying white blood cells.

[0478] 37, the analysis unit 300 classifies cells in a specimen into nucleated red blood cells and basophils by AI analysis (first analysis operation) based on waveform data (first data) obtained from a WDF measurement specimen (first measurement specimen). Furthermore, the analysis unit 300 classifies cells in the specimen into neutrophils, lymphocytes, monocytes, and eosinophils by computational analysis (second analysis operation) based on the waveform data (first data) obtained from the WDF measurement specimen (first measurement specimen). By using AI analysis to classify nucleated red blood cells and basophils, which are difficult to classify with high accuracy by computational analysis, while also using computational analysis to classify neutrophils, lymphocytes, monocytes, and eosinophils, which can be classified with high accuracy by computational analysis, the load on the analysis unit 300 can be reduced compared to when all classifications are performed by AI analysis.

[0479] The WDF channel has multiple reaction chambers 440a (first sample preparation units) that concurrently prepare multiple WDF measurement samples (first measurement samples) corresponding to different samples. This configuration can improve the throughput of sample analysis.

[0480] 37, the analysis unit 300 performs an analysis of the measurement items (first measurement items) related to nucleated red blood cells and basophils on the WDF measurement sample (first measurement sample) using AI analysis (first analysis operation). Furthermore, the analysis unit 300 analyzes the measurement items (second measurement items) related to lymphocytes, monocytes, eosinophils, and neutrophils by determining the type of analyte that did not fall under the measurement items (first measurement items) related to nucleated red blood cells and basophils in the AI ​​analysis (first analysis operation) using computational analysis (second analysis operation). With this configuration, even if it is difficult to classify white blood cells into five types using only computational analysis based on the WDF measurement sample, by performing both AI analysis and computational analysis, it is possible to classify white blood cells into five types based on the WDF measurement sample.

[0481] 38, the analysis unit 300 performs an analysis of measurement items (second measurement items) related to lymphocytes, monocytes, and eosinophils in a WDF measurement sample (first measurement sample) by computational analysis (second analysis operation). Furthermore, the analysis unit 300 analyzes measurement items (first measurement items) related to neutrophils and basophils by determining the type of analyte that did not fall under the second measurement item in the computational analysis (second analysis operation) by AI analysis (first analysis operation). Even in this configuration, by performing both AI analysis and computational analysis, it is possible to classify white blood cells into five groups based on the WDF measurement sample.

[0482] The embodiments of the present invention can be modified in various ways as appropriate within the scope of the technical idea defined in the claims. [Explanation of symbols]

[0483] 60 AI Algorithms (Artificial Intelligence Algorithms) 440a Reaction chamber (first sample preparation section) 440b, 440c, 440d Reaction chamber (second sample preparation section) 80a, 80b, 80c Optical signals (first and second optical signals) 82a, 82b, 82c Waveform data (first and second data) 300, 600 analysis units 400 measurement units 410 FCM detector (optical detector) 3001, 6001 processor (host processor) 3002, 6002 parallel processing processor 4000 Sample Analyzer 4111 Light source 4113 Flow Cell 4116, 4121, 4122 Photodetector

Claims

1. A sample analyzer for analyzing an analyte in a sample, comprising: a measurement unit including a plurality of first sample preparation units that prepare first measurement samples based on the specimen and a first reagent, a second sample preparation unit that prepares second measurement samples based on the specimen and a second reagent, and an optical detection unit that acquires a first optical signal from the first measurement samples and acquires a second optical signal from the second measurement samples; an analysis unit for analyzing first data corresponding to the first optical signal and second data corresponding to the second optical signal; The analysis unit performing an analysis of a first measurement item for the first measurement sample by a first analysis operation that processes the first data using an artificial intelligence algorithm; performing an analysis of a second measurement item for the first measurement sample by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to a characteristic of the analyte from the first data; analyzing the second measurement sample by a third analysis operation that processes a second representative value corresponding to a characteristic of the analyte from the second data; A sample analyzer characterized by:

2. The sample analyzer according to claim 1 , wherein the plurality of first sample preparation sections concurrently prepare a plurality of first measurement samples, each corresponding to a different sample.

3. The sample analyzer according to claim 1 , wherein the analysis unit, in the second analysis operation, identifies the first representative value based on the first data, and processes the identified first representative value.

4. The sample analyzer according to claim 3 , wherein the analysis unit, in the second analysis operation, identifies the first representative value based on the magnitude of the first data.

5. the first optical signal has a region corresponding to each of the analytes in the sample; The sample analyzer according to claim 3 , wherein the analysis unit identifies the first representative value based on the first data corresponding to each of the regions of the first optical signal.

6. The sample analyzer of claim 5 , wherein the analysis unit identifies a peak value in the region of the first data as the first representative value.

7. the first optical signal has a region corresponding to each of the analytes in the sample; The sample analyzer according to claim 1 , wherein the analysis unit inputs the first data corresponding to each of the regions of the first optical signal into the artificial intelligence algorithm in the first analysis operation.

8. The sample analyzer according to claim 1 , wherein the measurement unit acquires the first data based on a signal that is greater than a predetermined threshold corresponding to the intensity of the first optical signal.

9. the measurement unit obtains the first representative value based on the first optical signal; The sample analyzer according to claim 1 , wherein the analysis unit processes the first representative value acquired by the measurement unit in a second analysis operation.

10. The sample analyzer according to claim 1 , wherein the first optical signal is a signal that reflects the presence of an analyte in the sample.

11. 11. The sample analyzer according to claim 1, wherein the optical detection unit includes a light source, a flow cell, and a photodetector, and irradiates the flow cell with light to detect light generated from an analyte in the sample flowing through the flow cell.

12. The sample analyzer of claim 11 , wherein the first and second data correspond to the first and second optical signals, respectively, acquired while the analyte passes through the light irradiation position.

13. The sample analyzer according to claim 1 , wherein the analysis unit analyzes the first data by a convolution operation using the artificial intelligence algorithm.

14. The sample analyzer according to claim 1 , wherein the analysis unit analyzes the first data using a matrix operation based on the artificial intelligence algorithm.

15. The sample analyzer according to claim 14, wherein the analysis unit executes the matrix calculations by the artificial intelligence algorithm in parallel processing by a parallel processor.

16. The sample analyzer of claim 15, wherein the analysis unit executes the first analysis operation by the parallel processing processor, and executes the second and third analysis operations by a host processor of the parallel processing processor.

17. The sample analyzer of claim 1 , wherein the artificial intelligence algorithm is a deep learning algorithm.

18. 18. The sample analyzer of claim 1, wherein the first sample preparation unit prepares the first measurement sample based on the sample and the first reagent used to classify white blood cells in the sample.

19. 20. The sample analyzer of claim 18, wherein the first sample preparation unit prepares the first measurement sample based on the sample and the first reagent used to classify cells in the sample into neutrophils, lymphocytes, monocytes, and eosinophils.

20. The sample analyzer of claim 18 or 19, wherein the analysis unit classifies cells in the specimen as at least one of nucleated red blood cells and basophils in the analysis of the first measurement sample using the artificial intelligence algorithm.

21. The analysis unit classifying cells in the specimen into nucleated red blood cells and basophils by the first analyzing operation based on the first data obtained from the first measurement specimen; 21. The sample analyzer of claim 18, wherein the second analysis operation classifies cells in the specimen into neutrophils, lymphocytes, monocytes, and eosinophils based on the first data obtained from the first measurement sample.

22. A sample analyzer according to any one of claims 1 to 21, wherein the analysis unit performs an analysis of the first measurement item on the first measurement sample through the first analysis operation, and analyzes the second measurement item by determining the type of analyte that did not correspond to the first measurement item through the first analysis operation through the second analysis operation.

23. A sample analyzer according to any one of claims 1 to 22, wherein the analysis unit performs an analysis of the second measurement item on the first measurement sample by the second analysis operation, and analyzes the first measurement item by determining the type of analyte that did not correspond to the second measurement item by the second analysis operation by the first analysis operation.

24. 1. A sample analysis method for analyzing an analyte in a sample, comprising: preparing a first measurement sample based on the specimen and a first reagent, and preparing a second measurement sample based on the specimen and a second reagent; acquiring a first optical signal from the first measurement sample and a second optical signal from the second measurement sample; analyzing first data corresponding to the first optical signal and second data corresponding to the second optical signal; In the analyzing step, performing an analysis of a first measurement item for the first measurement sample by a first analysis operation that processes the first data using an artificial intelligence algorithm; performing an analysis of a second measurement item for the first measurement sample by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to a characteristic of the analyte from the first data; analyzing the second measurement sample by a third analysis operation that processes a second representative value corresponding to a characteristic of the analyte from the second data; A sample analysis method characterized by:

25. A program for causing a computer to execute a process for analyzing an analyte in a sample, analyzing first data corresponding to a first optical signal acquired from a first measurement sample prepared based on the specimen and a first reagent, and second data corresponding to a second optical signal acquired from a second measurement sample prepared based on the specimen and a second reagent; The process comprises: performing an analysis of a first measurement item for the first measurement sample by a first analysis operation that processes the first data using an artificial intelligence algorithm; performing an analysis of a second measurement item for the first measurement sample by at least one of the first analysis operation and a second analysis operation that processes a first representative value corresponding to a characteristic of the analyte from the first data; analyzing the second measurement sample by a third analysis operation that processes a second representative value corresponding to a characteristic of the analyte from the second data; A program characterized by:

Images