[0001]The present invention relates to a blood analyzer having a function of determining whether a blood sample has a possibility of being malaria positive, a computer program and a method for determining the possibility of a blood sample being malaria positive.
Background technology
[0002]
　In the present specification, "malaria positive" means a state in which Plasmodium is present in blood, and "malaria negative" means a state in which Plasmodium is not present in blood. Hereinafter, the determination as to whether or not the blood sample has the possibility of being malaria positive is also referred to as "malaria determination".
[0003]
　Conventionally, a blood analyzer having a function of determining malaria on a blood sample has been known (Patent Documents 1, 2, etc.). In the blood analyzers described in Patent Documents 1 and 2, malaria protozoa present in malaria-positive blood are regarded as particles having a specific volume, and how much volume is contained in the sample solution prepared from the patient's blood sample. The volume frequency distribution of the number of particles present is analyzed, and malaria is automatically determined based on the volume frequency distribution.
Prior art literature
Patent documents
[0004]
Patent Document 1: Japanese
Patent Application Laid-Open No. 2007-525674 Patent Document 2: Japanese Patent Application Laid-Open No. 2007-024844
Patent Document 3: Japanese Patent Application Laid-Open No. 11-101798
Outline of the invention
Problems to be solved by the invention
[0005]
　However, when the present inventors have examined in detail a blood analyzer having a function of performing a conventional malaria determination as described above, such a blood analyzer has a relatively high function for performing a malaria determination. It was essential to have an expensive configuration, and therefore it was not suitable for widespread use of the malaria determination function in areas where it is needed on a daily basis. Therefore, it was found that a determination method capable of performing malaria determination even with an inexpensive device having fewer functions is required.
[0006]
　An object of the present invention is to provide a blood analyzer capable of solving the above-mentioned problems, determining malaria with an inexpensive configuration, and providing a computer program for the blood analyzer. It is in.
Means to solve problems
[0007]
　The main configuration of the present invention is as follows.
[1] A blood analyzer
　having a measurement value receiving unit that receives a test measurement value of a blood sample to be analyzed, and the
　　test measurement value includes a CRP value based on a CRP parameter value and a blood cell count value. It is a measurement value for each of a predetermined blood test items including, and the
　　measurement value receiving unit has a CRP value receiving unit that receives a CRP parameter value in a sample solution containing the blood sample, and the
　blood analyzer. the inspection measurements for the blood test items other than CRP values, and / or, based on specific test measurement for CRP values, has a determination section for determining the possibility of malaria positive of said blood sample,
wherein Blood analyzer.
[2] The measurement value receiving unit has a volume value receiving unit that receives a volume measurement value of particles in a sample solution containing the blood sample, and the
　determination unit has a volume frequency distribution based on the volume measurement value. When the frequency peak portion is present within the predetermined volume range in the above, or when the CRP value based on the CRP parameter value is equal to or higher than a predetermined threshold value, the blood sample is highly likely to be malaria positive.
The blood analyzer according to the above [1] for determination .
[3] A leukocyte measurement chamber is further provided as a first particle measurement unit.
　In the leukocyte measurement chamber, the supplied blood sample is subjected to a hemolytic treatment to dissolve erythrocytes and diluted to form the sample solution, and the volumes of leukocytes and other particles in the sample solution are measured. ,
　volume value receiving unit, said receiving said volumetric measurements from the first particle measuring unit,
blood analyzer according to the above [2].
[4] The blood
　according to the
above [2] or [3] , further comprising a CRP measuring unit for measuring the CRP parameter value, and the CRP value receiving unit receives the CRP parameter value from the CRP measuring unit. Analysis equipment.
[5] When the frequency peak portion exists within a predetermined volume range in the volume frequency distribution based on the volume measurement value and the CRP value based on the CRP parameter value is equal to or higher than a predetermined threshold value, the above.
The blood analyzer according to any one of [2] to [4] above, wherein the determination unit determines that the blood sample has a high possibility of being malaria positive .
[6] The determination unit uses
　important test items related to malaria positive, which are identified from predetermined blood test items by the following first trained model (M1), and , 　The first trained model (M1) is configured as 　teacher data
　to determine the possibility of malaria positive in the blood sample based on the test measurements for the important test items. ,

　　A test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target, and a blood sample whose test measurement value (A1) is a malaria positive blood sample. The
　　test measurement value (B1) obtained for the predetermined blood test item using the information (A2) that the blood test item is, and the blood sample (B) that is known to be negative for malaria, and the test measurement value. (B1) and information (B2) that is of the blood samples of malaria-negative
with,
　that which blood test items of said predetermined blood test items to machine learning or have relevance to malaria positive
The blood analyzer according to the above [1] , which is formed by the blood analyzer.
[7] The important test item is
　whether or not there is a frequency peak in the volume range of 25 fL to 45 fL in the volume frequency distribution of leukocytes in the
　blood sample to be analyzed, the number of leukocytes in the blood sample to be analyzed, and
　analysis. The blood
　analyzer according to
the
above [6] , which comprises 1 or more selected from the group consisting of the number of white blood cells in the blood sample to be analyzed and the CRP value of the blood sample to be analyzed.
[8] The determination unit uses
　the degree of relevance of each of the predetermined blood test items to malaria positivity specified by the second trained model (M2) below.
　Based on the test measurements for the predetermined blood test item and the degree of relevance identified, the blood sample is configured to determine the likelihood of being malaria positive
　, wherein the second. In the trained model (M2) of the above, as
　teacher data,
　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be positive for malaria as a measurement target, and the test measurement value (A1). Information (A2) that the test measurement value (A1) is for a blood sample positive for malaria
　　and a blood sample (B) known to be negative for malaria are obtained for the predetermined blood test items as measurement targets. inspection measurements and (B1), and information (B2) of the test measurement (B1) is of a blood sample of malaria-negative
using,
　what blood test items of said predetermined blood test item what
The blood analyzer according to the above [1] , which is formed by machine-learning whether or not it is related to malaria positivity to a certain degree .
[9] A step of accepting a test measurement value of a blood sample to be analyzed, wherein the test measurement value is a measurement value for each of a predetermined blood test item including a CRP value and a blood cell count value. The step and the step
　of determining the possibility of malaria positive in the blood sample based on the test measurement value for a specific blood test item other than the CRP value and / or the test measurement value for the CRP value.
A computer program that you want your computer to run.
[10] 　Specific to the malaria protozoa in the
　step of accepting the volume measurement value of the particles in the sample solution containing the blood sample, the step of accepting the CRP parameter value in the sample solution, and the
volume frequency distribution based on the volume measurement value. determining whether the frequency peaks in the volume range are present,
　CRP value based on the CRP parameter value, determining whether the threshold value or more predetermined
　said frequency peak portion The computer program according to [9] above
,
wherein the computer is made to perform a step of determining that the blood sample is likely to be malaria positive if it is present or the CRP value is equal to or higher than the threshold value. ..
[11] Further, a computer program for causing a computer to execute a step of determining that the blood sample is likely to be malaria positive when the frequency peak portion is present and the CRP value is equal to or higher than the threshold value. The computer program according to the above [10].
[12] The determination step uses
　important test items related to malaria positivity, which are identified from predetermined blood test items by the following first trained model (M1). In addition, 　the first trained model (M1)
　has a step of determining the possibility of malaria positive in the blood sample based on the test measurement values ​​for the important test items
.
　As teacher data,
　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target, and the test measurement value (A1) are malaria positive. Information (A2) that
　　it belongs to the blood sample of the above, and the test measurement value (B1) obtained for the predetermined blood test item with the blood sample (B) known to be negative for malaria as the measurement target. and information (B2) of the test measurement (B1) is of a blood sample of malaria-negative
using,
　what blood test items of said predetermined blood test items have relevance to malaria positive
The computer program according to the above [9] , which is formed by machine learning .
[13] The determination step uses
　the degree of relevance of each of the predetermined blood test items to the positive malaria identified by the second trained model (M2) below, and the
　predetermined blood test item.
　The second trained model (M2) comprises the step of determining the possibility of malaria positive in the blood sample based on the test measurement value of the above and the degree of the identified relevance. As
　teacher data,
　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be positive for malaria as a measurement target, and the test measurement value (A1) are Information (A2) that it belongs to a blood sample positive for malaria,
　　A test measurement value (B1) obtained for the predetermined blood test item using a blood sample (B) known to be malaria-negative as a measurement target, and a blood sample whose test measurement value (B1) is malaria-negative. It is 　formed by machine learning which blood test item among the predetermined blood test items is related to malaria positivity to what degree
by using the information (B2) that the
blood test item is a thing. ,
The computer program according to the above [9].
[14] It has a step of preparing a test measurement value of a blood sample to be analyzed, and the test measurement value is a measurement value for each of a predetermined blood test items including a CRP value and a blood cell count value. and,
　test measurement for a particular blood test items other than CRP values, and / or, based on the test measurement for CRP values, with the step of determining the possibility of malaria positive of the blood sample,
the blood Analysis method.
[15] In the determination step,
　the important test items related to malaria positive, which are identified from the predetermined blood test items by the following first trained model (M1), are used. In addition, 　the first trained model (M1)
　has a step of determining the possibility of malaria positive in the blood sample based on the test measurement values ​​for the important test items, and
the first trained model (M1) is used as
　teacher data.
　　A test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target, and a blood sample whose test measurement value (A1) is a malaria positive blood sample. The
　　test measurement value (B1) obtained for the predetermined blood test item using the information (A2) that the blood test item is, and the blood sample (B) that is known to be negative for malaria, and the test measurement value. (B1) and information (B2) that is of the blood samples of malaria-negative
with,
　that which blood test items of said predetermined blood test items to machine learning or have relevance to malaria positive
The blood analysis method according to the above [14] , which is formed by .
[16] The determination step uses
　the degree of relevance of each of the predetermined blood test items to the positive malaria identified by the second trained model (M2) below for the
　predetermined blood test item.
　The second trained model (M2) comprises the step of determining the possibility of malaria positive in the blood sample based on the test measurement value of the above and the degree of the identified relevance. As
　teacher data,
　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be positive for malaria as a measurement target, and the test measurement value (A1) are Information (A2) that it belongs to a blood sample positive for malaria,
　　A test measurement value (B1) obtained by measuring the predetermined blood test item using a blood sample (B) known to be malaria-negative as a measurement target, and the test measurement value (B1) are malaria-negative. 　Formed by machine learning which blood test item among the predetermined blood test items is related to malaria positive to what degree
using the information (B2) that it belongs to a blood sample. The blood analysis method according to
the
above [14].
The invention's effect
[0008]
　According to the present invention, it is possible to determine malaria even with an inexpensive device configuration.
　Conventionally, it has not been found that it is effective to use the CRP value and the volume frequency distribution in a complementary manner for the malaria determination. Therefore, the volume frequency distribution and the CRP value of the particles are used as the conditions for the malaria determination. It was not associated with each other. On the other hand, in the present invention, it has been found that if the malaria determination is performed based on both the volume frequency distribution of the particles and the CRP value, the accuracy or reliability of the malaria determination is synergistically improved.
[0009]
　In a preferred embodiment of the present invention, (correlation between each test measurement value for a given blood test item of a blood sample known to be positive for malaria and the fact that it is positive for malaria), and (negative for malaria). The correlation between each test measurement value for a predetermined blood test item of a blood sample known to exist and the fact that it is malaria negative) is machine-learned by artificial intelligence, and the predetermined blood test item and malaria positive Build a trained model of association with (which of the given blood test items is associated with or to the extent that it is associated with malaria positivity) do. Then, according to the trained model, a specific blood test item (important test item) related to positive malaria is selected from predetermined blood test items (that is, various blood test items). The specific test measurement value is used for determining malaria. As a result, in addition to the blood test items that medical professionals such as doctors and technicians have traditionally used to determine malaria as being empirically important, new blood test items can be added to determine malaria positivity. This makes it possible and improves the accuracy of malaria determination.
A brief description of the drawing
[0010]
FIG. 1 is a block diagram showing an example of a preferable configuration of the blood analyzer of the present invention.
FIG. 2 is a flowchart showing the operation of the blood analyzer of the present invention. The flowchart is also a flowchart for explaining the structure of the computer program of the present invention.
[Fig. 3] Fig. 3 is a graph showing an example of a frequency peak portion appearing due to the presence of Plasmodium malaria in a volume frequency distribution diagram of particles in a sample solution (a blood sample subjected to hemolysis treatment and diluted). It is a figure. The figure shows the frequency distribution in the region where the volume of particles is small.
FIG. 4 is a scatter plot showing the relationship between the height of the frequency peak portion that appears due to the presence of Plasmodium malaria in the volume frequency distribution diagram of the particles in the sample solution and the CRP value of the sample solution. ..
[Fig. 5] Fig. 5 is a graph showing that the height of the frequency peak portion that appears due to the presence of Plasmodium changes depending on the type of Plasmodium in the volume frequency distribution map of the particles in the sample solution. be.
FIG. 6 is a graph showing an example of a frequency peak portion appearing due to dengue virus infection in a volume frequency distribution diagram of particles in a sample solution prepared from a dengue virus-positive blood sample. In the figure, as in FIG. 3, the frequency distribution in the region where the volume of the particles is small is shown.
FIG. 7 is a box-and-whisker plot showing the CRP values ​​of a malaria-positive blood sample and a dengue-positive blood sample. FIG. 7 (a) is a graph of a malaria-positive blood sample, and FIG. 7 (b) is a graph of a dengue fever-positive blood sample.
[Fig. 8] Fig. 8 shows [the height of the frequency peak appearing in the volume range of 25 fL to 45 fL in the graph of the volume frequency distribution obtained by counting the leukocytes of malaria-positive blood] and [the unit volume of the blood. It is a graph which shows the correlation with the number of malaria protozoa contained in the hit.
FIG. 9 is a flowchart showing the steps of the blood analysis method of the present invention. The flowchart is also a flowchart for explaining a preferred embodiment of the computer program of the present invention.
Mode for carrying out the invention
[0011]
[Structural Example of Blood Analyzer of the
　Present Invention ] First, the blood analyzer of the present invention (hereinafter, also referred to as the device) will be described in detail with reference to examples.
　As shown in FIG. 1 as an example of the configuration, the apparatus includes at least a measured value receiving unit 410 and a determination unit 430. The measurement value receiving unit 410 is a portion configured to receive a predetermined test measurement value for a blood sample to be analyzed, and includes a CRP value receiving unit. In the example of FIG. 1, the measured value receiving unit 410 has a volume value receiving unit and a CRP value receiving unit.
[0012]
　The test measurement value is a measurement value for each of a predetermined blood test item (various conventionally known blood test items). In a preferred embodiment of the present invention, the CRP value and the blood cell count value are included in the predetermined blood test item, but the CRP value may not be included. The test measurement value may be the value of the electric signal as it is output from the blood measuring device, and the value obtained by arithmetically processing the value, and further, the value having a unit that the medical staff can understand. It may be a value converted to.
[0013]
　The determination unit 430 is a portion that determines malaria of the blood sample to be analyzed. The determination unit is configured to perform the malaria determination based on one or both of (test measurement values ​​for blood test items other than CRP value) and (test measurement values ​​for CRP value). Is.
[0014]
　In the example shown in FIG. 1, the device is a device configured to receive a sample container X1 containing a blood sample M1 at a predetermined position and measure and analyze the blood sample M1 in the sample container X1. A measurement mechanism unit 10 (including a first particle measurement unit 11 and a CRP measurement unit 13) for performing measurement and measurement of a CRP value, and a control unit 20 (analysis) for controlling the measurement and analyzing the measured value. (Including part 40). In the example of FIG. 1, the first particle measurement unit 11 is a white blood cell measurement chamber (hereinafter, also referred to as WBC chamber), and the CRP measurement unit 13 is a CRP measurement chamber (hereinafter, also referred to as CRP chamber). Further, in the example of FIG. 1, as a preferred embodiment, a second particle measuring unit 12 is further provided, and the second particle measuring unit 12 is an red blood cell measuring chamber (also referred to as an RBC chamber). A detailed description of the entire device will be described later. Even in the configuration shown in FIG. 1, it is important that the apparatus has at least a volume value receiving unit, a CRP value receiving unit, and a determination unit, which will be described later, in the analysis unit.
[0015]
　The first particle measuring unit (WBC chamber) 11 is configured to measure the volume of each particle (particle containing leukocytes) in the sample liquid to be measured and output the volume measurement value of each particle. .. The sample solution measured by the first particle measuring unit 11 contains a hemolyzed blood sample M1, and in a preferred embodiment, the blood sample M1 is hemolyzed and diluted. .. The first particle measuring unit 11 shown in FIG. 1 indicates that the measuring container (chamber) is provided with a particle counting mechanism for measuring the volume of each particle.
[0016]
　The CRP measuring unit (CRP chamber) 13 is configured to measure a CRP parameter value which is an instruction value corresponding to the CRP value of the blood sample M1 and output the measured value. The CRP measuring unit 13 shown in FIG. 1 indicates that the measuring container (chamber) is provided with an optical device for measuring the CRP parameter value.
[0017]
　The analysis unit 40 has at least a volume value receiving unit, a CRP value receiving unit, and a determination unit 430. In the example shown in FIG. 1, the volume value receiving unit and the CRP value receiving unit are included in the measured value receiving unit 410. In the example shown in FIG. 1, the analysis unit 40 has a frequency distribution calculation unit 421 and a CRP value determination unit 423. The frequency distribution calculation unit 421 receives the volume measurement value of each particle from the first particle measurement unit 11 and calculates the volume frequency distribution of the particles (particles containing white blood cells) in the blood sample M1. In malaria-positive blood samples, these particles are mixed with Plasmodium. The CRP value determination unit 423 receives the CRP parameter value output from the CRP measurement unit 13 and determines the CRP value of the blood sample M1. Therefore, in the example of FIG. 1, the analysis unit 40 is provided with a measured value receiving unit (including a volume value receiving unit and a CRP value receiving unit) 410 and a measured value processing unit 420.
[0018]
　The measured value receiving unit (including the volume value receiving unit and the CRP value receiving unit) 410 accepts the measured values ​​output from each measuring unit (11, 12, 13) in the measuring mechanism unit 10 as data that can be calculated. .. The measured value receiving unit 410 may be configured to have an interface as hardware and software (computer program) that holds the received measured value signal as a measured value data set. The measured value receiving unit 410 may store each received measured value in a storage device (not shown).
　The measurement value processing unit 420 includes a first frequency distribution calculation unit 421 and a CRP value determination unit 423. Further, in the example of FIG. 1, the measured value processing unit 420 also includes a second frequency distribution calculation unit 422.
　The first frequency distribution calculation unit 421 processes the measured value output from the WBC chamber 11 and calculates the volume frequency distribution of the particles (white blood cells, Plasmodium malaria) in the hemolyzed blood sample M1.
　The CRP value determination unit 423 processes the measured value (CRP parameter) output from the CRP chamber 13 to determine the CRP value.
　The second frequency distribution calculation unit 422 processes the measured values ​​output from the second particle measurement unit (RBC chamber) and calculates the volume frequency distribution of the particles (particles containing red blood cells and platelets) in the blood sample M1. ..
[0019]
　In the determination unit 430, the blood analysis value obtained by the measurement value processing unit 420 (in particular, the volume frequency distribution obtained by the first frequency distribution calculation unit 421 and the CRP value obtained by the CRP value determination unit 423) is obtained. , When either of the following conditions (i) and (ii) is satisfied, it is determined that the blood sample M1 is likely to be malaria positive. Further, in a preferred embodiment, the blood sample M1 is likely to be malaria-positive (or the blood sample M1 may be malaria-positive) when both the following conditions (i) and (ii) are satisfied. Higher).
　(I) The frequency peak portion exists within the volume range peculiar to Plasmodium malaria in the volume frequency distribution.
　(Ii) The CRP value is equal to or higher than a predetermined threshold value.
　Although the details will be described later, in order to perform the condition determination of the above (i) and (ii), in the example of FIG. 1, the determination unit 430 includes the condition determination unit 431, the reference data holding unit 432, and the determination unit 433. Is provided. The reference data holding unit 432 stores data to be referred to for the condition determination unit 431 to determine, such as a threshold value for determining the CRP value.
　The condition determination unit 431 refers to the necessary data held in the reference data holding unit 432 and performs the condition determinations (i) and (ii) above.
　The determination unit 433 accepts the result of the condition determination by the condition determination unit 431, and if only one of the above conditions (i) and (ii) is satisfied, "there is a high possibility of malaria positive". Select the judgment of. If both of the above conditions (i) and (ii) are not satisfied, the determination of "not positive for malaria" may be selected. Even when both of the above conditions (i) and (ii) are satisfied, the determination that "there is a high possibility of being malaria positive" may be selected. However, if both of the above conditions (i) and (ii) are met, it is more likely (or more likely to be malaria positive) than if only one is met. The accuracy of the judgment itself is higher). Therefore, when both of the above conditions (i) and (ii) are satisfied, in order to make a difference in judgment from the case where only one of them is satisfied, "the possibility of malaria positive is higher. You may choose to determine "high" or, instead, "malaria positive". In the example of FIG. 1, the analysis unit 40 has a determination result output unit 440, through which the determination result selected by the determination unit 433 is transmitted to a peripheral device such as a display device 50, a communication device, an external computer, or the like. It is output.
　As another aspect of the analysis unit, the measurement value receiving unit 410 may accept the already calculated frequency distribution. Further, the CRP parameter value may be the CRP value itself.
[0020]
　By having the above configuration, the device can perform malaria determination even though it has an inexpensive configuration. The device may be an analysis device that only includes a first particle measurement unit, a CRP measurement unit, and a control unit (including an analysis unit), and is an analysis dedicated to malaria determination that outputs only the results of malaria determination. Although it may be an apparatus, as illustrated in FIG. 1, it has a configuration in which an analysis function for determining malaria is added to a conventionally known blood analyzer, and analyzes the total blood count, analyzes the CRP value, and malaria. It is preferable that the device makes a determination. In the example shown in FIG. 1, the WBC chamber 11 is used as the first particle measuring unit in the blood analyzer provided with the WBC chamber 11, the RBC chamber 12, and the CRP chamber 13. This is because in the WBC chamber, blood samples are hemolyzed (treated to dissolve red blood cells with a hemolytic agent) to dilute them to form a sample solution, and white blood cells are counted, so malaria lurking in the red blood cells. This is because it is suitable for exposing protozoans and counting them as particles. Further, it is preferable to use the measurement result of platelets counted in the RBC chamber 12 in order to further improve the accuracy and reliability of malaria determination (described later).
[0021]
[Particle Counting Mechanism]
　Particle counting includes not only simply counting the number of particles but also measuring the volume of each particle. This makes it possible to analyze how many volumes of particles are present (that is, the volume frequency distribution of particles). The "frequency" in the volume frequency distribution is the number of particles for each volume (or a channel described later).
　As a particle counting mechanism in the particle measuring unit, a mechanism that executes an impedance method (electrical resistance method) (using a change in the electrical characteristics (impedance) in the particle when the particle passes through the aperture, is used to obtain the particle. A mechanism for measuring volume) and a mechanism for performing flow cytometry (particles in the sample liquid traveling through the flow cell flow cell are irradiated with predetermined irradiation light, and the resulting optical characteristics such as light scattering and light absorption are obtained. (Mechanism for measuring the volume of the particles), Condensing flow impedance method (Mechanism for performing the impedance method and flow cytometry in one flow path so that each particle can be measured by the impedance method and flow cytometry). (Mechanism in which) is mixed. Among these particle counting mechanisms, the mechanism that executes the impedance method is inexpensive and is a useful mechanism for finding the frequency peak portion peculiar to Plasmodium malaria from the frequency distribution.
[0022]
[Particle Measuring Unit Mounted on the Device] In
　the example of FIG. 1, the first particle measuring unit (WBC chamber) is configured to count particles by the impedance method described later. The analysis unit classifies leukocytes into lymphocytes, monocytes, and granulocytes (3 classifications of leukocytes) as shown in the volume frequency distribution in FIG. 5, so that the state of their frequency distribution can be displayed as an analysis result. It is composed. Platelets may also be counted in the RBC chamber, and the WBC chamber may be configured to be capable of measuring hemoglobin concentration (Hgb). In addition, it may be configured to be able to measure items similar to those of a conventional blood analyzer, such as red blood cell volume (MCV) and hematocrit value (Hct).
　Further, a particle measuring unit (measurement chamber) for classifying granulocytes into neutrophils, eosinophils, and basophils may be added to the apparatus. Such a particle measuring unit uses an LMNE chamber (cell staining technology and a particle counting mechanism (particularly a mechanism for performing a concentrating flow impedance method) to count lymphocytes, monospheres, neutrophils, and eosinophils. (Measurement chamber) and BASO chamber (measurement chamber in which components other than basophils are hemolyzed and contracted by a hemolytic agent and basophils are counted by a particle counting mechanism).
　A blood analyzer equipped with two measurement chambers (RBC chamber and WBC chamber for white blood cell 3 classification) configured to carry out the impedance method and a CRP chamber is an inexpensive configuration and is a basic blood cell. Therefore, it is possible to inexpensively provide a preferable blood analyzer useful for determining malaria and diagnosing medical treatment.
[0023]
[CRP measuring unit mounted on the device] The
　CRP chamber is configured to measure the CRP parameter value corresponding to the CRP value of the blood sample. Examples of the method for obtaining such a CRP parameter value in the measurement chamber include the CRP latex immunoturbidimetric RATE method. For example, in the CRP latex immunoturbidimetric RATE method, a latex reagent for immunoassay (eg, anti-human CRP sensitized latex immunoreagent) is added to the blood sample in the chamber. A light irradiation unit and a photodetection unit are provided on the lower wall surface of the chamber, and the detected intensity of transmitted light is output as a CRP parameter value. This CRP parameter value is associated with the CRP value in the CRP value determination unit 423 of the analysis unit 40. As another aspect, the CRP measuring unit may have a CRP value determining unit and output the CRP parameter value as data converted into a CRP value.
　For the technique itself for measuring the CRP value in the blood analyzer, conventionally known techniques such as Patent Document 3 can be referred to.
[0024]
[Measuring mechanism unit]
　Each unit of the measuring mechanism unit 10 is controlled and operated by the mechanism control unit 30 included in the control unit 20. The mechanism control unit 30 and the analysis unit 40 cooperate with each other to cause the mechanism control unit 30 to perform measurement, analyze blood, and determine malaria.
　In the example of FIG. 1, the measuring mechanism unit (indicated by being surrounded by a broken line) 10 has a dispensing mechanism 14 in addition to the CRP chamber 13, the RBC chamber 12, and the WBC chamber 11. The dispensing mechanism 14 has a sampling nozzle 15 and a moving mechanism 16, and the moving mechanism has a horizontal moving mechanism 16a and a vertical moving mechanism 16b so that the sampling nozzle 15 can be moved in the horizontal direction and the vertical direction. It is composed of. The sampling nozzle 15 moves to the sample container 10, the reagent container, the red blood cell measurement chamber, and the white blood cell measurement chamber under the control of the mechanism control unit 30, and sucks and discharges the blood sample, the reagent solution, the sample solution created in the chamber, and the like. It is an elongated tube that performs.
　In addition, the measuring mechanism unit 10 is appropriately provided with reagent containers (s) containing reagents necessary for measurement, a dilution mechanism, a suction / discharge pump, a syringe, an electromagnetic valve, and the like (these are not shown). There is). For the measuring mechanism unit, conventionally known particle analyzers, blood analyzers, blood cell counters, etc., such as Patent Document 3, can be referred to.
[0025]
[Control unit] The
　control unit may be constructed by a logic circuit or the like, but a computer is suitable. In the computer, a computer program configured to control the operation of each part of the measurement mechanism unit and to perform an operation for analysis is executed.
[0026]
[Main operation and analysis flow of the
　device ] FIG. 2 is a flowchart illustrating an outline of the operation of the device. In the example of the flowchart, the analysis steps s1 to s4 of the CRP value and the analysis steps s5 to s8 of the particle measurement (white blood cell count) are shown in parallel, but the operations of these two series are in series. You may. The number assigned to each step is for identification. In practice, an RBC chamber and additional measurement chambers are added, and blood samples are distributed to each chamber in a predetermined order, in which sample solution is formed (a part of the blood sample diluted in the WBC chamber is placed in the RBC chamber). It also includes the movement of the sample solution between chambers, such as being transferred and further diluted there), after which the measurement proceeds independently within each chamber, and the time required for the measurement is also independent for each chamber. Is. Hereinafter, an example of the main operation and analysis flow in the apparatus of FIG. 1 will be described.
[0027]
[CRP value analysis steps s1 to s4]
(Step s1) A
　sampling nozzle (hereinafter referred to as a nozzle) operates to measure the CRP parameter value, and a reagent required for measuring the CRP parameter value (reagent container is not shown). ) And the blood sample M1 in the sample container X1 are supplied to the CRP measuring unit (CRP chamber) 13 in a predetermined amount, respectively, and a sample solution for measuring the CRP parameter value is formed there. Nozzle cleaning is performed as appropriate.
[0028]
(Step s2) The
　CRP parameter value is measured in the CRP chamber, and the measured value is output to the analysis unit.
[0029]
(Step s3) The
　analysis unit 40 (measured value receiving unit 410) receives the CRP parameter value (absorbance).
[0030]
(Step s4) The
　analysis unit 40 (CRP value determination unit 423 of the measurement value processing unit 420) determines the CRP value from the CRP parameter value. The CRP value determination unit 423 may have a reference table for converting the CRP parameter value into the CRP value.
[0031]
[Analysis of Volume Volume Distribution of Particles Steps s5 to s8]
(Step s5) The
　nozzle supplies a predetermined amount of blood sample M1 in the sample container X1 to the first particle measurement unit (WBC chamber) 11, and there. A predetermined amount of diluent (water, physiological saline, phosphate buffered diluent, etc.) is supplied to dilute the blood sample M1. Although the flowchart of FIG. 2 does not specifically show the detailed operation, a part of the blood sample diluted in the WBC chamber is transferred to the RBC chamber, where it is further diluted to count red blood cells and platelets. .. For these dilution ratios, a conventionally known counting technique may be referred to. In the WBC chamber, a hemolytic agent that dissolves (hemolyzes) red blood cells is further added to the diluted blood sample to form a sample solution in which red blood cells are dissolved.
[0032]
(Step s6)
　Particle counting is performed in the first particle measuring unit (WBC chamber) 11, the volume of each particle is measured, and the measured value is output to the analysis unit. In the WBC chamber, the hemoglobin concentration may be measured by an optical device for performing a colorimetric method (non-cyan method), and the measured value may be output. On the other hand, particle counting is also performed in the second particle measuring unit (RBC chamber) 12, the volume of each particle is measured, and the measured value is output to the analysis unit.
[0033]
(Step s7) The
　analysis unit 40 (measurement value receiving unit 410) receives measurement results such as particle volume measurement values ​​from each particle measurement unit.
[0034]
(Step s8) The
　analysis unit 40 (frequency distribution calculation unit 421 of the measurement value processing unit 420) processes the particle volume measurement value from the WBC chamber and calculates the volume frequency distribution of the particles containing leukocytes. On the other hand, the second frequency distribution calculation unit 422 processes the particle volume measurement value from the RBC chamber and calculates the volume frequency distribution of erythrocytes and platelets. These volume frequency distributions may be output or displayed as a histogram.
[0035]
[Analysis Steps s9 to s15 for Malaria Judgment by Judgment Unit 430] In the
　condition judgment shown in the flowchart of FIG. The order of the condition determination regarding the volume frequency distribution and the condition determination regarding the CRP value may be reversed or simultaneous and parallel.
[0036]
(Step s9) In
　the example of the flowchart of FIG. 2, in the condition determination unit 431 of the determination unit 430, the volume frequency distribution obtained by the measurement in the WBC chamber in step s9 is the condition (i) above (for malaria protozoan). It is determined whether or not the frequency peak portion exists within the peculiar volume range) is satisfied.
[0037]
(Step s10) When
　the condition determination unit 431 determines in step s9 that there is a frequency peak portion (YES), the condition determination unit 431 determines the condition (CRP value) of the above (ii) in advance in step s10. It is determined whether or not the condition (that is equal to or higher than the threshold value) is satisfied.
[0038]
(Step s11) When
　the condition determination unit 431 determines in step s10 that the CRP value is equal to or greater than the threshold value (YES), both the above condition (i) and the above condition (ii) are satisfied. In step s11, the determination unit 433 selects the determination result that "there is a higher possibility of being malaria positive". This determination result is a preferable example in which the possibility of malaria positivity is set in stages, and even if the determination result of "high possibility of malaria positivity" is selected in steps s11, s12, and s14, all of them are similarly selected. good.
[0039]
(Step s12) When
　the condition determination unit 431 determines in step s10 that the CRP value is not equal to or greater than the threshold value (NO), the determination unit 433 determines in step s12 that "there is a high possibility of being malaria positive". select.
[0040]
(Step s13)
　On the other hand, when the condition determination unit 431 determines in step s9 that there is no frequency peak portion (NO), the condition determination unit 431 determines in step s13 the condition (CRP value) of the above (ii) in advance. It is determined whether or not the above threshold value is satisfied.
[0041]
(Step s14) When
　it is determined in step s13 that the CRP value is equal to or greater than the threshold value (YES), the determination unit 433 selects the determination result that "there is a high possibility of being malaria positive" in step s14.
[0042]
(Step s15) When
　the condition determination unit 431 determines in step s13 that the CRP value is not equal to or greater than the threshold value (NO), the determination unit 433 selects the determination result of "not malaria positive" in step s15.
[0043]
　By the above steps, the malaria determination is preferably performed.
[0044]
[Detailed Description of Conditions (i) and (ii) Above] As
　shown in the volume frequency distribution charts of Experimental Examples 1 and 3 described later, in the present invention, the sample solution (blood that has been subjected to hemolytic treatment and diluted) is used. It was found that a peculiar frequency peak part may appear due to the presence of malaria protozoa within a predetermined range in the volume frequency distribution map of the particles in the sample). In the example of the volume frequency distribution diagram of FIG. 3, in a malaria-positive blood sample, the range of channels 14 to 26 corresponding to 25 fL (femtolitre) to 45 fL, particularly the range of channels 17 to 23 corresponding to 30 fL to 40 fL. A clear frequency peak appears inside (“channel” will be described later). In the example of FIG. 3, the center of the frequency peak portion is concentrated in the vicinity of the channel 20. This is a result of the lysis of erythrocytes, which caused the Plasmodium contained in the erythrocytes to float in the sample solution, which was counted as particles. The frequency peak portion centered around the channel 40 indicates the presence of lymphocytes, and the frequency peak portion peculiar to Plasmodium malaria is next to the side having a smaller volume than the frequency peak portion of lymphocytes. Existing.
[0045]
[Channel] The
　volume measurement value indicating the volume of the particle is not a specific numerical value, but the total width from the minimum value (which may be zero) to the maximum value of the volume measurement value is set to 256 steps (0 to 255) or 1024. It is preferable to divide into stages (0 to 1023) and the like. This facilitates the handling of data when the control unit (computer) performs the process of obtaining the frequency distribution. Each section divided as described above is called a channel. In this specification, an example in which channels 0 to 255 are used as the name of the volume section is given. 3 and 6 show the frequency distribution of the region (channels 0 to 100) in which the volume of the particles is small among all channels 0 to 255.
[0046]
[Existence of frequency peak portion peculiar to malaria protozoa] In the
　present invention, the existence of a frequency peak portion in the volume frequency distribution of particles typically means that the graph curve of the volume frequency distribution (smoothly connects the frequencies of each channel). It means that there is a part where the slope is 0 in the middle of the part where the slope of the tangent line of the curve) is positive or negative, and in a more remarkable state, it means that there is a part where the maximum value is taken. As shown in the graph curve of FIG. 3, the frequency of the channel on the side having a smaller volume than the frequency peak portion is less than the frequency of the frequency peak portion, and therefore the slope of the tangent line on the side having a smaller volume than the frequency peak portion is positive ( It is positive).
　The height of the frequency peak portion peculiar to Plasmodium malaria is preferably determined by the peak to valley based on the frequency from the bottom of the valley adjacent to the frequency peak portion on the side having a larger volume. However, the height may be based on zero frequency (= the value of the frequency of the frequency peak portion itself), or the height may be based on a predetermined frequency.
[0047]
[First threshold value: lower limit of the height of the frequency peak portion peculiar to Plasmodium malaria]
　According to the experiments of the present inventors, the height (peak to valley) of the frequency peak portion peculiar to Plasmodium malaria is, for example, 1 or more. That is, the graph curve of the volume frequency distribution does not increase monotonically within the volume range peculiar to Plasmodium malaria, and the direction of the curve changes like an inflection point (the height of the frequency peak part is zero). ) Is present, it is suspected that malaria is positive, but in the present invention, 1 is set as the lower limit for determining the peak to valley, which is an appropriate value based on experiments. It is adopted. This lower limit value "1" may be stored in advance in the reference data holding unit 432. Further, this lower limit value may be appropriately changed by the operator of the device.
　When the frequency peak portion is present, the determination unit determines that there is a high possibility of malaria positive, and the final malaria determination is made together with the determination result of the CRP value.
[0048]
[Threshold: Lower limit of CRP value peculiar to malaria positive]
　The CRP value of a healthy person is generally about 0 mg / L to 5 mg / L. According to the experiments of the present inventors, a malaria-positive blood sample may show a high CRP value of about 50 mg / L. From the results of the experiment, an appropriate threshold value (lower limit value for determination) that can be determined to be highly likely to be malaria positive is, for example, a value higher than 5 mg / L to 50 mg / L, and a more preferable threshold value is 20 mg. / L to 30 mg / L can be mentioned. The threshold can be selected from these ranges. In Experimental Example 2 described later, a CRP value of 26 mg / L is adopted as a threshold value (lower limit value for determination), and if the CRP value is equal to or higher than the threshold value, it is determined that there is a high possibility of malaria positive. The adopted threshold value is stored in advance in the reference data holding unit 432. The threshold value may be appropriately changed by the operator of the device.
[0049]
[Relationship between frequency peaks peculiar to Plasmodium malaria and CRP value] As
　shown in the scatter diagrams of Experimental Examples 2 and 4 described later, in the present invention, malaria is based on both the volume frequency distribution of particles and the CRP value. It has been found that if the judgment is made, the accuracy or reliability of the malaria judgment is synergistically improved.
　According to the study by the present inventors, in the scatter plot of FIG. 4, malaria is positive in all cases where the height of the frequency peak portion of the volume frequency distribution of the particles in which leukocytes and Plasmodium are present is 1 or more. It is preferable to determine that the possibility is high. Further, in the volume frequency distribution, it is preferable to determine that the possibility of malaria positive is high in all cases where the CRP value is 26 mg / L or more. Further, in the scatter plot of FIG. 4, when the height of the frequency peak portion of the volume frequency distribution of the particles in which leukocytes and Plasmodium are present is 1 or more and the CRP value is 26 mg / L or more. It is preferable to determine that the probability of malaria positive is high (or higher).
[0050]
[Display of
　malaria judgment result ] For display of malaria judgment result, see the operator of the blood analyzer or the analysis result, such as "(malaria) positive" or "(malaria) positive is likely". It may be a sign or symbol that allows a person to understand whether or not he / she is positive for malaria.
　Further, the higher the peak to valley and / or the higher the CRP value, the higher the possibility of malaria positive may be displayed.
[0051]
[Identification of types of
　malaria parasites ] There are many types of malaria parasites, but the ones that infect humans and cause clinical problems are mainly Plasmodium falciparum (P. falciparum) and Malaria parasite (P. falciparum). vivax), ovate malaria parasite (P. ovale), quartan malaria protozoa (P. malariae).
　Among these malaria protozoa, in the present invention, as shown in the frequency distribution charts of Experimental Example 3 and FIG. 5 described later, the height of the frequency peak portion is set between the Plasmodium falciparum and the Plasmodium falciparum. We found that there were clear differences and that the frequency of Plasmodium falciparum tended to be higher than that of Plasmodium falciparum. Therefore, in a preferred embodiment of the present invention, a malaria protozoan discrimination unit may be further provided in the analysis unit. The malaria protozoan discriminating unit can discriminate the type of malaria protozoan based on the height of the frequency peak portion within the volume range peculiar to the malaria protozoan. In the example of FIG. 1, the malaria protozoan discrimination unit is included in the condition determination unit 431 in the determination unit 430.
[0052]
[Correction of malaria determination by analysis of platelets]
　Platelets are also counted in the WBC chamber. When platelets aggregate, they may become one large particle equivalent to Plasmodium malaria. In the present invention, attention has been paid to the possibility that the frequency of aggregated platelets is included in the frequency peak portion within the volume range peculiar to Plasmodium malaria. By analyzing the platelets, it is possible to correct the height of the frequency peak portion by subtracting the frequency of the aggregated platelets from the height of the frequency peak portion.
[0053]
[Differentiation between Malaria Positive and Dengue Virus Positive] As
　shown in the frequency distribution charts of Experimental Examples 4 and 6 described later, in the present invention, even when dengue virus is positive, the leukocyte frequency distribution changes, which is peculiar to Plasmodium malaria. It is noted that a frequency peak portion is generated in the vicinity of the volume range of the above. Dengue virus is a minute particle that cannot be detected by the particle counting mechanism, but as shown in the frequency distribution diagram of FIG. 6, a frequency peak portion may occur in the vicinity of the volume range peculiar to Plasmodium malaria. However, as is clear from the frequency distribution map (malaria positive) in FIG. 3 and the frequency distribution map (dengue virus positive) in FIG. 6, the width of the frequency peak portion differs greatly between the malaria positive and the dengue virus positive. In malaria positive, the frequency peak is relatively sharp, whereas in dengue virus positive, the shape of the frequency peak is clearly wide.
　Further, in the present invention, attention is paid to the fact that the CRP values ​​of malaria-positive and dengue virus-positive are significantly different from each other. When the range of CRP value in the central part of the distribution (box part in FIG. 7) containing 50% of the samples is malaria positive, as shown in Experimental Example 5 and Box Histogram Frequency Distribution Map shown later. Is larger than in the case of dengue virus positive.
　From the above points, it is determined whether the possibility of malaria positive or dengue virus positive is high based on the width of the frequency peak portion or based on the width of the frequency peak portion and the CRP value. It is also possible to provide a malaria-dengue discriminator configured to do so.
[0054]
[Computer Program According to the Present Invention]
　Next, a computer program according to the present invention (hereinafter, also referred to as the program) will be described. The program may be provided as recorded on a computer-readable medium, or may be provided via a network from another computer or external storage device.
[0055]
　The program is preferably used to determine malaria in the control unit of the blood analyzer of the present invention. The equipment for each measurement and the process for analysis in the program are as described above, and the flow of measurement and analysis of measured values ​​and malaria determination, and the threshold value for determining CRP values ​​are described above. As you did.
[0056]
　The program
　:: The step of accepting the test measurement value of the blood sample to be analyzed (steps S1 to S4 in the examples described later);
　the test measurement value for a specific blood test item other than the CRP value, and / or It is
a computer program that causes a computer to execute steps (steps S5 to S7 in the examples described later) for determining the possibility of malaria positiveness in the blood sample based on the test measurement value for the CRP value .
　The test measurement value is a measurement value for each of a predetermined blood test item including a CRP value and a blood cell count value. In addition, the specific blood test item is a blood test item related to malaria determination.
[0057]
　In a preferred embodiment, the program is a computer program that causes a computer to perform the following steps S1 to S7. For the reference in the following steps, the reference numerals of each part of the apparatus illustrated in FIG. 1 are used for explanation.
[0058]
(Step S1) A
　first particle measuring unit (WBC) for measuring the volume of each particle (particle containing white blood cells) in a sample solution (diluted in a preferable embodiment) obtained by subjecting blood sample X1 to a hemolytic treatment. The step of accepting the volume measurement value of each particle from the chamber) 11.
(Step S2)
　A step of calculating the volume frequency distribution of the particles based on the volume measurement value of each particle.
(Step S3) A step
　of receiving a CRP parameter value from a CRP measuring unit (CRP chamber) 13 for measuring a CRP parameter value corresponding to the CRP value of the blood sample X1.
(Step S4)
　A step of determining a CRP value based on a CRP parameter value.
(Step S5) A step
　of determining whether or not a frequency peak portion exists within a volume range peculiar to Plasmodium malaria in the volume frequency distribution.
(Step S6) A step
　of determining whether or not the determined CRP value is equal to or higher than a predetermined threshold value.
(Step S7) A step of
　determining that the blood sample is likely to be malaria-positive when the frequency peak portion is present and the determined CRP value is equal to or higher than the threshold value.
　By using a computer program that causes a computer to perform the above steps, it is possible to perform malaria determination with higher accuracy or higher reliability than before.
[0059]
　In a preferred embodiment of the program, a further Plasmodium malaria discrimination step is provided to determine the type of Plasmodium (Plasmodium falciparum, Plasmodium falciparum) based on the height of the frequency peak within the volume range specific to Plasmodium falciparum. May be determined.
　Further, in a preferred embodiment of the program, a step of calculating the volume frequency distribution of the particles containing platelets received from the second particle measuring unit is further provided, and the volume and frequency of the agglomerate formed by agglutination of platelets are determined. It may be calculated. Then, as described above, a step may be provided in which the height of the frequency peak portion peculiar to Plasmodium malaria is corrected according to the frequency distribution of the agglutinating mass of platelets, and the determination regarding the positive malaria is corrected. ..
　Further, in a preferred embodiment of the program, a malaria-dengue determination step is further provided, and the possibility of malaria positive is possible based on the width of the frequency peak portion or based on the width of the frequency peak portion and the CRP value. It may be determined whether it is high or likely to be dengue virus positive.
[0060]
[Experimental Example 1] In
　this experimental example, a blood sample (sample number 147) confirmed to be malaria-positive by a morphological diagnostic method using a blood smear was subjected to hemolysis treatment, and particles (white blood cells) were subjected to hemolysis treatment. ) Was counted, and the frequency distribution of the particles was calculated from the measurement results. The graph of FIG. 3 shows the result of superimposing the volume frequency distribution map (range from channel 0 to 100) of each blood sample in one graph.
　As is clear from the graph of FIG. 3, it was found that the frequency peak portion peculiar to Plasmodium malaria appeared in channels 17 to 23.
[0061]
[Experimental Example 2] For
　the malaria-positive blood sample used in Experimental Example 1, particles (white blood cells) are counted and the CRP value is also measured, and channel 17 of the volume frequency distribution map is measured for each blood sample. The height of the frequency peak portion peculiar to the malaria protozoan appearing in No. 23 was associated with the CRP value. The height of the frequency peak portion was determined by a peak to valley based on the frequency of the bottom of the valley portion adjacent to the frequency peak portion on the side having a larger volume. Then, each blood sample was plotted on a scatter plot having the height of the frequency peak portion as the horizontal axis and the CRP value as the vertical axis.
　A scatter plot of the results is shown in the graph of FIG. The vertical broken line in the graph is a line indicating that the height of the frequency peak portion is 1, and the horizontal broken line in the graph is a line indicating that the CRP value is 26 mg / L.
　In the scatter plot of FIG. 4, in the region where the height of the frequency peak portion of the volume frequency distribution of the particles in which leukocytes and Plasmodium are present is 1 or more and the CRP value is 26 mg / L or more. More than 94% of malaria-positive specimens were included. It was found that malaria positive can be accurately determined by relatively simple measurements such as particle counting of blood samples and CRP value measurement.
[0062]
[Experimental Example 3] In
　this experimental example, the blood sample falci confirmed to be infected with Plasmodium falciparum by the RDT (Rapid Diagnotic. Test) method and the malaria parasite vivax were infected. The confirmed blood sample vivax was subjected to hemolytic treatment in the same manner as in Experimental Example 1, particles (leukocytes) were counted by the same blood analyzer, and the frequency distribution of the particles was calculated from the measurement results. The graph of FIG. 5 shows the result of superimposing the volume frequency distribution map (range from channels 1 to 254) of each blood sample in one graph. The thick line in the graph of FIG. 5 is the volume frequency distribution diagram of the particles contained in the blood sample vivox, and the thin line is the volume frequency distribution diagram of the particles contained in the blood sample falci.
　As is clear from the graph of FIG. 5, in all blood samples, a frequency peak portion peculiar to Plasmodium appears near channel 20, but the blood sample vitax has a higher frequency peak portion than the blood sample falci ( It was found that the peak to valley) was higher and the height of the frequency peak portion based on the frequency 0 was also higher in the blood sample vivax than in the blood sample falci.
[0063]
[Experimental Example 4] In
　this experimental example, a blood sample (170 samples) confirmed to be positive for dengue virus by an enzyme-linked immunosorbent assay (ELISA method) was subjected to hemolytic treatment to count particles (white blood cells). Was performed, and the frequency distribution of the particles was calculated from the measurement results. The graph of FIG. 6 shows the result of superimposing a representative volume frequency distribution map (range from channel 0 to 100) of each blood sample in one graph.
　As is clear from the graph of FIG. 6, a frequency peak portion appeared in the vicinity of the same channel as the peak caused by Plasmodium in FIG.
[0064]
[Experimental Example 5] In
　this Experimental Example, a blood sample confirmed to be malaria-positive as in Experimental Example 1 and a blood sample confirmed to be dengue virus-positive as in Experimental Example 4 were subjected to. Each CRP value was measured.
　FIG. 7A is a box-and-whisker plot showing the CRP value of a malaria-positive blood sample, and FIG. 7 (b) is a box-and-whisker plot showing the CRP value of a dengue fever-positive blood sample. Further, in each box plot, the horizontal line on the lower side of the box is the 9% line from the smallest distribution of CRP values, the lower limit of the box is the 25% line, and the horizontal line in the box is the average value. The upper limit of the box is the 75% line, and the horizontal line on the upper side of the box is the 91% line. A plot of 9% line or less and a plot of 91% line or more are displayed, and plots of the same CRP value are displayed side by side in the horizontal direction.
　From the graphs of FIGS. 7 (a) and 7 (b), the CRP value is significantly different between malaria positive and dengue virus positive, and the malaria positive case has a larger CRP value than the dengue virus positive case. It was found that it was distributed in.
[0065]
(Aspect 1 of Malaria Judgment Using Artificial Intelligence) In
　a preferred embodiment 1 of the device, a blood test item used for malaria determination is selected using artificial intelligence. In this aspect, the determination unit 430 of FIG. 1 is malaria positive (or malaria determination) identified from predetermined blood test items (various blood test items) by the first learned model M1 described later. Use important test items that are relevant to. In addition, the determination unit 430 is configured to perform malaria determination based on the inspection measurement values ​​for the specified important inspection items.
[0066]
(First trained model M1)
　The first trained model M1 uses the following data (a1) and (b1) as teacher data, and which of the predetermined blood test items is positive for malaria. It is formed by letting artificial intelligence machine-learn whether or not it is related to.
　(A1) A test measurement value (A1) obtained for the predetermined blood test item and a test measurement value (A1) obtained from a blood sample (A) known to be malaria positive as a measurement target are malaria positive. Information that it belongs to the blood sample of (A2).
　(B1) A test measurement value (B1) obtained for the predetermined blood test item and a test measurement value (B1) obtained from a blood sample (B) known to be malaria-negative as a measurement target are malaria-negative. Information that it belongs to the blood sample of (B2).
[0067]
　The blood test item identified by the first trained model M1 is an important test item (ie, a blood test item associated with malaria positivity). In this aspect 1, the important test items may include all blood test items associated with malaria positivity as much as possible.
[0068]
　The important inspection item specified by the first trained model M1 may be held in the reference data holding unit 432 in FIG. 1 or may be incorporated in the condition determining unit 431 as a determination condition, for example. The condition determination unit 431 refers to the important inspection item held in the reference data holding unit 432, receives the inspection measurement value related to the important inspection item from the measurement value receiving unit 410 or from the measurement value processing unit 420, and is predetermined. Condition judgment is performed. The determination unit 433 accepts the result of the condition determination by the condition determination unit 431 and makes a malaria determination. That is, in this aspect 1, the important inspection item is specified in advance by the trained model, and the malaria determination is performed using the inspection measurement value of the important inspection item. The same applies to the following other aspects.
[0069]
　In the present invention, the artificial intelligence and the trained model are preferably built on a computer other than the device (such as a neurocomputer configured to conform to the principle of artificial intelligence), but the device is provided. May be done.
　In addition, the artificial intelligence and the trained model built on a computer other than the device may be connected to the device through a communication line.
　In addition, a blood sample (or a blood sample separately collected from a patient from which the blood sample was collected) for which the malaria was determined by the device is subjected to a more accurate malaria determination by a conventional precise test method. The test measurement value by the device and the malaria determination result by the conventional test method may be fed back to artificial intelligence as teacher data to further update the learning state of the trained model.
[0070]
(Important inspection items
　specified by the trained model M1 ) Examples of the important inspection items specified by the trained model M1 include the following items (e1) to (e4).
　(E1) Whether or not there is a frequency peak in the volume range of 25 fL to 45 fL in the volume frequency distribution of leukocytes in the blood sample to be analyzed. Hereinafter, this frequency peak is also referred to as "L1 peak in the volume frequency distribution of leukocytes". The L1 peak in the volume frequency distribution of leukocytes is, for example, as described above with reference to FIG. 3, and is shown as a “frequency peak portion caused by Plasmodium malaria” in the frequency distribution diagram of FIG.
　According to the study of the present inventors, the height of the L1 peak in the volume frequency distribution of leukocytes and the number of Plasmodium contained per unit volume of malaria-positive blood are shown in the graph of FIG. There is such a clear correlation. As explained with reference to FIG. 3, the height of the L1 peak is a value determined by peak to valley, a height based on zero frequency (= the frequency value itself of the frequency peak portion), and the like. There may be.
　(E2) The number of white blood cells in the blood sample to be analyzed. Blood test items include items that measure the number of white blood cells in a specific volume (1 μL, etc.) of blood without classifying white blood cells in detail. In malaria-positive blood, the white blood cell count is known to show abnormal values ​​(increase or decrease), which is consistent with conventional findings.
　(E3) The number of platelets in the blood sample to be analyzed. Blood test items include items that measure the number of platelets in a specific volume (1 μL, etc.) of blood. It is known that the number of platelets shows an abnormal value (decrease) in malaria-positive blood, which is consistent with the conventional findings.
　(E4) CRP value of blood sample to be analyzed. Blood test items include an item for measuring the amount of CRP (C-reactive protein) in a specific volume (1 μL or the like) of serum. In the present invention, it is proposed for the first time that a CRP value is used for determining malaria. As revealed by the present invention, the CRP value correlates with malaria positive, and in malaria positive blood, the CRP value shows an abnormal value (increase).
　The important inspection item can include one or more items selected from the group consisting of the items (e1) to (e4), and in order to obtain a more accurate determination result, for example, with the item (e1). Using (e2), or using the items (e1) and (e3), or using the items (e1) and (e4), or using the items (e2) and (e3). Or, using the items (e2) and (e4), or using the items (e3) and (e4), or using the items (e1), (e2), and (e3). Or, using the items (e1), (e2), and (e4), or using the items (e1), (e3), and (e4), or using the items (e2) and (e3). It is preferable to perform malaria determination using (e4) and all of the above items (e1) to (e4).
[0071]
(Modification of
　Phase 1 of Malaria Judgment Using Artificial Intelligence) In this modification, the determination unit was identified from the blood test items by the trained model M11 modified from the first trained model. Use important test items that are more or more relevant to malaria positivity. Then, the determination unit is configured to determine the possibility of malaria positive in the blood sample based on the test measurement values ​​for the specified important test items.
[0072]
　In this modified mode as well, the important inspection items specified by the first trained model M11 may be held in the reference data holding unit 432 in FIG. 1, for example, as in the above-described first mode. The condition determination unit 431 refers to the important inspection item held in the reference data holding unit 432, receives the inspection measurement value related to the important inspection item from the measurement value receiving unit 410 or from the measurement value processing unit 420, and is predetermined. Calculates whether or not the above conditions are met.
[0073]
(Trained model M11 modified from the first trained model M1)
　This trained model M11 uses the following data (a11) and (b11) as teacher data, and which blood of the predetermined blood test items It is formed by letting artificial intelligence machine-learn to what extent the test items are associated with positive malaria.
　(A11) A test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target, and the test measurement value (A1) are malaria positive. Information that it belongs to a blood sample (A2).
　(B11) A test measurement value (B1) obtained for the predetermined blood test item using a blood sample (B) known to be malaria-negative as a measurement target, and the test measurement value (B1) are malaria-negative. Information that it belongs to a blood sample (B2).
[0074]
　In this variant, the important test items include only blood test items that have a greater or greater association with malaria positivity and include blood test items that have a smaller association that can be ignored. not present. On the other hand, in the above aspect 1, the important test items include all blood test items determined to be related, even if the relevance is small.
　The determination unit uses important test items having a greater or greater association with malaria positivity, which is identified from the predetermined blood test items by the modified trained model M11, and regards the important test items. It is configured to determine malaria in a blood sample based on the test measurement values ​​of. For example, among all blood test items that have any association with malaria positivity, the top n items in descending order of association (for example, n = 1 to 5, preferably n = 2 to 4). Items up to the above may be selected as important inspection items, or the relevance is represented by a weighting coefficient, and only those whose weighting coefficient is equal to or greater than a predetermined value (highly relevant) are selected as important inspection items. You may.
　The important inspection items specified by the modified trained model M11 may be the same as the items (e1) to (e4) described above.
[0075]
(Aspect 2 of Malaria Judgment Using Artificial Intelligence) In
　Preferred Aspect 2 of the device, a blood test item used for malaria determination is selected using artificial intelligence, and further, each of the above blood test items is associated with positive malaria. The degree of sexuality is specified by artificial intelligence. In this aspect 2, the determination unit 430 of FIG. 1 is a blood test item (important test item) related to malaria positive (or malaria determination) identified by the second learned model M2 described later. ), And the degree of relevance of each of the blood test items to the malaria positive (or malaria determination) identified by the second trained model M2. Here, specifying the degree of relevance of each blood test item means specifying a weight according to the degree of relevance to malaria positive (or malaria determination) for each blood test item. .. Then, the determination unit 430 is configured to determine malaria of the blood sample based on the test measurement value for the specified important test item and the degree (weight) of the specified relevance.
[0076]
　In this aspect 2, the degree of relevance (weight) to the important inspection item specified by the second trained model M2 may be held by the reference data holding unit 432 in FIG. 1, or the condition determining unit 431. It may be incorporated as a judgment condition in.
　For example, the condition determination unit 431 first refers to an important inspection item held in the reference data holding unit 432, and receives an inspection measurement value related to the important inspection item from the measurement value receiving unit 410 or the measurement value processing unit 420. Next, the condition determination unit 431 calculates whether or not the conditions set by incorporating the weights for each important inspection item are met.
[0077]
(Second trained model M2) As
　the second trained model M2, the same one as the above-mentioned trained model M11 can be used. Even in the embodiment using the same trained model, in the modified embodiment of the above-described aspect 1, the purpose was to eliminate unnecessary blood test items, and the obtained important test items were used under equal conditions. On the other hand, in the second aspect, the purpose is to acquire important inspection items and to acquire weights for each important inspection item, and when determining malaria, weights are given to each inspection measurement value for each important inspection item. Attach and make a judgment.
[0078]
　In this aspect 2, the degree of relevance of each important inspection item is added as a weighting coefficient to the inspection measurement value of the important inspection item specified by the second trained model M2 to form a discriminant for malaria determination. , Malaria may be determined according to the value of the determination formula. For example, a threshold value and the presence / absence of a state are provided in advance for the value of the determination formula, and whether the value of the judgment formula is equal to or less than or equal to the threshold value, or whether or not the value matches the presence / absence of the state. May make a malaria determination. The determination formula, the threshold value, the presence / absence of the state, and the like may be held in the condition determination unit 431 of FIG. 1 or the reference data holding unit 432.
[0079]
　When there are a plurality of important inspection items, a judgment formula is constructed by adding a weighting coefficient corresponding to each inspection measurement value of the plurality of important inspection items, and according to the value (judgment value) indicated by the judgment formula. The determination may be made. In the discriminant, each weighting coefficient is given to the inspection measurement value of the specified important inspection item according to a predetermined operation (sum, difference, quotient, product, etc.), and (f2) the above-mentioned weighting coefficient. The test measurement values ​​to which are given may be an expression in which the test measurement values ​​are combined into one according to a predetermined operation (sum, difference, quotient, product).
　From the viewpoint that the calculation is simple and the influence of each item is easy to understand for the designer, operator, user, etc. of the device, the above discriminant is based on (f1) the inspection measurement value of the specified important inspection item. It is preferable that each weighting coefficient is given in the form of a product, and (f2) the product of the weighting coefficient and the inspection measurement value is combined into one in the form of a sum.
[0080]
　That is, in this aspect 2,
　each important inspection item (Q1, Q2, Q3, ..., Qk) among the important inspection items and the weighting coefficient (p1, p2, p3 ,. , Pk) (p1 × Q1, p2 × Q2, p3 × Q3, ..., pk × Qk) (where k is an integer of 1 or more, preferably 2 (It is about 4) and
　the sum (p1 × Q1 + p2 × Q2 + p3 × Q3 +. ＋ pk × Qk) is calculated as a judgment value, and a step
　of judging that the blood sample is highly likely to be malaria positive by comparing the above judgment value with a preset threshold value is
executed. It may be a device configured in.
　Further, it may be a computer program configured to perform the above steps by a computer, or it may be a blood analysis method (malaria determination method, malaria positive estimation method) that performs these steps. ..
[0081]
　The more important the test item is associated with positive malaria, the more the test measurement value of the important test item has a greater influence on the determination of malaria. The above-mentioned judgment formula (p1 × Q1 + p2 × Q2 + p3 × Q3 + ... + PK × Qk) or its value (judgment value) comprehensively expresses the influence of each change of the inspection measurement value of each important inspection item as a numerical value. ing. The weighting coefficient (p1, p2, p3, ..., PK) is given a larger value as the important test item has a greater influence on the malaria determination. In other words, important test items given a higher numerical weighting factor have a greater effect on malaria determination.
[0082]
　There may be a plurality of important inspection items. When there are a plurality of important test items, for example, a predetermined number (for example, about 1 to 5, preferably about 2 to 4) of blood test items may be adopted as important test items in order from the one having the largest weighting coefficient. good.
　When there are a plurality of important inspection items, malaria can be determined more accurately by using the weighting coefficient corresponding to each important inspection item.
　In addition, an important test item may be adopted from a large number of blood test items by providing a predetermined calculation or threshold value.
[0083]
　The trained models (M1, M11, M2) described above provide known artificial intelligence, such as neural networks (particularly those configured to perform deep learning), with the above-mentioned teacher data, thereby providing the artificial intelligence. Can be constructed by letting them perform machine learning. The operation of giving the above-mentioned teacher data to artificial intelligence to perform machine learning and identifying important test items (and the degree of association with malaria positive) from predetermined blood test items corresponds to classification (discrimination). do.
[0084]
　Predetermined blood test items to be subject to machine learning by artificial intelligence include CRP value, volume frequency distribution of blood cells to be measured, L1 peak in volume frequency distribution of leukocytes, platelet count, leukocyte count, erythrocyte count, hemoglobin. Amount, hematocrit value, average erythrocyte volume, average erythrocyte hemoglobin amount, average erythrocyte hemoglobin concentration, erythrocyte volume particle size distribution width average deviation, erythrocyte volume particle size distribution width, average platelet volume, platelet volume, platelet volume particle size distribution width, large Blood platelet ratio, large platelet count, lymphocyte count, monocyte count, neutrophil count, eosinophil count, basophil count, lymphocyte ratio, monocyte ratio, neutrophil ratio, eosinophil ratio, basophil Hematological test items, biochemical test items, immunological test items such as monocyte ratio, atypical lymphocytes, large immature cells, and various blood count values ​​measured by the above-mentioned particle counting mechanism. .. Artificial intelligence identifies important test items from these test items. The blood test item may be determined according to the blood analysis mechanism provided in the blood analyzer. For example, when the blood analyzer does not include a CRP measuring unit, the important test item does not have to include the CRP value, but the blood analyzer provided with a red blood cell measuring unit, a white blood cell measuring unit, and a CRP measuring unit. Is more preferable, and it is preferable to include the CRP value as an important inspection item.
[0085]
　As the weighting coefficient given by the second trained model, the degree of relevance (for example, a numerical value (ratio) according to each degree of relevance when the total is 1 or 100, etc.) is adopted as it is. Alternatively, the degree of relevance may be subjected to a predetermined calculation to obtain a weighting coefficient.
　In the above aspect, the case where a plurality of important inspection items are adopted has been described, but when the degree of relevance of one important inspection item is significantly larger than the degree of relevance of other important inspection items. May be changed to adopt only one important inspection item having a significantly higher degree of relevance from the plurality of specified important inspection items. In this case, the weighting coefficient may be set only for one important inspection item adopted.
　In addition, a criterion for re-adopting important inspection items may be incorporated into the second trained model by providing a predetermined calculation or threshold value.
[0086]
(Other Aspects of Malaria Judgment Using Artificial Intelligence) In
　the construction of the learned model, the above-mentioned aspect 1 (or a modified aspect of aspect 1) and aspect 2 are combined in two stages, and a third learned model has been learned. Model M3 may be constructed. That is, the third trained model M3 uses the following data (a3) ​​and (b3) as teacher data, and which of the specified important test items is malaria positive to what extent. It is formed by letting artificial intelligence machine-learn whether it is related to.
　(A3) A test measurement value (a3) ​​obtained for an important test item specified by the first learned model M1 (or M11) of a blood sample (A) known to be malaria positive as a measurement target. A11) and information (A21) that the test measurement value (A11) is that of a blood sample positive for malaria.
　(B3) Test measurement values ​​(b3) obtained for important test items identified by the first trained model M1 (or M11) above, using a blood sample (B) known to be malaria-negative as a measurement target. B11) and information (B21) that the test measurement value (B11) is for a malaria-negative blood sample.
[0087]
　In the third trained model M3, when the second trained model M2 is constructed, the blood test items are specified in advance by the first trained model M1, and the items that do not contribute to the determination are excluded. .. As a result, the weight obtained by the third trained model M3 (which of the identified important test items is related to malaria positivity to what extent) is determined by the second learning. In some cases, it is possible to give a malaria determination with higher accuracy than the weight obtained by the completed model M2 alone.
[0088]
　On the other hand, when the above aspect 2 is carried out independently, the important test items and their weights are specified from all the blood test items, so that the malaria is determined by the influence of the blood test items having a minute relevance. May be less accurate. This is thought to be due to the fact that blood test items with minor relevance are not adopted as important test items. In such cases, the accuracy of malaria determination may be improved by applying principal component analysis or sparse optimization to the degree of relevance.
[0089]
　In the above embodiment, the threshold value for determining the CRP value is obtained experimentally in advance and stored in the reference data holding unit 432, but the threshold value for determining the CRP value is learning constructed by machine learning. It may be obtained by a finished model. Further, in each aspect of the malaria determination using the artificial intelligence described above, various trained models may be similarly constructed and acquired for each threshold value used for the malaria determination.
[0090]
(Computer Program for Malaria Judgment Using Artificial Intelligence (1))
　The judgment step in the
　above computer program is related to the malaria positivity identified by the above-mentioned first trained model M1 or M11. It may be configured to have a step of determining malaria of a blood sample based on the test measurement values ​​for the above-mentioned important test items by using the important test items.
　Important inspection items and malaria determination are as described in the explanation of the device.
[0091]
(Computer Program for Malaria Judgment Using Artificial Intelligence (2))
　The judgment step in the
　above computer program is the association of each of the above blood test items for malaria positivity specified by the above second learned model M2. It is composed of a step of determining the possibility of malaria positive in the blood sample based on the test measurement value for the blood test item and the specified degree of relevance using the degree of sex. It's okay.
　Important inspection items and malaria determination are as described in the explanation of the device.
[0092]
　The computer program according to the present invention is configured to cause a computer to execute various aspects of the determination operation of the above-mentioned device.
[0093]
(Method of the
　present invention ) The blood analysis method according to the present invention is a method for determining malaria. The method includes steps s20 of preparing test measurements of blood samples to be analyzed, as shown in the flowchart of FIG. Here, the test measurement value is a measurement value for each of a predetermined blood test item including a CRP value and a blood cell count value. Next, the method includes step s30 for determining the possibility of malaria positive in the blood sample based on the test measurement values ​​for blood test items other than the CRP value and / or the test measurement values ​​for the CRP value. Have.
　Important inspection items and malaria determination are as described in the explanation of the device.
[0094]
(Aspect 1 of Malaria Judgment in Blood Analysis Method Using Artificial Intelligence) In
　this aspect 1, the determination step s30 is specified from the blood test items by the first learned model M1 described above. , Use important test items that are relevant for malaria positivity. Then, it has a step of determining the possibility of malaria positive in the blood sample based on the test measurement value for the important test item.
[0095]
(Aspect 2 of malaria determination in a blood analysis method using artificial intelligence) In
　this aspect 2, the determination step s30 is the blood test item for malaria positivity specified by the second learned model M2 described above. Use the degree of relevance of. Then, it has a step of determining the possibility of malaria positive in the blood sample based on the test measurement value for the blood test item and the degree of the identified relevance.
[0096]
　The description of the operation contents of each part for malaria determination described above in the description of the device is also a description of the computer program of the present invention and also a description of the blood analysis method of the present invention.
Industrial applicability
[0097]
　INDUSTRIAL APPLICABILITY According to the present invention, it has become possible to inexpensively provide a blood analyzer capable of determining malaria even though it has an inexpensive configuration. Further, the present invention has made it possible to provide a computer program capable of causing a computer to perform malaria determination more preferably, and a blood analysis method capable of performing malaria determination more preferably.
[0098]
　This application is based on Japanese Patent Application No. 2018-246186 filed in Japan (Filing date: December 27, 2018), the contents of which are incorporated herein by reference in its entirety.
Code description
[0099]
　　M1 Blood sample
　　10 Measuring mechanism
　　11 First particle measuring unit
　　12 Second particle measuring unit
　　13 CRP measuring unit
　　14 Dispensing mechanism
　　20 Control unit
　　30 Mechanism control unit
　　40 Analytical unit
　　410 Measured value receiving unit (volume value receiving unit and (Including CRP value receiving unit)
　　420 Measured value processing unit
　　421 Frequency distribution calculation unit
　　423 CRP value determination unit
　　430 Judgment unit
　　440 Judgment result output unit

WE CLAIMS

[Claim 1]A blood analyzer
　having a measurement value receiving unit that receives a test measurement value of a blood sample to be analyzed, and the
　　test measurement value includes a CRP value based on a CRP parameter value and a blood cell count value. The
　　measured value receiving unit has a CRP value receiving unit that receives a CRP parameter value in a sample solution containing the blood sample, and the
　blood analyzer has a CRP. The
blood analyzer having a determination unit for determining the possibility of malaria positive in the blood sample based on the test measurement values ​​for blood test items other than the values ​​and / or the specific test measurement values ​​for the CRP value. ..
[Claim 2]
　The measurement value receiving unit has a volume value receiving unit that receives a volume measurement value of particles in a sample solution containing the blood sample, and the
　determination unit has a predetermined volume frequency distribution based on the volume measurement value. When the frequency peak portion is present in the volume range, or when the CRP value based on the CRP parameter value is equal to or higher than a predetermined threshold value, it is determined that the blood sample is likely to be malaria positive.
The blood analyzer according to claim 1.
[Claim 3]
　A leukocyte measurement chamber is further provided as a first particle measurement unit,
　and in the leukocyte measurement chamber, the supplied blood sample is subjected to a hemolytic treatment for dissolving erythrocytes and diluted to form the sample solution, and the sample is formed. The blood analyzer according to claim 2,
　wherein the volumes of the leukocytes and other particles in the liquid are measured, and the volume value receiving unit receives the volume measurement value from the first particle measuring unit
.
[Claim 4]
The blood analyzer according to claim 2 or 3 　, further comprising a CRP measuring unit for measuring the CRP parameter value, and the
　CRP value receiving unit receives the CRP parameter value from the CRP measuring unit
.
[Claim 5]
　When the frequency peak portion exists within a predetermined volume range in the volume frequency distribution based on the volume measurement value and the CRP value based on the CRP parameter value is equal to or higher than a predetermined threshold value, the determination unit determines.
The blood analyzer according to any one of claims 2 to 4 , wherein the blood sample is determined to have a high possibility of being malaria positive .
[Claim 6]
　The determination unit uses
　　the important test items related to the positive for malaria, which are identified from the predetermined blood test items by the following first trained model (M1), and the
　　above-mentioned It is configured to determine the possibility of malaria positive in the blood sample based on the test measurement values ​​for important test items
　, wherein the first trained model (M1) is 　　　malaria positive
　　as teacher data.
The test measurement value (A1) obtained for the predetermined blood test item with the blood sample (A) known to be the measurement target and the test measurement value (A1) of the blood sample positive for malaria. The
　　　test measurement value (B1) obtained for the predetermined blood test item with the information (A2) that there is, and the blood sample (B) known to be negative for malaria as the measurement target, and the test measurement value (B1). ) Is from a blood sample that is negative for malaria (B2), and
is
　　formed by machine learning which blood test item among the predetermined blood test items is related to positive for malaria. It is,
blood analyzer according to claim 1.
[Claim 7]
　The important test items are
　　whether or not there is a frequency peak in the volume range of 25 fL to 45 fL in the volume frequency distribution of leukocytes in the
　　blood sample
　　to be analyzed, the number of leukocytes in the blood sample to be analyzed, and the blood to be analyzed. The
　　blood analyzer according to claim 6, wherein the blood analyzer
comprises one or more selected from the group consisting of the number of white blood cells in the sample and the CRP value of the blood sample to be
analyzed.
[Claim 8]
　The determination unit uses
　　the degree of relevance of each of the predetermined blood test items to the positive malaria identified by the second trained model (M2) below, and the
　　test measurement value for the predetermined blood test item. And, based on the degree of association identified, the blood sample is configured to determine the likelihood of being malaria positive
　, wherein the second trained model (M2) is
　　teacher data. As
　　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target, and blood whose test measurement value (A1) is malaria positive. The information (A2) that
　　　it belongs to a sample, the test measurement value (B1) obtained for the predetermined blood test item with the blood sample (B) known to be negative for malaria as the measurement target, and the test. measured value (B1) and information (B2) that is of the blood samples of malaria-negative
using,
　　with relevance to any blood test items What about malaria positive degree of one of the predetermined blood test item
The blood analyzer according to claim 1, which is formed by machine-learning whether or not the blood is present.
[Claim 9]
　The step of accepting a test measurement value of a blood sample to be analyzed, wherein the test measurement value is a measurement value for each of a predetermined blood test item including a CRP value and a blood cell count value.
　Have the
computer perform a step of determining the likelihood of malaria positive in the blood sample based on the test measurements for a particular blood test item other than the CRP value and / or the test measurements for the CRP value . Computer program.
[Claim 10]
　A
　step of accepting a volume measurement value of particles in a sample solution containing the blood sample, a step of accepting a CRP parameter value in the sample solution, and
　a volume range peculiar to malaria protozoa in the volume frequency distribution based on the volume measurement value. A step of determining whether or not a frequency peak portion exists in the inside, a step of determining whether or not the
　CRP value based on the CRP parameter value is equal to or higher than a predetermined threshold value, and
　a case where the frequency peak portion exists. The computer program according to claim 9 , wherein the
computer is made to perform a step of determining that the blood sample is likely to be malaria positive when the CRP value is equal to or higher than the threshold value
.
[Claim 11]
　A computer program further comprises a computer program that causes a computer to perform a step of determining that the blood sample is likely to be malaria positive when the frequency peak portion is present and the CRP value is equal to or higher than the threshold value. Item 10. The computer program according to item 10.
[Claim 12]
　The determination step uses
　　the important test items related to the positive for malaria, which are identified from the predetermined blood test items by the following first trained model (M1), and
　　described above.
　The first trained model (M1) has a step of determining the possibility of being positive for malaria in the blood sample based on the test measurement values ​​for the important test items of the above , wherein the first trained model (M1) has 　　　malaria
　　as teacher data.
A test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be positive as a measurement target, and a blood sample whose test measurement value (A1) is malaria positive. (A2),
　　　a test measurement value (B1) obtained for the predetermined blood test item using a blood sample (B) known to be negative for malaria as a measurement target, and the test measurement value (B1).
By using the information (B2) that B1) belongs to a blood sample negative for malaria
　　, machine learning which of the predetermined blood test items is related to positive malaria is performed.
The computer program according to claim 9, which is formed .
[Claim 13]
　The determination step uses
　　the degree of relevance of each of the predetermined blood test items to the positive malaria identified by the second trained model (M2) below, and the
　　test measurement for the predetermined blood test item.
　The second trained model (M2) comprises a step of determining the likelihood of a malaria positive in the blood sample based on the value and the degree of association identified above , wherein the second trained model (M2) is a
　　teacher. As data,
　　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target and the test measurement value (A1) are malaria positive. The information (A2) that
　　　it belongs to a blood sample, the test measurement value (B1) obtained for the predetermined blood test item with the blood sample (B) known to be negative for malaria as the measurement target, and the said and information (B2) of test measurement (B1) is of a blood sample of malaria-negative
using a
　　relevance malaria positive in any degree of blood test items as many of said predetermined blood test item
The computer program according to claim 9, which is formed by machine learning what to have .
[Claim 14]
　It has a step of preparing a test measurement value of a blood sample to be analyzed, and the test measurement value is a measurement value for each of a predetermined blood test items including a CRP value and a blood cell count value, and
　A
blood analysis method comprising the step of determining the possibility of malaria positive in the blood sample based on the test measurement value for a specific blood test item other than the CRP value and / or the test measurement value for the CRP value .
[Claim 15]
　The determination step uses
　　the important test items related to the positive for malaria, which are identified from the predetermined blood test items by the following first trained model (M1), and
　　described above.
　The first trained model (M1) has a step of determining the possibility of being positive for malaria in the blood sample based on the test measurement values ​​for the important test items of the above, and the first trained model (M1) is 　　　positive for malaria
　　as teacher data.
It is said that the test measurement value (A1) obtained for the predetermined blood test item using the blood sample (A) known to be measured and the test measurement value (A1) are those of a blood sample positive for malaria. The information (A2),
　　　the test measurement value (B1) obtained for the predetermined blood test item with the blood sample (B) known to be negative for malaria as the measurement target, and the test measurement value (B1) are 　　It is formed by machine learning which blood test item among the predetermined blood test items is related to the positive malaria
using the information (B2) that the blood sample is negative for malaria.
It is,
blood analysis method according to claim 14.
[Claim 16]
　The determination step uses
　　the degree of relevance of each of the predetermined blood test items to the positive malaria identified by the second trained model (M2) below, and the
　　test measurement for the predetermined blood test item.
　The second trained model (M2) comprises a step of determining the likelihood of a malaria positive in the blood sample based on the value and the degree of association identified above , wherein the second trained model (M2) is a
　　teacher. As data,
　　　a test measurement value (A1) obtained for the predetermined blood test item using a blood sample (A) known to be malaria positive as a measurement target and the test measurement value (A1) are malaria positive.
　　　Test measurement value (B1) obtained by measuring the predetermined blood test items with the information (A2) that it belongs to a blood sample and the blood sample (B) known to be negative for malaria as measurement targets. And,
using the information (B2) that the test measurement value (B1) is that of a blood sample negative for malaria
　　, which blood test item among the predetermined blood test items becomes positive for malaria to what extent.
The blood analysis method according to claim 14, which is formed by machine learning whether or not it has relevance .