Automatic analysis device

Abstract

Provided is an automatic analysis device capable of calculating a stable measurement value even when temperature fluctuation occurs in a light source or a photometer. The automatic analysis device includes: a light source 104 that comprises a first temperature measurement unit 209 and irradiates a container 203 having stored therein a mixed liquid of a sample and a reagent with light; a photometer 208 comprising a second temperature measurement unit 210 and whereby the amount of light from the light source that has passed through the mixed liquid is measured; a light amount correction value calculation unit 211 that calculates a light amount correction value on the basis of the light source temperature measured by the first temperature measurement unit and the photometer temperature measured by the second temperature measurement unit; and an absorbance calculation unit 205 that calculates the absorbance of the mixed liquid on the basis of the light amount measured by the photometer and the light amount correction value.

Description

automatic analyzer

[0001] The present invention relates to an automatic analyzer.

[0002] In automated analyzers that analyze samples such as blood, the light source is one of the important components that determines analytical performance. Light emitted from the light source passes through a reaction solution containing a mixture of reagents and samples, is split into a specific number of wavelengths by a spectroscope, and is then detected by a detector. Because of the detection process described above, it is important to ensure stability in the amount of light emitted from the light source and the amount of light received by the detector. It is known that the optical system of an automated analyzer is affected by the ambient temperature or the temperature of the light source and detector. Therefore, technology is needed to stabilize the temperature of the light source and detector, or to compensate for these temperature fluctuations.

[0003] As a method for correcting the light intensity related to the optical system of an automatic analyzer, Patent Document 1 discloses a technique for correcting absorbance using the environmental temperature around the light source and a correction factor calculated before analysis.

[0004] Japanese Patent Application Laid-Open No. 2022-164188

[0005] The method disclosed in Patent Document 1 uses only the temperature on the light source side for correction, which has the problem of not reflecting temperature fluctuations on the detector side. For example, if the temperature on the light source and detector sides rise by 1°C, there are cases in which the light intensity on the light source side decreases and the light intensity on the detector side increases. In such a situation, if correction is performed based only on temperature fluctuations on the light source side, the correction will be made in the direction of excessively increasing the light intensity, which may result in an inaccurate measurement value.

[0006] Furthermore, the method of Patent Document 1 corrects for absorbance rather than light intensity. Generally, absorbance is calculated using the amount of incident light before passing through a reaction solution and the amount of transmitted light after passing through the reaction solution, so it is appropriate to correct for light intensity rather than absorbance.

[0007] Furthermore, there are individual differences between light sources and photometers, so it is desirable to have a mechanism that can correct the light intensity or adjust for temperature fluctuations while taking into account the individual differences.

[0008] An object of the present invention is to provide an automatic analyzer that can calculate stable measurement values ​​by canceling or reducing the temperature effects on both the light source and detector sides.

[0009] An automated analyzer according to one embodiment of the present invention includes a light source having a first temperature measurement unit that irradiates light onto a container containing a mixed solution of a sample and a reagent; a photometer having a second temperature measurement unit that measures the amount of light from the light source that has passed through the mixed solution; a light intensity correction value calculation unit that calculates a light intensity correction value based on the light source temperature measured by the first temperature measurement unit and the photometer temperature measured by the second temperature measurement unit; and an absorbance calculation unit that calculates the absorbance of the mixed solution based on the light intensity and light intensity correction value measured by the photometer.

[0010] It is possible to provide an automatic analyzer that can calculate stable measurement values ​​even when temperature fluctuations occur in the light source or photometer (detector). Other objects and novel features will become apparent from the description of this specification and the accompanying drawings.

[0011] FIG. 1 is a schematic diagram showing the overall configuration of an automatic analyzer. FIG. 1 is a diagram for explaining an optical system of an automatic analyzer according to Example 1 and devices arranged in the periphery thereof. FIG. 2 is a functional block diagram for explaining a light intensity correction function of Example 1. FIG. 3 is a measurement value calculation flow of Example 1. FIG. 4 is a graph showing the relationship between temperature and light intensity for each wavelength. FIG. 5 is a diagram for explaining an optical system of an automatic analyzer according to Example 2 and devices arranged in the periphery thereof. FIG. 6 is a functional block diagram for explaining a temperature adjustment function of Example 2. FIG. 7 is a measurement value calculation flow of Example 2. FIG. 8 is a graph showing the relationship between light source temperature and photometer adjustment temperature for each wavelength. FIG. 9 is a hardware configuration example of a learning device. FIG. 10 is a functional block diagram for explaining training of a learning model according to Example 3. FIG. 11 is an example of a learning model. FIG. 12 is a functional block diagram for explaining a light intensity correction function of Example 3. FIG. 13 is a functional block diagram for explaining a temperature adjustment function of Example 3.

[0012] Figure 1 is a schematic diagram of the automatic analyzer. A display unit 101 displays operations and measurement information being performed by the automatic analyzer. An input unit 102 inputs requests for analytical operations to be performed by the automatic analyzer. An analysis unit 103 is equipped with various operating mechanisms for executing analytical operations. A light source 104 is one of the elements that make up the analysis unit 103, and projects light onto the object to be measured. A control unit 105 controls the various operating mechanisms that make up the analysis unit 103 based on operation instructions input from the input unit 102, etc., and also calculates measurement data.

[0013] FIG. 2 illustrates the optical system used in the automated analyzer of Example 1 and the surrounding devices. The light source 104 is composed of LED elements. There is no limit to the number of LED elements constituting the light source 104, and multiple elements can be used. When the light source 104 has multiple LED elements, the amount of current supplied to each LED element is set individually. During operation of the analysis unit 103, light from the light source 104 is irradiated onto the reaction vessel 203, which passes through the photometry position between the light source 104 and the diffraction grating 207. In the mixture in the reaction vessel 203, the test sample components react with the reagent, producing or consuming a substance to be measured in proportion to the concentration of the test component. Note that reaction vessel water (a constant-temperature medium) is located between the reaction vessel 204 and the reaction vessel 203.

[0014] Of the light irradiated onto the mixed liquid, light with wavelengths in an absorption region corresponding to the substance being measured is absorbed by the substance being measured. The light that passes through the mixed liquid is incident on a concave diffraction grating 207 installed in a photometer 208. The diffraction grating 207 separates the incident light into wavelengths and outputs the separated light to a photodetector 206. The photodetector 206 converts the amount of light into an electrical signal and outputs the electrical signal to an absorbance calculation unit 205. The absorbance calculation unit 205 calculates the absorbance and outputs it to an apparatus control unit 212. As will be described later, the absorbance calculation unit 205 calculates the absorbance based on a light intensity correction value calculated by a light intensity correction value calculation unit 211 based on the electrical signal output from the photodetector 206 and temperature data acquired by temperature measurement units 209 and 210.

[0015] The device control unit 212 performs colorimetric analysis based on the absorbance output from the absorbance calculation unit 205. When performing colorimetric analysis, cell blank water is dispensed into all reaction cells, and the absorbance at each wavelength from 340 to 800 nm is measured (cell blank measurement). The device control unit 212 retains (stores) this measurement result as a cell blank value (also referred to as baseline light intensity). The device control unit 212 compares the cell blank value with the absorbance of the mixed solution to be analyzed, corrects the absorbance, and outputs the corrected absorbance as measurement data to a user interface (such as a screen displayed on the display unit 101).

[0016] The current detection unit 201 monitors (measures) the current flowing through the light source 104. The current adjustment unit 202 has a circuit that reduces the amount of current supplied to the light source 104 or turns off the power supply to the light source 104 at a timing that does not affect the analysis.

[0017] The hardware of the control unit 105 shown in FIG. 1 is a computer that functions as a functional unit that provides predetermined functions by executing a program. The device control unit 212, light intensity correction value calculation unit 211, and absorbance calculation unit 205 shown in FIG. 2, and the measurement value calculation unit 301 shown in FIG. 3 are all examples of functional units. In the following, the computer (control unit 105) is the main hardware of these functional units. The computer executes processing according to a program loaded into memory using a processor, appropriately using resources such as memory and a communication interface. Furthermore, the processing required to realize the functions is not limited to software program processing, but can also be implemented using a dedicated circuit. Examples of the dedicated circuit include a field programmable gate array (FPGA) and an application-specific integrated circuit (ASIC).

[0018] The light intensity correction function of the first embodiment will be described with reference to FIG. 3 . The light source 104 emits light when a current is supplied from the current adjustment unit 202. The light emitted from the light source 104 is measured as a light intensity by the photometer 208, and the light intensity data is sent to the absorbance calculation unit 205. The light source 104 includes a temperature measurement unit 209 that measures the light source temperature, and the photometer 208 includes a temperature measurement unit 210 that measures the photometer temperature. The temperature data measured by the temperature measurement units 209 and 210 is sent to the light intensity correction value calculation unit 211. The light intensity correction value calculation unit 211 calculates a light intensity correction value based on a light intensity correction factor calculated and stored in advance and the temperature data sent from the temperature measurement units 209 and 210. In this way, by using both the temperature information of the light source 104 and the temperature information of the photometer 208, it is possible to obtain light intensity information that is free from the effects of temperature fluctuations on both the light source side and the detector side.

[0019] The calculated light intensity correction value is sent to the absorbance calculation unit 205. The absorbance calculation unit 205 calculates the absorbance from the light intensity data and the light intensity correction value. The calculated absorbance is sent to the measurement value calculation unit 301. The measurement value calculation unit 301 calculates the measurement value set for each measurement item from the absorbance data sent to it. The calculated measurement value is displayed to the user on the display unit 101.

[0020] Here, the light intensity correction value calculation unit 211 calculates a light intensity correction value based on the temperature data and light intensity correction factor sent from the temperature measurement units 209 and 210. The light intensity correction value calculation unit 211 may store a light intensity correction factor for light source temperature correction and a light intensity correction factor for photometer temperature correction. The light intensity correction factor for light source temperature correction can be obtained by measuring the light intensity from the light source with a photometer at a constant photometer temperature while changing the ambient temperature of the light source. Furthermore, the light intensity correction factor for photometer temperature correction can be obtained by measuring the light intensity from the light source with a photometer at a constant light source temperature while changing the photometer temperature. In this case, the light intensity correction value is obtained as the sum of the light intensity correction value based on the light source temperature change of the light source 104 and the light intensity correction value based on the photometer temperature change of the photometer 208.

[0021] Furthermore, the light intensity correction rate may be a value specific to each device. Therefore, it is preferable to be able to calculate the light intensity correction rate on the actual device. In particular, when an LED light source is used as the light source 104, since LED light sources are consumables, it is preferable that the device itself be able to calculate the light intensity correction rate for light source temperature correction for the LED light source. For example, when starting up the device from power off, the light intensity correction rate for light source temperature correction can be calculated based on the change in light intensity from when the light source 104 is turned on until the light intensity stabilizes, and the temperature data measured by the temperature measurement unit 209 during that period.

[0022] In this way, by using the calculation method of the light intensity correction value shown in Fig. 3, it is possible to calculate a light intensity correction value that corresponds to temperature fluctuations at two locations: the light source and the photometer. Furthermore, if the light intensity correction rate is set to a value specific to the device, it is possible to calculate a light intensity correction value that reflects individual differences between devices.

[0023] Next, the measurement value calculation flow of the first embodiment will be described with reference to FIG. 4 . This flow is executed by the control unit 105. When measurement by the automatic analyzer is started, first, a measurement preparation operation S401 is started. During the measurement preparation operation S401, a reference temperature measurement S402 is performed. In the reference temperature measurement S402, the temperature measurement value at that time measured by the temperature measurement units 209 and 210 is stored. Next, the process proceeds to a measurement operation S403. During the measurement operation S403, in parallel with a light intensity measurement S404, a temperature information acquisition S405 is performed by the temperature measurement units 209 and 210, and a light intensity correction value calculation S406 is performed using the acquired temperature data. In the light intensity correction value calculation S406, the difference between the reference temperature measured in the reference temperature measurement S402 and the temperature at the time of light intensity measurement acquired in the temperature information acquisition S405 is calculated, and a light intensity correction value is calculated using the difference and a light intensity correction factor. The absorbance calculation S407 is performed using the light intensity data measured in the light intensity measurement S404 and the light intensity correction value calculated in the light intensity correction value calculation S406. Thereafter, the measurement value is calculated in the measurement value calculation S408 using the absorbance data calculated in the absorbance calculation S407.

[0024] Figure 5 is a graph showing the relationship between temperature and light intensity for each wavelength. The horizontal axis represents temperature, and the vertical axis represents light intensity. The solid line shows the change in light intensity with temperature for wavelength A, and the dashed line shows the change in light intensity with temperature for wavelength B. The slope of this graph corresponds to the light intensity correction factor. Let's assume that the light intensity at a reference temperature t0 is 100%, and the light intensity at wavelength A is a% and the light intensity at wavelength B is b% at an arbitrary temperature t1. Figure 5 shows an example where the relationship b<100<a is true. As such, the temperature dependence of light intensity generally differs for each wavelength of light. Note that this difference in temperature dependence with wavelength applies to both light source temperature and photometer temperature. Therefore, it is desirable for the light intensity correction value calculation unit 211 to store a light intensity correction factor for light source temperature correction and a light intensity correction factor for photometer temperature correction for each wavelength used in measurement.

[0025] For example, the light intensity correction factor is calculated for each wavelength, the temperature before the cell blank water measurement is set to the reference temperature t0, the light intensity at the start of the cell blank water measurement is set to 100%, and the baseline light intensity used to calculate absorbance is corrected for temperature fluctuations from the reference temperature t0. For example, if the light intensity at a certain wavelength is +1% / 1°C when the temperature rises by 0.1°C from the reference temperature t0, the corrected baseline light intensity will be 100.1%. By using the light intensity correction factor for each wavelength in this way, it is possible to perform appropriate light intensity correction for fluctuations from the reference temperature for each wavelength.

[0026] An optical system used in the automatic analyzer of Example 2 and devices arranged around it will be described with reference to Figure 6. Components common to Example 1 will be assigned the same reference numerals, and duplicated descriptions will be omitted. The following description will focus on differences from Example 1.

[0027] Photometer 208 is equipped with a temperature adjustment unit 602 that adjusts the ambient temperature. Temperature data measured by temperature measurement units 209 and 210 is sent to temperature adjustment value calculation unit 601. Temperature adjustment value calculation unit 601 calculates a temperature adjustment value from the sent temperature data and uses this value to adjust the output of temperature adjustment unit 602 installed in photometer 208. The device control unit 212, temperature adjustment value calculation unit 601, and absorbance calculation unit 205 shown in Figure 6, and measurement value calculation unit 301 shown in Figure 7 are all functional units, and the main hardware component is a computer (control unit 105).

[0028] The temperature adjustment function of the second embodiment will be described with reference to FIG. 7 . Temperature data measured by the temperature measurement units 209 and 210 is sent to a temperature adjustment value calculation unit 601. The temperature adjustment value calculation unit 601 calculates a temperature adjustment value based on a temperature adjustment rate calculated and stored in advance, and the differences between the light source temperature and photometer temperature sent from the temperature measurement units 209 and 210 and their respective reference temperatures. The calculated temperature adjustment value is sent to a temperature adjustment unit 602. The temperature adjustment unit 602 changes its output in accordance with the sent temperature adjustment value, adjusting the temperature of the photometer 208. Light emitted from the light source 104 is measured as a light intensity in the photometer 208, and the light intensity data is sent to an absorbance calculation unit 205, which calculates the absorbance. The calculated absorbance is sent to a measurement value calculation unit 301. The measurement value calculation unit 301 calculates a measurement value set for each measurement item from the sent absorbance data.

[0029] Here, the temperature adjustment rate is set so as to offset changes in light intensity due to temperature fluctuations in the light source 104 by adjusting the temperature of the photometer 208. The temperature adjustment rate may also be a value specific to each device. As described in the first embodiment, it is possible to calculate the light source correction rate for light source temperature correction using an actual device, and the temperature dependence of light intensity due to changes in the temperature of the light source can be determined. Meanwhile, the temperature dependence of light intensity due to temperature changes in the non-consumable photometer is pre-defined for the device, and the temperature dependence of light intensity due to changes in the temperature of the photometer is known. With information on these two temperature dependencies, it is possible to determine a temperature adjustment rate for the photometer 208 for each device that offsets changes in light intensity due to temperature fluctuations in the light source 104.

[0030] By using the method for calculating the temperature adjustment value shown in Figure 7 in this way, it is possible to cancel out or reduce the fluctuation in light intensity caused by temperature fluctuations on the light source side by using the change in light intensity caused by temperature changes on the photometer side.

[0031] Next, the measurement value calculation flow of the second embodiment will be described with reference to Fig. 8. This flow is executed by the control unit 105. Steps common to Fig. 3 are assigned the same reference numerals, and duplicated explanations will be omitted. During the measurement operation S403, temperature information acquisition S405 is performed, and the acquired temperature data is used to calculate a temperature adjustment value S801. The temperature adjustment value calculated in temperature adjustment value calculation S801 is used to adjust the temperature of the photometer S802.

[0032] Figure 9 shows the relationship between the light source temperature and the photometer adjustment temperature for each wavelength. The photometer adjustment temperature is the amount of adjustment of the photometer temperature to offset changes in light intensity due to changes in the light source temperature. The horizontal axis represents the light source temperature, and the vertical axis represents the photometer adjustment temperature. The relationship of the photometer adjustment temperature to the light source temperature for wavelength A is shown by a solid line, and the relationship of the photometer adjustment temperature to the light source temperature for wavelength B is shown by a dashed line. Let the photometer adjustment temperature at a reference temperature t0 be 0°C, and let's say that the photometer adjustment temperature for wavelength A at an arbitrary temperature t1 is c°C, and the photometer adjustment temperature for wavelength B is d°C. As such, the temperature adjustment rate generally differs for each wavelength of light.

[0033] For this reason, it is desirable to set a temperature adjustment rate according to the wavelength of light and select a temperature adjustment rate for one of the wavelengths. For example, select the temperature adjustment rate for the wavelength that is most affected by the change in light intensity due to temperature fluctuation. For example, when the temperature rises by +0.01°C at the selected wavelength, if the temperature adjustment rate is +5°C / 1°C, the adjusted temperature value is calculated as 0.05°C, and the output of the temperature adjustment unit 602 is adjusted to cause a temperature fluctuation of that value in the photometer. In this way, the change in light intensity due to temperature fluctuation on the light source side is controlled to be offset by causing a temperature fluctuation on the photometer side.

[0034] In principle, it is possible to provide a temperature adjustment unit in the light source and offset changes in light intensity due to temperature fluctuations on the photometer side with temperature fluctuations on the light source side; however, since the light source is a replaceable part, installing a temperature adjustment unit would increase operating costs. The configuration of Example 2 avoids increases in operating costs by providing a temperature adjustment unit 602 in the photometer 208, which is not a replaceable part. Furthermore, the fluctuations in light intensity due to temperature fluctuations on the photometer side are smaller than on the light source side, making it possible to eliminate the need for more precise control than controlling the light source side.

[0035] In the third embodiment, a trained machine learning model is applied. Specifically, the light intensity correction value calculated by the light intensity correction value calculation unit 211 in the first embodiment or the temperature adjustment value calculated by the temperature adjustment value calculation unit 601 in the second embodiment can be corrected using the trained machine learning model. The light intensity correction value calculation unit 211 and the temperature adjustment value calculation unit 601 are assumed to calculate the light intensity correction value and the temperature adjustment value based on a model with the light source temperature and the photometer temperature as variables. However, variations due to individual differences in the light source and the photometer, the influence of the device structure, and the like may result in an insufficient effect of counteracting the influence of temperature changes on the light intensity correction value and the temperature adjustment value. Therefore, a trained model is used that enables the light intensity correction value and the temperature adjustment value to be corrected using additional explanatory variables.

[0036] The following description will be given taking as an example the learning model used to calculate the light intensity correction value in Example 1. Fig. 10A shows an example of the hardware configuration of a learning device 1000 that trains the learning model. The learning device 1000 is configured to be able to transmit and receive data to and from the control unit 105.

[0037] The learning device 1000 is realized by an information processing device including, as its main components, a processor (CPU) 1001, a memory 1002, a storage device 1003, an input interface (I / F) 1004, an output I / F 1005, a communication I / F 1006, and a bus 1007. The processor 1001 functions as a functional unit (functional block) that provides a predetermined function by executing processing according to a program loaded into the memory 1002. The storage device 1003 stores data and programs used by the functional unit. The storage device 1003 may be a non-volatile storage medium such as an HDD (Hard Disk Drive) or an SSD (Solid State Drive). The input I / F 1004 is an interface for connecting an input device 1008 such as a keyboard or pointing device, and the output I / F 1005 is an interface for connecting a display device 1009. The communication I / F 1006 enables communication with the control unit 105 and other information processing devices. These are communicatively connected to each other via the bus 1007.

[0038] Note that all or part of the software functions may be implemented using hardware such as FPGAs or ASICs. Furthermore, all or part of the functions may be implemented using multiple information processing devices, or as a cloud application. Figure 10A also shows the processor 106, storage device 107, and communication I / F 108 of the control unit 105, which have the same hardware configuration as the learning device 1000.

[0039] 10B shows a functional block diagram for training (learning) the learning model of Example 3. The storage device 107 of the control unit 105 stores the learning model 1014 that has been trained by the learning device 1000. Learning data 1010 for generating training data is input to the learning device 1000.

[0040] The training data 1010 includes, for example, experimental data 1011 collected by repeatedly measuring multiple samples having a predetermined absorbance relationship using a pre-shipment device, optical system individual data 1012 about the optical system of the device used in the experiment, and a light intensity correction value calculation formula 1013 used to correct the light intensity. The experimental data 1011 collected at this time includes temperature data from the temperature measurement units 209 and 210, light intensity data from the photometer 208, as well as room temperature data. The optical system individual data 1012 also includes the measurement wavelength, photometer wavelength adjustment value, initial light intensity, and light source usage time. The measurement wavelength is the wavelength measured by the photometer 208, the photometer wavelength adjustment value is an adjustment value used to correct variations in the wavelength measured by the photometer 208, the initial light intensity is the light intensity when the light source 104 begins to be used, and the usage time is the cumulative lighting time of the light source 104. The light intensity correction value calculation formula 1013 is a correction formula used by the light intensity correction value calculation unit 211 to calculate the light intensity correction value.

[0041] FIG. 11 illustrates an example of a neural network learning model 1100. The learning model 1100 includes an input layer 1101, a hidden layer 1102, and an output layer 1103. Explanatory variables input to the input layer 1101 are propagated to the hidden layer 1102 and then to the output layer 1103. The output layer 1103 outputs an inference result (objective variable) based on the explanatory variables input to the input layer 1101. Note that the hidden layers of a neural network generally consist of multiple layers, and are shown simplified here. The input layer 1101, the hidden layer 1102, and the output layer 1103 each include multiple input units, hidden units, and output units, indicated by circles. Input data (explanatory variables) input to each input unit of the input layer 1101 is weighted by the coupling coefficient between the input unit and the hidden unit and then input to each hidden unit. The value of the hidden unit in the hidden layer 1102 is calculated by adding values ​​from the input units. Furthermore, the output from each hidden unit in the hidden layer 1102 is weighted by the coupling coefficient between the hidden unit and the output unit, and is input to each output unit. The values ​​from the hidden units are added together to calculate the value of the output unit in the output layer 1103. In this way, the processing in the hidden layer 1102 corresponds to nonlinearly transforming the value of the input data input to the input layer 1101 and outputting it as output data from the output layer 1103.

[0042] For example, input data 1104, which is an explanatory variable, is data acquired as learning data 1010. Output data 1105, which is an objective variable, has three labels: "maintain light intensity adjustment value," "correct light intensity adjustment value," and "unable to calculate light intensity adjustment value (abnormal)."

[0043] The training data generation unit 1022 shown in FIG. 10B generates training data, which is a combination of input data 1104 for the learning model 1100 and its correct label, from the learning data 1010. The input data 1104 serving as training data is obtained by extracting parameter values ​​from the learning data 1010. The correct label is calculated from the learning data 1010. For example, the estimated light intensity for each sample is compared with the light intensity calculated using the light intensity correction value calculation formula 1013. If the two are deemed to be approximately equivalent, the unit determines that the light intensity adjustment value is to be maintained; if they are not equivalent but the difference is within a predetermined value, the unit determines that the light intensity adjustment value is to be corrected; and if the difference exceeds the predetermined value, the unit determines that the light intensity adjustment value cannot be calculated (abnormal). Furthermore, in the case of correcting the light intensity adjustment value, it is desirable to also learn the amount of correction to the light intensity adjustment value.

[0044] Using the training data generated by the training data generation unit 1022, the learning model update unit 1023 trains the learning model 1100. The trained learning model 1014 is transmitted from the output unit 1025 to the control unit 105, stored in the storage device 107, and used to determine and correct the light intensity correction value. Note that the prototype of the learning model 1100 is assumed to be stored in the algorithm database 1024. Here, an example of a learning model using a neural network is shown, but the present invention is not limited to this.

[0045] The light intensity correction function of the third embodiment will be described with reference to FIG. 12 . The difference from FIG. 3 (first embodiment) is the light intensity correction value correction determination unit 1201, which is a functional unit that determines whether or not the light intensity correction value needs to be corrected using the trained learning model 1014. The light intensity correction value correction determination unit 1201 determines whether or not the light intensity correction value calculated by the light intensity correction value calculation unit 211 needs to be corrected using the trained learning model 1014, and outputs the determination result to the light intensity correction value calculation unit 211. If the determination result is “maintain the light intensity adjustment value,” the light intensity correction value calculation unit 211 outputs the calculated light intensity correction value to the absorbance calculation unit 205. If the determination result is “correct the light intensity adjustment value,” the light intensity correction value calculation unit 211 corrects the calculated light intensity correction value based on the correction amount also output by the trained learning model 1014, and outputs the corrected value to the absorbance calculation unit 205. If the determination result is "light intensity adjustment value calculation impossible (abnormal)", the light intensity correction value calculation unit 211 stops the subsequent processing and requests the device control unit 212 to issue an alarm.

[0046] While the above describes an example of determining whether or not the light intensity correction value needs to be corrected using the trained learning model 1014 in the third embodiment, it is also possible to determine whether or not the temperature correction value in the second embodiment needs to be corrected in a similar manner. A trained learning model for determining whether or not the temperature adjustment value needs to be corrected can also be generated in a similar manner as described above. The learning data uses experimental data, optical system individual data about the optical system associated with the device from which the experimental data was acquired, and a temperature adjustment value calculation formula used by the temperature adjustment value calculation unit 601 to calculate the temperature adjustment value.

[0047] The temperature adjustment function of the third embodiment will be described with reference to FIG. 13 . The difference from FIG. 7 (second embodiment) is the temperature adjustment value correction determination unit 1301, which is a functional unit that determines whether the temperature adjustment value needs to be corrected using a trained learning model. The temperature adjustment value correction determination unit 1301 uses a trained learning model to determine whether the temperature adjustment value calculated by the temperature adjustment value calculation unit 601 needs to be corrected, and outputs the determination result to the temperature adjustment value calculation unit 601. If the determination result is “maintain the temperature adjustment value,” the temperature adjustment value calculation unit 601 outputs the calculated temperature adjustment value to the temperature adjustment unit 602. If the determination result is “correct the temperature adjustment value,” the temperature adjustment value calculation unit 601 corrects the calculated temperature adjustment value based on the correction amount also output by the trained learning model, and outputs the corrected value to the temperature adjustment unit 602. If the determination result is “unable to calculate the temperature adjustment value (abnormal),” the temperature adjustment value calculation unit 601 cancels subsequent processing and requests the device control unit 212 to issue an alarm.

[0048] The present invention is not limited to the above-described embodiments and includes various modifications. For example, the above-described embodiments have been described in detail to clearly explain the present invention, and the present invention is not necessarily limited to those including all of the described configurations. It is possible to replace part of the configuration of one embodiment with the configuration of another embodiment, and it is also possible to add the configuration of another embodiment to the configuration of one embodiment. It is possible to add, delete, or replace part of the configuration of each embodiment with other configurations.

[0049] 101: display unit, 102: input unit, 103: analysis unit, 104: light source, 105: control unit, 106: processor, 107: storage device, 108: communication interface, 201: current detection unit, 202: current adjustment unit, 203: reaction vessel, 204: reaction tank, 205: absorbance calculation unit, 206: photodetector, 207: diffraction grating, 208: photometer, 209: temperature measurement unit, 210: temperature measurement unit, 211: light intensity correction value calculation unit, 212: device control unit, 301: measurement value calculation unit, 601: temperature adjustment value calculation unit, 602: temperature adjustment unit, 1000: learning device, 1001: processor (CPU), 1002: memory, 1003: storage device, 1004 : Input interface, 1005: Output interface, 1006: Communication interface, 1008: Input device, 1009: Display device, 1010: Learning data, 1011: Experimental data, 1012: Optical system individual data, 1013: Light intensity correction value calculation formula, 1014: Learning model, 1021: Input unit, 1022: Training data generation unit, 1023: Learning model update unit, 1024: Algorithm database, 1025: Output unit, 1100: Learning model, 1101: Input layer, 1102: Intermediate layer, 1103: Output layer, 1104: Input data, 1105: Output data, 1201: Light intensity correction value correction determination unit, 1301: Temperature adjustment value correction determination unit.

Claims

1. An automatic analyzer having a light source including a first temperature measurement unit that irradiates light onto a container that contains a mixed liquid of a sample and a reagent; a photometer including a second temperature measurement unit that measures the amount of light from the light source that has passed through the mixed liquid; a light intensity correction value calculation unit that calculates a light intensity correction value based on the light source temperature measured by the first temperature measurement unit and the photometer temperature measured by the second temperature measurement unit; and an absorbance calculation unit that calculates the absorbance of the mixed liquid based on the light intensity measured by the photometer and the light intensity correction value.

2. An automatic analyzer according to claim 1, wherein the light intensity correction value calculation unit holds a first light intensity change rate which is the change in light intensity in response to a change in the light source temperature, and a second light intensity change rate which is the change in light intensity in response to a change in the photometer temperature, and calculates the light intensity correction value as the sum of a light intensity correction value based on the difference between the light source temperature measured by the first temperature measurement unit and a reference temperature and the temperature change of the light source based on the first light intensity change rate, and a light intensity correction value based on the difference between the photometer temperature measured by the second temperature measurement unit and the reference temperature and the temperature change of the photometer based on the second light intensity change rate.

3. An automatic analyzer according to claim 2, wherein the photometer comprises a diffraction grating which separates incident light into wavelengths and a photodetector which detects the separated light, and the light quantity correction value calculation unit holds the first light quantity change rate and the second light quantity change rate for each wavelength.

4. An automatic analysis device according to claim 1, further comprising a light intensity correction value correction judgment unit which evaluates and corrects the light intensity correction value calculated by the light intensity correction value calculation unit using a trained learning model, wherein the learning model is a learning model trained using at least experimental data, optical system individual data of an optical system including the light source and the photometer related to the device which acquired the experimental data, and training data generated based on a light intensity correction value calculation formula used by the light intensity correction value calculation unit to calculate the light intensity correction value.

5. An automatic analyzer having a light source including a first temperature measurement unit that irradiates light onto a container containing a mixed liquid of a sample and a reagent; a photometer including a second temperature measurement unit that measures the amount of light from the light source that has passed through the mixed liquid; a temperature adjustment value calculation unit that calculates a temperature adjustment value based on the light source temperature measured by the first temperature measurement unit and the photometer temperature measured by the second temperature measurement unit; a temperature adjustment unit that adjusts the temperature of the photometer based on the temperature adjustment value; and an absorbance calculation unit that calculates the absorbance of the mixed liquid based on the amount of light measured by the photometer.

6. An automatic analysis device according to claim 5, wherein the temperature adjustment value calculation unit holds a temperature adjustment rate that is set so that the change in light intensity corresponding to the change in the light source temperature is offset by the change in temperature of the photometer, and the temperature adjustment value is calculated based on the difference between the light source temperature measured by the first temperature measurement unit and a reference temperature, the difference between the photometer temperature measured by the second temperature measurement unit and the reference temperature, and the temperature adjustment rate.

7. An automatic analyzer according to claim 6, wherein the photometer comprises a diffraction grating that separates incident light into wavelengths and a photodetector that detects the separated light, and the temperature adjustment value calculation unit holds the temperature adjustment rate for each wavelength.

8. An automatic analysis device as claimed in claim 5, further comprising a temperature adjustment value correction judgment unit which evaluates and corrects the temperature adjustment value calculated by the temperature adjustment value calculation unit using a learned learning model, wherein the learning model is a learning model which has been learned using experimental data, optical system individual data for an optical system including the light source and the photometer relating to the device which acquired the experimental data, and training data generated based on a temperature adjustment value calculation formula which the temperature adjustment value calculation unit uses to calculate the temperature adjustment value.

Images