JP7264375B2 - Information processing device, information processing method and information processing program - Google Patents

Description

The present disclosure relates to an information processing device, an information processing method, and an information processing program.

Various methods for detecting dyslipidemia, which is one of lifestyle-related diseases, have been investigated. For example, Patent Literature 1 discloses a technique for noninvasively testing blood cholesterol levels by measuring the brightness of light that passes through a living body including blood vessels. In addition, in Patent Document 2, the scattered light of the light irradiated to the living body is received at two different points, and lipids such as CM (Chylomicron) and VLDL (Very Low-Density Lipoprotein) in the blood are detected based on the attenuation of the light intensity. Techniques for measuring the concentration of components are disclosed.

JP 2010-246578 A Japanese Patent No. 6029128

In order to appropriately evaluate dyslipidemia, it is necessary to grasp the concentration of lipid components such as bad cholesterol (LDL-C), good cholesterol (HDL-C), and triglycerides (TG). . However, the methods described in Patent Literatures 1 and 2 cannot be said to sufficiently evaluate lipid components necessary for evaluating dyslipidemia, and there is room for improvement.

The present disclosure has been made in view of the above, and aims to provide an information processing device, an information processing method, and an information processing program capable of accurately evaluating lipid components contained in a sample solution. and

In order to achieve the above object, an information processing apparatus according to an aspect of the present disclosure includes a light irradiation unit that irradiates a sample solution with light; and a teacher data set for estimating the lipid components contained in the sample solution from the light receiving data received by the light receiving unit. and a computing unit for estimating the lipid components contained in the sample solution based on the received light data and the machine learning model.

According to the information processing apparatus described above, the lipid component contained in the sample solution is estimated based on the machine learning model and the received light data relating to the intensity information of the light from the sample solution. A machine-learning model is created from a training data set for estimating the lipid components contained in the sample solution. Therefore, since the lipid component is estimated using such a machine learning model, it is possible to accurately evaluate the lipid component contained in the sample solution.

According to the present disclosure, an information processing device, an information processing method, and an information processing program capable of accurately evaluating lipid components contained in a sample solution are provided.

1 is a block diagram showing the functional configuration of an information processing apparatus according to an embodiment; FIG. It is a flow figure explaining an information processing method. It is a figure explaining the machine-learning model calculated from the data set for teachers. It is a figure which shows an example of light reception data. It is a figure explaining the process using a machine-learning model. FIG. 10 is a diagram illustrating a teacher data set used in the first analysis example; FIG. FIG. 10 is a diagram illustrating a teacher data set used in the first analysis example; FIG. FIG. 10 is a diagram showing evaluation results of estimation performance of the machine learning model created in the first analysis example; FIG. 4 is a diagram showing light reception data of a sample solution used in the first analysis example; It is a figure which shows the feature-value used by the 1st analysis example. FIG. 11 is a diagram illustrating a teacher data set used in the second analysis example; FIG. FIG. 11 is a diagram illustrating a teacher data set used in the second analysis example; FIG. FIG. 10 is a diagram showing evaluation results of the estimation performance of the machine learning model created in the second analysis example; FIG. 10 is a diagram showing light reception data of a sample solution used in the second analysis example; It is a figure which shows the feature-value used by the 2nd analysis example. It is a figure which shows the hardware constitutions of an information processing apparatus. It is a figure which shows the structure of an information processing program.

EMBODIMENT OF THE INVENTION Hereinafter, with reference to an accompanying drawing, the form for implementing this disclosure is demonstrated in detail. In the description of the drawings, the same elements are denoted by the same reference numerals, and overlapping descriptions are omitted.

[Information processing device]
FIG. 1 is a diagram illustrating the configuration of an information processing apparatus according to one embodiment. The information processing apparatus 1 according to the present embodiment is an apparatus that estimates the lipid components contained in the sample solution S by irradiating the sample solution S with light. The sample solution includes, for example, blood or a solution obtained by subjecting blood to a predetermined treatment. The sample solution S may be housed in a container or the like that is permeable to the light used for measurement (measurement light). Lipid components to be analyzed include LDL-C (Low Density Lipoprotein-Cholesterol: bad cholesterol), HDL-C (High Density Lipoprotein-Cholesterol: good cholesterol), and TG (Triglyceride: neutral fat). be done. However, it is not limited to these components. In the following embodiments, a case of analyzing the above components will be described.

As shown in FIG. 1 , the information processing apparatus 1 includes a light irradiation section 11 , a light receiving section 12 , a model holding section 13 , a calculation section 14 , an output section 15 and a control section 16 . Note that these functional units may be integrated into one device (computer), or may be distributed among a plurality of devices. For example, the light irradiation unit 11 and the light receiving unit 12 may be provided separately from the device in which other functional units are mounted.

The light irradiation unit 11 irradiates the sample solution S with the measurement light based on an instruction from the control unit 16 .

The light receiving unit 12 receives light transmitted, reflected, or scattered through the sample solution S when the measurement light from the light irradiation unit 11 is irradiated, and acquires light reception data.

The received light data acquired by the light receiving unit 12 is information related to the light intensity according to the wavelength of the light from the sample solution S irradiated with the measurement light. When analyzing the lipid components contained in the sample solution S, the light irradiation unit 11 irradiates the measurement light within the wavelength range of at least 190 nm to 270 nm. The measurement light may be light including all light in a predetermined wavelength band. Examples of wavelength bands include, but are not limited to, 800 nm to 2500 nm. Also, the measurement light may be light with a single wavelength of 1315 nm, for example. Furthermore, a combination of light with a plurality of wavelengths selected from a predetermined wavelength range may be used as the measurement light.

If the measurement light is not light of a single wavelength, the wavelength of the light emitted from the light irradiation unit 11 may be gradually changed so that the light receiving unit 12 acquires a spectrum as light reception data. As a configuration for gradually changing the wavelength of light, for example, providing a spectrometer in the light irradiation unit 11 can be considered. Moreover, the measurement light emitted from the light irradiation unit 11 may include all light of multiple wavelengths. In this case, by providing a spectroscope in the light receiving section 12, the light receiving section 12 may acquire a spectral spectrum as light reception data.

The model holding unit 13 stores a machine learning model for calculating the concentration of lipid components (LDL-C, HDL-C, TG) in the sample solution S based on the received light data acquired by the light receiving unit 12. there is This model is created by machine learning. Details of the model (machine learning model) will be described later.

The calculation unit 14 calculates the lipid components (LDL-C, HDL-C, TG) in the sample solution S based on the received light data acquired by the light receiving unit 12 and the machine learning model held by the model holding unit 13. It has a function to calculate the concentration of A method for calculating the concentration of the lipid component will be described later.

The output unit 15 has a function of outputting the calculation result by the calculation unit 14 .

The control unit 16 has a function of controlling the operation of each unit described above.

[Information processing method]
Next, an information processing method by the information processing apparatus 1 will be described with reference to FIG. First, before analyzing the sample solution S, a training data set is used to create a machine learning model to be used for analysis (S01). The training data set includes light reception data obtained using the information processing device 1 for training samples with known concentrations of lipid components (HDL-C, LDL-C, TG) to be analyzed, and the concentration of lipid components (HDL-C, LDL-C, TG). A plurality of teacher samples having different concentrations of lipid components are prepared. Although the method for measuring the concentration of the lipid component in each of the plurality of teacher samples is not particularly limited, for example, a chemical analysis method such as electrophoresis can be used. Further, the received light data for each of the plurality of teaching samples is preferably data for light with wavelengths in the widest possible band. A “broad band” can be, for example, 190 nm to 2700 nm.

Using the training data set obtained as described above, the concentration of each lipid component (which may be any of them) is set as the objective variable, and the light intensity information (absorbance, (reflectance, scattering coefficient, etc.) is used as a feature quantity, and machine learning is performed. As a result, a feature amount that contributes to the estimation of the concentration of the lipid component in the sample solution S is selected. After that, machine learning is used to determine what kind of model should be constructed in order to estimate the concentration of the lipid component contained in the sample solution S using the feature values selected in the above procedure.

FIG. 3(a) exemplifies information included in the teacher data set. In FIG. 3(a), light intensity information at specific wavelengths for three types (A to C) of training samples are shown as x ₁ to x _d , respectively, and each lipid component (HDL-C, LDL-C, TG ) are shown as y ₁ to y ₃ . The training data set contains such information for each sample that constitutes the data set. Here, using machine learning, the concentrations of lipid components (y ₁ to y ₃ ) are used as target variables, and light intensity information highly related to these is selected from the total light intensity information (x ₁ to x _d ). do. Machine learning is then used to create a relational expression for estimating the concentration of lipid components from the selected light intensity information. The created relational expression becomes a machine learning model. FIG. 3B is a description example of a machine learning model. In the example shown in FIG. 3(b), the lipid component concentrations (y ₁ to y ₃ ) are described as functions using six light intensity information (x _a to x _f ). As described above, before performing analysis related to the sample solution S, the information processing apparatus 1 creates a machine learning model (S01). Note that the creation of the machine learning model can be performed in the calculation unit 14 based on instructions from the control unit 16 . Also, the machine learning model created by the above procedure is held in the model holding unit 13 .

Note that the machine learning used in the information processing apparatus 1 of the present embodiment refers to information (concentration, class, etc.) related to the lipid component to be calculated from the teacher data in the process for creating the relational expression as described above. ) is used as the objective variable, the feature amount for determining the objective variable is selected (if necessary), and then the relational expression for determining the objective variable is calculated. Machine learning techniques used in such processes include SVR (Support Vector Regression), Ridge Regression, Lasso Regression, Elastic Net, Neural Network, Deep Learning, and the like. Any one of these techniques may be used, or a combination thereof may be used. Also, statistical processing methods such as principal component analysis may be used in combination.

Next, in the information processing apparatus 1, the light-receiving unit 12 acquires light reception data by irradiating the sample solution S with the measurement light from the light irradiation unit 11 (S02: acquisition step). After placing the sample solution S at a predetermined position, the light irradiation unit 11 irradiates the sample solution S with measurement light based on instructions from the control unit 16, and the light receiving unit 12 receives transmitted, reflected, or scattered light. By doing so, the light receiving unit 12 can acquire received light data. FIG. 4 shows an example of received light data. The example shown in FIG. 4 is an example of a spectral spectrum obtained by irradiating the sample solution S with light in the wavelength range of 190 nm to 2700 nm as measurement light. Thus, the received light data includes light intensity information associated with each wavelength. The received light data acquired by the light receiving unit 12 is sent to the calculation unit 14 .

Next, the calculation unit 14 of the information processing device 1 calculates the concentration of the lipid component based on the received light data acquired by the light receiving unit 12 and the machine learning model held in the model holding unit 13 (S03: calculation step).

The calculation of the lipid component concentration by the calculation unit 14 (S03) will be further described with reference to FIG. In the machine learning model, light intensity information to be used as a feature amount is determined. Therefore, the calculation unit 14 first selects light intensity information to be used as a feature amount from the received light data (S11). This step corresponds to the process of extracting the light intensity information corresponding to the six light intensity information (x _a to x _f ) shown in FIG. 3(b). Then, the selected light intensity information is substituted as a feature quantity into the machine learning model (S12). This step corresponds to a process of substituting the extracted six pieces of light intensity information (x _a to x _f ) into the relational expression shown in FIG. 3(b). After that, the lipid component concentration is calculated by calculating the relational expression in which the feature amount is substituted (S13). Through such a procedure, the lipid component concentration can be calculated using the light intensity information included in the received light data.

Returning to FIG. 2, the calculation result of the lipid component concentration obtained by the above procedures (S01 to S03) is output from the output unit 15, for example, according to an instruction from the control unit 16 (S04). Thereby, it becomes possible to notify the operator of the apparatus or the like of information related to the concentration of the lipid component calculated based on the received light data.
In the above procedure, a case of calculating the "concentration" of the lipid component has been described, but the estimation relating to the lipid component is not limited to the "concentration". For example, instead of calculating a numerical value like concentration, it may be required to specify a classification (class) indicating how much lipid component is contained for each sample solution S. In that case, by using the classification (class) as the objective variable, it is possible to perform the same processing as in the case of estimating the concentration.

[First analysis example]
Next, a first analysis example by the information processing apparatus 1 will be described with reference to FIGS. 6 to 10. FIG. In the first analysis example, a case where the information processing apparatus 1 is used to calculate the concentration of LDL-C contained in the sample solution S will be described.

Assuming that the sample solution S is a solution obtained by dissolving hemoglobin, LDL-C, HDL-C, and TG in a buffer solution, first, a machine learning model was created (Step S01 in FIG. 2).

In creating the machine learning model, 144 kinds of teacher samples (Samples 1 to 144) were prepared by dissolving combinations of different concentrations of hemoglobin, LDL-C, HDL-C, and TG in a buffer solution. Each teacher sample was irradiated with light having a wavelength of 190 nm to 2,700 nm from the light irradiation unit 11 of the information processing apparatus 1 , and the light transmitted through the sample was received by the light receiving unit 12 . FIG. 6 shows an example of received light data for 144 samples. As shown in FIG. 6, absorbance data was obtained for each sample between 190 nm and 2700 nm in increments of 1 nm. Also, the concentrations of LDL-C, HDL-C and TG are known for each of the 144 samples. An example of each lipid component for 144 samples is shown in FIG. The data shown in FIGS. 6 and 7 are the teacher data set.

In the first analysis example, the number of teaching samples is 144, but the number of samples may be more or less than 144.

After that, the calculation unit 14 of the information processing apparatus 1 performs machine learning using the information related to the light absorption characteristics shown in FIG. 6 as a feature amount and the LDL-C concentration shown in FIG. 7 as an objective variable. In this way, a feature amount useful for estimating the LDL-C concentration is selected. The feature amount selected here refers to light intensity information that is highly related to the LDL-C concentration. As an example, the light intensity information of each wavelength shown in Table 1 below can be used as the feature quantity.

Note that there are a method of limiting useful data items by means of Lasso regression, Ridge regression, Elastic Net, etc., and a method of limiting data items based on the contribution rate of principal component analysis, etc., for extracting feature amounts. However, it is not limited to these.

After that, a machine learning model is created using the selected features. When creating a machine learning model, the data contained in the training dataset may be utilized to assess accuracy. Alternatively, a separate evaluation data set may be prepared separately from the teacher data set, and accuracy may be evaluated using this data set.

Machine learning models include, but are not limited to, SVM (Support Vector Machine), logistic regression models, and the like. Also, evaluation methods include coefficient of determination, correlation coefficient, etc., but are not limited to these. FIG. 8 shows an example of evaluating the estimation performance of the created machine learning model. In FIG. 8, the horizontal axis is the correct value of the LDL-C concentration, and the vertical axis is the estimated value of the LDL-C concentration estimated using the machine learning model, and each sample included in the data set is plotted. . As shown in FIG. 8, the results of estimation by the machine learning model show a tendency similar to that of the correct values. The created machine learning model is held in the model holding unit 13 .

Next, a sample solution S whose concentration is to be estimated is prepared, the sample solution S is irradiated with measurement light in a wavelength range of 190 nm to 2700 nm from the light irradiation unit 11, and the transmitted light transmitted through the sample solution S is received. Light is received at the portion 12 . The measurement light may include light of wavelengths included in the feature amounts shown in Table 1. Through the above-described procedure, light receiving data relating to the sample solution S is obtained in the light receiving unit 12 . FIG. 9 shows an example of received light data.

Next, the calculation unit 14 of the information processing apparatus 1 extracts a feature amount used for estimating the concentration of LDL-C from the received light data shown in FIG. FIG. 10 shows the extracted results. By substituting these feature values into the previously created machine learning model, the concentration of LDL-C in the sample solution S can be estimated. In the case of the sample solution S shown in FIGS. 9 and 10, for example, the concentration of LDL-C can be estimated to be 1.69752 [mg/ml].

[Second analysis example]
Next, a second example of analysis by the information processing apparatus 1 will be described with reference to FIGS. 11 to 15. FIG. In the first analysis example, the information processing apparatus 1 is used to calculate the concentration of LDL-C contained in the sample solution S. In the second analysis example, the concentration of the lipid component itself is calculated. It is different from the first analysis example in that the concentration of the lipid component tends to be classified instead of the first analysis example. Specifically, a case of estimating whether the concentration of HDL-C contained in the sample solution S is determined to be dyslipidemia, probable dyslipidemia, or health will be described. In addition, in this analysis example, the judgment criteria shall be based on Table 2 below.

In this analysis example, it is determined to which class the sample solution S belongs based on the above criteria. However, the number of classes is not limited to three, and the threshold values are not limited to the values shown in Table 2 either.

Assuming that the sample solution S is a solution obtained by dissolving hemoglobin, LDL-C, HDL-C, and TG in a buffer solution, first, a machine learning model was created (Step S01 in FIG. 2).

In creating the machine learning model, 144 kinds of teacher samples (Samples 1 to 144) were prepared by dissolving combinations of different concentrations of hemoglobin, LDL-C, HDL-C, and TG in a buffer solution. Each teacher sample was irradiated with light having a wavelength of 190 nm to 2,700 nm from the light irradiation unit 11 of the information processing apparatus 1 , and the light transmitted through the sample was received by the light receiving unit 12 . FIG. 11 shows an example of received light data for 144 samples. As shown in FIG. 11, absorbance data was obtained for each sample between 190 nm and 2700 nm in increments of 1 nm. Also, since the concentrations of LDL-C, HDL-C, and TG are known for each of the 144 samples, which of the three classes the concentration of each lipid component falls into based on Table 2 above. can be determined. FIG. 12 shows an example of class classification for each lipid component of 144 samples. In the case of the second analysis example, the data shown in FIGS. 11 and 12 are the teacher data set.

Although the number of teaching samples is 144 in the second analysis example as well as in the first analysis example, the number of samples may be more or less than 144.

After that, the calculation unit 14 of the information processing apparatus 1 performs machine learning using the information related to the light absorption characteristics shown in FIG. 11 as a feature amount and the HDL-C class classification shown in FIG. 12 as an objective variable. This selects a feature quantity useful for estimating class classification based on the concentration of HDL-C. As an example, light of each wavelength shown in Table 3 below can be used as the feature amount.

Note that there are a method of limiting useful data items by means of Lasso regression, Ridge regression, Elastic Net, etc., and a method of limiting data items based on the contribution rate of principal component analysis, etc., for extracting feature amounts. However, it is not limited to these.

After that, a machine learning model is created using the selected features. When creating a machine learning model, the data contained in the training dataset may be utilized to assess accuracy. Alternatively, a separate evaluation data set may be prepared separately from the teacher data set, and accuracy may be evaluated using this data set.

Machine learning models include, but are not limited to, k-nearest neighbor, logistic regression, and the like. Also, evaluation methods include accuracy, precision, and the like, but are not limited to these. FIG. 13 shows the result of analyzing the estimation performance of the created machine learning model. FIG. 13 shows a table of classes estimated using a machine learning model and a confusion matrix of actual classes of each data. The created machine learning model is held in the model holding unit 13 .

Next, a sample solution S whose concentration is to be estimated is prepared, the sample solution S is irradiated with measurement light in a wavelength range of 190 nm to 2700 nm from the light irradiation unit 11, and the transmitted light transmitted through the sample solution S is received. Light is received at the portion 12 . The measurement light may include light of wavelengths included in the feature values shown in Table 3. Through the above-described procedure, light receiving data relating to the sample solution S is obtained in the light receiving unit 12 . FIG. 14 shows an example of received light data.

Next, the calculation unit 14 of the information processing apparatus 1 extracts a feature amount used for class estimation based on the concentration of HDL-C from the received light data shown in FIG. FIG. 15 shows the extracted results. By substituting these feature values into the previously created machine learning model, the class associated with the HDL-C concentration of the sample solution S can be estimated. In the case of the sample solution S shown in FIGS. 14 and 15, for example, it can be estimated that the concentration of HDL-C corresponds to dyslipidemia.

[Hardware configuration]
Each function used in the description of the above embodiments is implemented by any combination of hardware and/or software. Moreover, means for realizing each function is not particularly limited. That is, each function may be implemented by one device physically and/or logically coupled, or may be implemented by two or more physically and/or logically separated devices directly and/or indirectly. (eg, wired and/or wirelessly) and implemented by these multiple devices.

For example, the information processing device 1 described above may function as a computer. FIG. 16 is a diagram showing an example of the hardware configuration of the information processing device 1. As shown in FIG. The information processing device 1 may be physically configured as a computer device including a processor 1001, a memory 1002, a storage 1003, a communication device 1004, an input device 1005, an output device 1006, a bus 1007, and the like.

Note that in the following description, the term "apparatus" can be read as a circuit, device, unit, or the like. The hardware configuration of the information processing device 1 may be configured to include one or a plurality of each device shown in the figure, or may be configured without including some of the devices.

Each function in the information processing apparatus 1 is performed by causing the processor 1001 to perform calculations by loading predetermined software (programs) on hardware such as the processor 1001 and the memory 1002, and performing communication by the communication device 1004, the memory 1002 and storage. It is realized by controlling reading and/or writing of data in 1003 .

The processor 1001, for example, operates an operating system to control the entire computer. The processor 1001 may be configured with a central processing unit (CPU) including an interface with peripheral devices, a control device, an arithmetic device, registers, and the like. For example, various processes of the information processing device 1 may be realized by the processor 1001 .

The processor 1001 also reads programs (program codes), software modules, and data from the storage 1003 and/or the communication device 1004 to the memory 1002, and executes various processes according to them. As the program, a program that causes a computer to execute at least part of the operations described in the above embodiments is used. For example, the function of executing various processes of the information processing apparatus 1 may be implemented by a control program stored in the memory 1002 and operated by the processor 1001, and other functional blocks may be similarly implemented. Although it has been described that the above-described various processes are executed by one processor 1001, they may be executed by two or more processors 1001 simultaneously or sequentially. Processor 1001 may be implemented with one or more chips. Note that the program may be transmitted from a network via an electric communication line.

The memory 1002 is a computer-readable recording medium, and is composed of at least one of, for example, ROM (Read Only Memory), EPROM (Erasable Programmable ROM), EEPROM (Electrically Erasable Programmable ROM), and RAM (Random Access Memory). may be The memory 1002 may also be called a register, cache, main memory (main storage device), or the like. The memory 1002 can store executable programs (program codes), software modules, and the like for executing various processes of the information processing apparatus 1 .

The storage 1003 is a computer-readable recording medium, for example, an optical disk such as a CD-ROM (Compact Disc ROM), a hard disk drive, a flexible disk, a magneto-optical disk (for example, a compact disk, a digital versatile disk, a Blu-ray disk), smart card, flash memory (eg, card, stick, key drive), floppy disk, magnetic strip, and/or the like. Storage 1003 may also be called an auxiliary storage device. The storage medium described above may be, for example, a database, server, or other suitable medium including memory 1002 and/or storage 1003 .

The communication device 1004 is hardware (transmitting/receiving device) for communicating between computers via a wired and/or wireless network, and is also called a network device, network controller, network card, communication module, or the like. For example, part of the various processes of the information processing device 1 may be realized by the communication device 1004 .

The input device 1005 is an input device (for example, keyboard, mouse, microphone, switch, button, sensor, etc.) that receives input from the outside. The output device 1006 is an output device (for example, display, speaker, LED lamp, etc.) that outputs to the outside. Note that the input device 1005 and the output device 1006 may be integrated (for example, a touch panel).

Devices such as the processor 1001 and the memory 1002 are connected by a bus 1007 for communicating information. The bus 1007 may be composed of a single bus, or may be composed of different buses between devices.

The information processing device 1 includes hardware such as a microprocessor, a digital signal processor (DSP), an ASIC (Application Specific Integrated Circuit), a PLD (Programmable Logic Device), and an FPGA (Field Programmable Gate Array). , and part or all of each functional block may be implemented by the hardware. For example, processor 1001 may be implemented with at least one of these hardware.

[Information processing program]
Next, an information processing program for causing a computer to function as the information processing apparatus 1 of this embodiment will be described. FIG. 17 is a diagram showing the configuration of the information processing program P1.

The information processing program P1 includes a main module m10, a light irradiation module m11, a light receiving module m12, a model holding module m13, an arithmetic module m14, an output module m15, and a control module m16 that collectively control the above processing in the information processing apparatus 1. configured with. Each module m11 to m16 realizes each function for the light irradiation unit 11, the light receiving unit 12, the model holding unit 13, the calculation unit 14, the output unit 15, and the control unit 16 in the information processing apparatus 1. . The information processing program P1 may be transmitted via a transmission medium such as a communication line, or may be stored in a recording medium M1 as shown in FIG.

[Action]
The information processing apparatus 1 of the present embodiment described above includes a light irradiation unit that irradiates light onto a sample solution, and a light irradiation unit that receives the light from the sample solution irradiated with the light from the light irradiation unit. A light-receiving unit that acquires light-receiving data related to intensity information, and a machine learning model created from a training data set for estimating a lipid component contained in the sample solution from the light-receiving data received by the light-receiving unit. and a calculation unit for estimating lipid components contained in the sample solution based on the received light data and the machine learning model.

Further, the information processing method according to the present embodiment includes an acquisition step of acquiring light reception data related to light intensity information by receiving light from the sample solution by irradiating the sample solution with light; a computing step of estimating the lipid components contained in the sample solution based on the data and a machine learning model created from a teacher data set for estimating the lipid components contained in the sample solution; have.

Further, the information processing program P1 according to the present embodiment is an information processing program for causing a computer to function as an information processing device, and irradiates a sample solution with light to receive light from the sample solution. and a teacher data set for estimating a lipid component contained in the sample solution from the received light data received by the light receiving function. A model holding function for holding a machine learning model and an arithmetic function for estimating a lipid component contained in the sample solution based on the machine learning model are realized.

According to the information processing device, information processing method, and information processing program described above, the lipid component contained in the sample solution can be estimated based on the machine learning model and the received light data related to the light intensity information from the sample solution. done. A machine-learning model is created from a training data set for estimating the lipid components contained in the sample solution. Therefore, since the lipid component is estimated using such a machine learning model, it is possible to accurately evaluate the lipid component contained in the sample solution.

In recent years, with the aim of extending healthy life expectancy, there has been growing interest in preventive techniques for lifestyle-related diseases. There are many lifestyle-related diseases that should be prevented, and among them, dyslipidemia is attracting attention. Dyslipidemia refers to a state in which the concentration of multiple lipid components in the blood is higher or lower than the threshold. When dyslipidemia develops, plaque forms on the surface of the blood vessel wall, making the blood vessels more likely to clog. can cause arteriosclerosis. However, although dyslipidemia is highly likely to lead to heart disease and cerebrovascular disease, it is characterized by the lack of subjective symptoms in those who develop it. Therefore, in order to prevent lifestyle-related diseases, it is desirable to evaluate whether or not a person suffers from dyslipidemia by some method.

The lipid components used in determining whether or not there is dyslipidemia are LDL-C, HDL-C, and TG described in the above embodiment, and different threshold values are set for each. Therefore, it is important for the prevention of dyslipidemia to know how much each lipid component is contained in the blood. However, a method for evaluating this method simply and accurately in a non-invasive manner has not been studied, and improvements have been desired.

On the other hand, according to the information processing apparatus, information processing method, and information processing program described above, based on the machine learning model and the received light data related to the light intensity information from the sample solution, Inferences can be made regarding the lipid content. That is, it is possible to accurately evaluate the lipid component by a non-invasive method. In addition, since the estimation of the lipid component described in the above embodiment is an evaluation using received light data obtained by an optical method, the operation is also simple.

In another embodiment, the calculation unit may estimate the concentration of the lipid component as the estimation of the lipid component.

By estimating the lipid component concentration using the machine learning model in the calculation unit, it becomes possible to accurately estimate the concentration of the lipid component contained in the sample solution.

In another embodiment, the estimation of the lipid component may be a mode of estimating a class related to the concentration of the lipid component.

By estimating a class related to lipid component concentration using a machine learning model in the computing unit, it is possible to more accurately estimate the tendency of lipid component concentration.

In another embodiment, the light irradiated from the light irradiating section may be light having a wavelength within the range of 190 nm to 2700 nm.

Estimation related to lipid components can be performed with high accuracy by using light in the above wavelength range. Therefore, by irradiating light in the above wavelength range from the light irradiating part, it is possible to accurately estimate the lipid component.

In another embodiment, the computing unit extracts a feature amount to be applied to the machine learning model from the received light data, and substitutes the feature amount into the machine learning model, thereby estimating the lipid component. can be

With the above configuration, it is possible to accurately estimate the lipid component in the calculation unit.

As described above, according to the present disclosure, it is possible to examine and confirm the concentration, class, etc. of lipid components that can be indicators of dyslipidemia by a noninvasive and simple method. When the information processing device, information processing method, and information processing program described above are applied to prevent dyslipidemia, by periodically performing the estimation related to the lipid component, lipid It becomes possible to notice abnormalities in the concentration of lipid components, which are indicators of abnormalities. That is, it is possible to prevent dyslipidemia before it develops, and prevent lifestyle-related diseases.

[others]
Although the present embodiments have been described in detail above, it will be apparent to those skilled in the art that the present embodiments are not limited to the embodiments described herein. This embodiment can be implemented as modifications and changes without departing from the spirit and scope of the present invention defined by the description of the claims. Therefore, the description in this specification is for the purpose of illustration and explanation, and does not have any restrictive meaning with respect to the present embodiment.

Notification of information is not limited to the aspects/embodiments described herein and may be done in other ways. For example, notification of information includes physical layer signaling (e.g., DCI (Downlink Control Information), UCI (Uplink Control Information)), higher layer signaling (e.g., RRC (Radio Resource Control) signaling, MAC (Medium Access Control) signaling, It may be implemented by broadcast information (MIB (Master Information Block), SIB (System Information Block)), other signals, or a combination thereof. RRC signaling may also be called an RRC message, and may be, for example, an RRC connection setup message, an RRC connection reconfiguration message, or the like.

The procedures, sequences, flow charts, etc. of each aspect/embodiment described herein may be interchanged so long as there is no inconsistency. For example, the methods described herein present elements of the various steps in a sample order, and are not limited to the specific order presented.

Input/output information and the like may be stored in a specific location (for example, memory), or may be managed in a management table. Input/output information and the like may be overwritten, updated, or appended. The output information and the like may be deleted. The entered information and the like may be transmitted to another device.

The determination may be made by a value represented by one bit (0 or 1), by a true/false value (Boolean: true or false), or by numerical comparison (for example, a predetermined value).

Each aspect/embodiment described herein may be used alone, in combination, or switched between implementations. In addition, the notification of predetermined information (for example, notification of “being X”) is not limited to being performed explicitly, but may be performed implicitly (for example, not notifying the predetermined information). good too.

Software, whether referred to as software, firmware, middleware, microcode, hardware description language or otherwise, includes instructions, instruction sets, code, code segments, program code, programs, subprograms, and software modules. , applications, software applications, software packages, routines, subroutines, objects, executables, threads of execution, procedures, functions, and the like.

Software, instructions, etc. may also be sent and received over a transmission medium. For example, the software can be used to access websites, servers, or other When transmitted from a remote source, these wired and/or wireless technologies are included within the definition of transmission media.

Information, signals, etc. described herein may be represented using any of a variety of different technologies. For example, data, instructions, commands, information, signals, bits, symbols, chips, etc. that may be referred to throughout the above description may refer to voltages, currents, electromagnetic waves, magnetic fields or magnetic particles, light fields or photons, or any of these. may be represented by a combination of

The terms explained in this specification and/or terms necessary for understanding this specification may be replaced with terms having the same or similar meanings.

As used herein, the terms "system" and "network" are used interchangeably.

In addition, the information, parameters, etc. described in this specification may be represented by absolute values, may be represented by relative values from a predetermined value, or may be represented by corresponding other information. .

The names used for the parameters described above are not limiting in any way. Further, the formulas, etc. using these parameters may differ from those explicitly disclosed herein.

As used herein, the terms "judging" and "determining" may encompass a wide variety of actions. "Judgement", "determining" are, for example, judging, calculating, computing, processing, deriving, investigating, looking up (e.g., table , searching in a database or other data structure), regarding ascertaining as "determining" or "determining". Also, "judgment" and "determination" are used for receiving (e.g., receiving information), transmitting (e.g., transmitting information), input, output, access (accessing) (for example, accessing data in memory) may include deeming that a "judgment" or "decision" has been made. In addition, "judgment" and "decision" are considered to be "judgment" and "decision" by resolving, selecting, choosing, establishing, comparing, etc. can contain. In other words, "judgment" and "decision" can include considering that some action is "judgment" and "decision".

The terms "connected", "coupled", or any variation thereof, mean any direct or indirect connection or coupling between two or more elements, It can include the presence of one or more intermediate elements between two elements being "connected" or "coupled." Couplings or connections between elements may be physical, logical, or a combination thereof. As used herein, two elements are referred to by the use of one or more wires, cables and/or printed electrical connections and, as some non-limiting and non-exhaustive examples, radio frequency They can be considered to be “connected” or “coupled” to each other through the use of electromagnetic energy, such as electromagnetic energy having wavelengths in the, microwave, and optical (both visible and invisible) regions.

As used herein, the phrase "based on" does not mean "based only on," unless expressly specified otherwise. In other words, the phrase "based on" means both "based only on" and "based at least on."

Wherever "include," "including," and variations thereof are used in the specification or claims, these terms are synonymous with the term "comprising." are intended to be inclusive. Furthermore, the term "or" as used in this specification or the claims is not intended to be an exclusive OR.

In this specification, plural devices shall be included unless the context or technicality clearly dictates that there is only one. Throughout this disclosure, the plural shall be included unless the context clearly indicates the singular.

DESCRIPTION OF SYMBOLS 1... Information processing apparatus, 11... Light irradiation part, 12... Light-receiving part, 13... Model holding part, 14... Calculation part, 15... Output part, 16... Control part, P1... Information processing program, S... Sample solution.

Claims (7)

a light irradiation unit that irradiates the sample solution with light;
a light-receiving unit that receives the light from the sample solution irradiated with the light from the light-irradiating unit and acquires received light data related to light intensity information;
a model holding unit that holds a machine learning model created from a teacher data set for estimating a lipid component contained in the sample solution from the received light data received by the light receiving unit;
a calculation unit that estimates a lipid component contained in the sample solution based on the received light data and the machine learning model;
has
The calculation unit estimates LDL-C, which is the lipid component, from light intensity information at each wavelength of 190 nm, 191 nm, 194 nm, 284 nm, 405 nm, 579 nm, 1314 nm, 1315 nm, 1316 nm, 1458 nm and 1888 nm, or An information processing device for estimating HDL-C, which is the lipid component, from light intensity information at wavelengths of 190 nm, 191 nm, 193 nm, 285 nm, 310 nm, 370 nm, 377 nm, 543 nm, 594 nm, 1445 nm, 1797 nm and 1877 nm. The information processing device according to claim 1, wherein the calculation unit estimates the concentration of the lipid component as the estimation of the lipid component. The information processing device according to claim 1, wherein the calculation unit estimates a class related to the concentration of the lipid component as the estimation related to the lipid component. 4. The information processing apparatus according to claim 1, wherein the light emitted from the light emitting unit has a wavelength within a range of 190 nm to 2700 nm. The computing unit extracts a feature amount to be applied to the machine learning model from the received light data, and substitutes the feature amount into the machine learning model, thereby estimating the lipid component. The information processing device according to any one of the items. an acquiring step of irradiating a sample solution with light to receive light from the sample solution and acquiring received light data related to light intensity information;
a computing step of estimating a lipid component contained in the sample solution based on the received light data and a machine learning model created from a teacher data set for estimating the lipid component contained in the sample solution; ,
has
In the calculation step, from the light intensity information of each wavelength of 190 nm, 191 nm, 194 nm, 284 nm, 405 nm, 579 nm, 1314 nm, 1315 nm, 1316 nm, 1458 nm and 1888 nm, estimate LDL-C, which is the lipid component, or An information processing method for estimating HDL-C, which is the lipid component, from light intensity information at wavelengths of 190 nm, 191 nm, 193 nm, 285 nm, 310 nm, 370 nm, 377 nm, 543 nm, 594 nm, 1445 nm, 1797 nm and 1877 nm. An information processing program for causing a computer to function as an information processing device,
a light receiving function of irradiating a sample solution with light to receive the light from the sample solution and acquiring light reception data related to light intensity information;
a model holding function for holding a machine learning model created from a training data set for estimating a lipid component contained in the sample solution from the received light data received by the light receiving function;
A computing function for estimating the lipid components contained in the sample solution based on the machine learning model;
to realize
The arithmetic function estimates LDL-C, which is the lipid component, from light intensity information at each wavelength of 190 nm, 191 nm, 194 nm, 284 nm, 405 nm, 579 nm, 1314 nm, 1315 nm, 1316 nm, 1458 nm and 1888 nm, or An information processing program for estimating HDL-C, which is the lipid component, from light intensity information at wavelengths of 190 nm, 191 nm, 193 nm, 285 nm, 310 nm, 370 nm, 377 nm, 543 nm, 594 nm, 1445 nm, 1797 nm and 1877 nm.

Images