JP7574660B2 - Information processing device, program and system - Google Patents

Description

The present invention relates to an information processing device, a program, and a system.

Technology has been proposed for examining the health condition of a subject. For example, Patent Document 1 describes a device that displays a warning if an abnormality is detected based on a signal output by a gas sensor attached to a cloth T-shaped belt. Patent Document 2 describes a system that accumulates and records measurement data, including detection data for sulfur-containing gas detected by a sulfur-containing gas sensor included in a subject-side device installed in a toilet room, in a database together with the date and time of defecation, and analyzes the subject's physical condition based on the tendency of the accumulated and recorded measurement data to change over time.

Japanese Patent Application Publication No. 9-43182 JP 2016-145808 A

However, while the techniques described in Patent Documents 1 and 2 can monitor the subject's health condition based on the output signal of the gas sensor, there is room for improvement in terms of improving the accuracy of determining the health condition.

One aspect of the present invention aims to provide a technology that can estimate the health condition of a subject with greater accuracy.

In order to solve the above problems, an information processing device according to aspect 1 of the present invention includes a controller, a memory, and a communication unit, and the communication unit receives the detection value of the sensor transmitted by a sensor module that includes a sensor that detects information related to excrement and a transmission unit that transmits the detection value of the sensor and is attached to a subject when used, the memory stores a trained model that derives output data indicating the subject's health condition from detection data that represents a time series of the detection value of the sensor, and the controller executes an estimation process that estimates the health condition using the detection data and the trained model.

According to the above configuration, the information processing device estimates the subject's health condition by inputting detection data representing a time series of detection values of a sensor that detects information about the subject's excrement into a trained model. This makes it possible to estimate the subject's health condition more accurately than when the trained model is not used.

In the information processing device according to aspect 2 of the present invention, in the aspect 1, the trained model may derive output data indicating the nutritional state of the subject from the detection data, and in the estimation process, the controller may estimate the nutritional state using the detection data and the trained model.

According to the above configuration, the information processing device estimates the nutritional state of the subject by inputting detection data representing a time series of detection values of a sensor that detects information about the subject's excrement into a trained model. This makes it possible to estimate the nutritional state of the subject more accurately than when the trained model is not used.

In the information processing device according to aspect 3 of the present invention, in the aspects 1 or 2, the trained model may be constructed by supervised machine learning using a training dataset that is a set of combinations of detection data representing a time series of detection values of the sensor and a class representing the health state.

According to the above configuration, the information processing device estimates the subject's health condition by inputting detection data representing a time series of detection values of a sensor that detects information about the subject's excrement into a trained model constructed by supervised machine learning. This makes it possible to estimate the subject's health condition with greater accuracy than when a trained model is not used.

In the information processing device according to aspect 4 of the present invention, in any one of aspects 1 to 3, the communication unit receives dietary data representing the contents of meals consumed by the subject, the trained model derives the output data from input data including the detection data and the dietary data, and the estimation process may be a process of estimating the health condition using the detection data, the dietary data, and the trained model.

According to the above configuration, the information processing device estimates the subject's health condition by inputting dietary data representing the subject's dietary content into the trained model in addition to the detection data. This allows the subject's health condition to be estimated more accurately than if dietary data is not used.

In the information processing device according to aspect 5 of the present invention, in any one of aspects 1 to 4, the controller may further execute a generation process for generating a trained model by machine learning the correlation between the detection data and a class that represents the subject's health condition, using detection data that represents a time series of the detection values of the sensor and a class that represents the subject's health condition.

According to the above configuration, the information processing device estimates the health state using a trained model generated using detection data indicating a time series of sensor detection values and a class indicating the subject's health state. This makes it possible to estimate the subject's health state more accurately than when this trained model is not used.

The information processing device according to aspect 6 of the present invention includes a controller, a memory, and a communication unit, and the communication unit receives the detection value of the sensor transmitted by a sensor module that includes a sensor for detecting information related to excrement and a transmission unit that transmits the detection value of the sensor, the sensor module being worn by a subject when used, and the controller executes a generation process that uses detection data representing a time series of the detection value of the sensor and a class representing the health condition of the subject to generate a trained model that uses machine learning to learn the correlation between the detection data and the class.

According to the above configuration, a trained model is used that is generated using detection data indicating a time series of sensor detection values and a class indicating the subject's health condition, making it possible to estimate the subject's health condition more accurately than if this trained model is not used.

In the information processing device according to aspect 7 of the present invention, in any one of aspects 1 to 6, the controller may include at least one processor that executes each of the processes according to a predetermined program, and at least one memory that stores the program.

The above configuration allows for more accurate estimation of the subject's health condition compared to when a trained model is not used.

The information processing device according to aspect 8 of the present invention is a program for controlling the information processing device according to any one of aspects 1 to 7, and is characterized in that it causes the controller to execute each of the processes.

The scope of the present invention also includes a program according to aspect 8 and a computer-readable recording medium on which the program is recorded.

The system according to aspect 9 of the present invention comprises a sensor module worn by a subject and an information processing device that estimates the health condition of the subject, the sensor module comprises a sensor that detects information related to excrement and a transmission unit that transmits the detection value of the sensor, the information processing device comprises a controller, a memory, and a communication unit, the communication unit receives the detection value of the sensor transmitted by the sensor module, the memory stores a trained model that derives output data indicating the health condition of the subject from detection data that represents a time series of the detection value of the sensor, and the controller executes an estimation process that estimates the health condition using the detection data and the trained model.

According to the above configuration, the information processing device estimates the subject's health condition by inputting detection data representing a time series of detection values of a sensor that detects information about the subject's excrement into a trained model. This makes it possible to estimate the subject's health condition more accurately than when the trained model is not used.

According to one aspect of the present invention, it is possible to estimate the subject's health condition more accurately.

1 is a diagram showing an overview of a health management system according to a first embodiment of the present invention; 1 is a diagram illustrating a state in which a sensor module according to a first embodiment of the present invention is mounted; 1 is a block diagram illustrating a functional configuration of a sensor module according to a first embodiment of the present invention; 1 is a diagram illustrating an internal configuration of a sensor module according to a first embodiment of the present invention; 1 is a block diagram illustrating a configuration of an information processing device according to a first embodiment of the present invention. FIG. 1 is a diagram illustrating an example of a trained model according to a first embodiment of the present invention. FIG. 1 is a flowchart showing the flow of a health condition estimation method according to a first embodiment of the present invention. FIG. 4 is a diagram showing an example of a screen displayed on the information processing device according to the first embodiment of the present invention. FIG. 11 is a flow chart showing the flow of a method for generating a trained model according to a second embodiment of the present invention. FIG. 11 is a diagram showing an overview of a health management system according to a third embodiment of the present invention. FIG. 11 is a sequence diagram showing the flow of a health condition estimation method according to a third embodiment of the present invention. FIG. 11 is a diagram showing an example of a screen displayed on an information processing device according to a third embodiment of the present invention. FIG. 13 is a diagram showing an overview of a health management system according to a fourth embodiment of the present invention.

[Embodiment 1]
Hereinafter, one embodiment of the present invention will be described in detail.

<Health Management System Configuration>
Fig. 1 is a diagram showing a schematic configuration of a health management system 1 according to this embodiment. The health management system 1 is a system for managing the health of a subject. In particular, the health management system 1 provides a service for estimating the health state of the subject. The subject is, for example, a person receiving care who needs nursing care or an infant who needs childcare.

The health management system 1 includes a sensor module 10 and an information processing device 20. The sensor module 10 is a small device for detecting excretion, etc., of a subject. The sensor module 10 is worn by the subject when in use.

The information processing device 20 has a function of estimating the health condition of the subject based on the detection value output from the sensor module 10. As an example, the information processing device 20 is a PC, a tablet PC, a mobile phone such as a smartphone or a feature phone, or a personal digital assistant (PDA). As an example, the information processing device 20 is installed at the entrance to the subject's room, such as a hospital room or a nursery room. As an example, the information processing device 20 may be a terminal installed in a nurse's station in a hospital ward. As an example, the information processing device 20 may be a terminal owned by a caregiver who cares for the subject or a childcare worker who cares for the subject.

The sensor module 10 and the information processing device 20 are connected via a communication path N1. The communication path N1 is a transmission path for wireless communication, for example, a wireless transmission path for Bluetooth (registered trademark). In order to prevent the drawing from becoming too cluttered, FIG. 1 illustrates one sensor module 10 and one information processing device 20, but multiple sensor modules 10 and/or multiple information processing devices 20 may be included in the health management system 1.

2 is a diagram showing a schematic diagram of how the sensor module 10 is attached. In the example shown in FIG. 2, the sensor module 10 is attached to a diaper 50 worn by the subject. The sensor module 10 is placed in a nonwoven bag 60, and the nonwoven bag 60 containing the sensor module 10 is attached inside the diaper 50. Note that FIG. 2 illustrates an example in which the sensor module 10 is attached to the diaper 50, but the method in which the subject attaches the sensor module 10 is not limited to the example shown in FIG. 2. As an example, the sensor module 10 may be attached to other wearable items such as underwear of the subject. As an example, the sensor module 10 may be attached to the subject by being sandwiched between the subject's body and the underwear worn by the subject.

(Configuration of the sensor module)
3 is a block diagram illustrating an example of a functional configuration of the sensor module 10. The sensor module 10 includes a sensor group 11, a transmitter 12, a controller 13, a power source 14, a substrate 15, and a housing 16.

(sensor group)
The sensor group 11 includes one or more sensors that detect information about excrement. Excrement is a substance released from an organism, such as feces, urine, or gas (fart). In the example of FIG. 3, the sensor group 11 includes a gas sensor 111 and a temperature and humidity sensor 112. The gas sensor 111 is a sensor that detects the concentration and/or components of gas generated by the excrement of the subject. The temperature and humidity sensor 112 is a sensor that detects temperature and humidity. In the example of FIG. 3, the sensor group 11 includes one gas sensor 111 and one temperature and humidity sensor 112, but the sensor group 11 may include multiple gas sensors 111 and/or multiple temperature and humidity sensors 112.

(Transmitter, controller, power supply)
The transmitter 12 is a wireless module that transmits detection values of the sensor group 11 by wireless communication. For example, the transmitter 12 performs wireless communication using Wi-Fi (registered trademark) or Bluetooth (registered trademark). The controller 13 controls each part of the sensor module 10. The power source 14 supplies power to each part of the sensor module 10. The sensor group 11, the transmitter 12, and the controller 13 are attached to a substrate 15.

(Housing)
The housing 16 houses the substrate 15 and each of the components attached to the substrate 15. The housing 16 has one or more openings 161 (see FIG. 1 ) through which gas flows into the inside of the housing 16. Inside the housing 16, the gas sensor 111 is disposed at a position closer to the openings 161 than the transmitting unit 12 is.

Figure 4 is a diagram illustrating the internal configuration of the sensor module 10. In the illustrated example, the gas sensor 111 is located closer to the opening 161 than the transmission unit 12. The power source 14 is located on the back side of the substrate 15 opposite the side on which the transmission unit 12 is located, far from the opening 161. The sensor module 10 is worn on the subject with the opening 161 oriented so that it is located on the lower body side of the subject. When worn in this manner, the gas sensor 111 is located on the lower body side of the person being cared for. Gas generated from excrement flows into the interior of the housing 16 through the opening 161 of the housing 16 and is detected by the gas sensor 111.

The sensor module 10 transmits the detection values of the sensor group 11 via wireless communication at a predetermined timing (e.g., every second, every minute, etc.). The transmitted detection values are received by the information processing device 20. When transmitting the detection values of the sensor group 11, the sensor module 10 may transmit identification information (ID) that identifies the sensor module 10 via wireless communication together with the detection values.

(Configuration of information processing device)
5 is a block diagram illustrating an example of the configuration of the information processing device 20. The information processing device 20 is realized using a general-purpose computer, and includes a processor 21, a primary memory 22, a secondary memory 23, an input/output IF 24, a communication IF 25, and a bus 26. The processor 21, the primary memory 22, the secondary memory 23, the input/output IF 24, and the communication IF 25 are connected to each other via the bus 26.

The health state estimation program P1 and the learned model LM1 are stored in the secondary memory 23. The processor 21 expands the health state estimation program P1 and the learned model LM1 stored in the secondary memory 23 onto the primary memory 22. The processor 21 then executes each step included in the health state estimation method M1 according to the instructions included in the health state estimation program P1 expanded onto the primary memory 22. The learned model LM1 expanded onto the primary memory 22 is used when the processor 21 executes the estimation step M11 (described later) of the health state estimation method M1. Note that the health state estimation program P1 being stored in the secondary memory 23 refers to the source code or the executable file obtained by compiling the source code being stored in the secondary memory 23. Also, the learned model LM1 being stored in the secondary memory 23 refers to the parameters defining the learned model LM1 being stored in the secondary memory 23.

Devices that can be used as the processor 21 include, for example, a CPU (Central Processing Unit), a GPU (Graphic Processing Unit), a DSP (Digital Signal Processor), an MPU (Micro Processing Unit), an FPU (Floating point number Processing Unit), a PPU (Physics Processing Unit), a microcontroller, or a combination of these. The processor 21 is sometimes called a "computing device."

In addition, a device that can be used as the primary memory 22 can be, for example, a semiconductor RAM (Random Access Memory). The primary memory 22 is sometimes called a "main memory". In addition, a device that can be used as the secondary memory 23 can be, for example, a flash memory, a hard disk drive (HDD), a solid state drive (SSD), an optical disk drive (ODD), a floppy (registered trademark) disk drive (FDD), or a combination thereof. The secondary memory 23 is sometimes called an "auxiliary memory". The secondary memory 23 may be built into the information processing device 20, or may be built into another computer (for example, a computer constituting a cloud server) connected to the information processing device 20 via the input/output IF 24 or the communication IF 25. In this embodiment, the storage in the information processing device 20 is realized by two memories (the primary memory 22 and the secondary memory 23), but is not limited to this. In other words, the storage in the information processing device 20 may be realized by one memory. In this case, for example, one storage area of the memory can be used as the primary memory 22, and another storage area of the memory can be used as the secondary memory 23.

An input device and/or an output device are connected to the input/output IF 24. Examples of the input/output IF 24 include interfaces such as USB (Universal Serial Bus), ATA (Advanced Technology Attachment), SCSI (Small Computer System Interface), and PCI (Peripheral Component Interconnect). Examples of input devices connected to the input/output IF 24 include a keyboard, a mouse, a touchpad, a microphone, or a combination of these. Data acquired from a user in the health condition estimation method M1 is input to the information processing device 20 via these input devices and stored in the primary memory 22. Examples of output devices connected to the input/output IF 24 include a display, a projector, a printer, a speaker, a headphone, or a combination of these. Information provided to a user in the health condition estimation method M1 is output from the information processing device 20 via these output devices. Note that the information processing device 20 may each have a built-in keyboard that functions as an input device and a display that functions as an output device, like a laptop computer. Alternatively, the information processing device 20 may have a built-in touch panel that functions as both an input device and an output device, such as a tablet computer.

The communication IF 25 (an example of a communication unit) is connected to other computers via a network in a wired or wireless manner. Examples of the communication IF 25 include interfaces such as Ethernet (registered trademark) and Wi-Fi (registered trademark). Available networks include a PAN (Personal Area Network), a LAN (Local Area Network), a CAN (Campus Area Network), a MAN (Metropolitan Area Network), a WAN (Wide Area Network), a GAN (Global Area Network), or an internetwork including these networks. The internetwork may be an intranet, an extranet, or the Internet. Data acquired by the information processing device 20 from other computers in the health condition estimation method M1, and data provided by the information processing device 20 to other computers in the health condition estimation method M1 are transmitted and received via these networks.

In this embodiment, a configuration is adopted in which health state estimation method M1 is executed using a single processor (processor 21), but the present invention is not limited to this. That is, a configuration may be adopted in which health state estimation method M1 is executed using multiple processors. In this case, the multiple processors that cooperate to execute health state estimation method M1 may be provided in a single computer and configured to be able to communicate with each other via a bus, or may be provided in a distributed manner in multiple computers and configured to be able to communicate with each other via a network. As an example, a processor built in a computer that constitutes a cloud server and a processor built in a computer owned by a user of the cloud server may cooperate to execute health state estimation method M1.

In addition, in this embodiment, a configuration is adopted in which the learned model LM1 is stored in a memory (secondary memory 23) built into the same computer as the processor (processor 21) that executes the health state estimation method M1, but the present invention is not limited to this. That is, a configuration may be adopted in which the learned model LM1 is stored in a memory built into a computer different from the processor that executes the health state estimation method M1. In this case, the computer with the built-in memory that stores the learned model LM1 is configured to be able to communicate with the computer with the built-in processor that executes the health state estimation method M1 via a network. As an example, a mode can be considered in which the learned model LM1 is stored in a memory built into a computer that constitutes a cloud server, and a processor built into a computer owned by a user of the cloud server executes the health state estimation method M1.

In addition, in this embodiment, a configuration is adopted in which the learned model LM1 is stored in a single memory (secondary memory 23), but the present invention is not limited to this. That is, a configuration may be adopted in which the learned model LM1 is distributed and stored in multiple memories. In this case, the multiple memories that store the learned model LM1 may be provided in a single computer (which may or may not be a computer with a built-in processor that executes the health state estimation method M1), or may be distributed and provided in multiple computers (which may or may not include a computer with a built-in processor that executes the health state estimation method M1). As an example, a configuration may be considered in which the learned model LM1 is distributed and stored in memories built into each of the multiple computers that make up the cloud server.

In this embodiment, the communication IF 25 performs short-range wireless communication (Wi-Fi (registered trademark) or Bluetooth (registered trademark), etc.) with the sensor module 10 and receives the detection values of the sensor group 11 output from the sensor module 10.

The processor 21, the primary memory 22, and the secondary memory 23 are examples of a controller according to this specification. In other words, the controller includes at least one processor 21 that executes each process according to a predetermined program, and at least one memory (the primary memory 22 and the secondary memory 23) that stores the program.

The trained model LM1 is a trained model that derives output data indicating the health condition of a subject from detection data representing a time series of detection values of the sensor group 11. The trained model LM1 may be any machine learning model capable of generating a class indicating a health condition based on detection data. In this embodiment, the trained model LM1 is a trained model constructed by supervised machine learning using a training dataset that is a set of combinations of detection data representing a time series of detection values of the sensors and classes indicating the health condition of the subject. The trained model LM1 can be realized, for example, by a convolutional neural network (CNN), a recurrent neural network (RNN), a long short-term memory (LSTM), a deep neural network (DNN), or a combination thereof.

The input of the trained model LM1 is detection data representing a time series of detection values of the sensor group 11. The detection data may be, for example, time series data in which the detection values of the sensors (e.g., gas concentration and temperature) are arranged in a time series, or image data representing the signal waveforms of the detection values of the sensors.

The output of the trained model LM1 is a class indicating the subject's health condition. As an example, the class indicates one of "good," "normal," or "needs observation." As another example, the class may indicate the stage of a certain medical condition, such as diabetes. As another example, the class may indicate the type of medical condition with which the subject is presumed to be ill.

In addition, the class indicates, for example, the nutritional status of the subject. The class indicates, for example, whether one or more nutrients are sufficient. More specifically, the class indicates, for example, whether there is a fiber (dietary fiber) deficiency and whether there is a moisture deficiency. In this case, the class is, for example, one of four types: "fiber OK, moisture OK," "fiber OK, moisture NG," "fiber NG, moisture OK," and "fiber NG, moisture NG." "Fiber OK, moisture OK" indicates that there is no problem with both fiber and moisture. "Fiber OK, moisture NG" indicates that there is enough fiber but not enough moisture. "Fiber NG, moisture OK" indicates that there is enough moisture but not enough fiber. "Fiber NG, moisture NG" indicates that there is a both fiber and moisture deficiency.

The processor 21 executes an estimation process to estimate the health state using the detection data representing the time series of the detection values of the sensor group 11 and the learned model LM1.

(Trained model)
FIG. 6 is a diagram showing a schematic diagram of an example of the trained model LM1 according to the present embodiment. As shown in the figure, input data is input to the trained model LM1. The trained model LM1 is composed of, for example, a convolutional layer, a pooling layer, and a combination layer. In the convolutional layer, the input data is convoluted by filtering. The convolutional data is subjected to a pooling process in the pooling layer. This improves the model's ability to recognize changes in the position of features in the data. The pooling data is processed in the combination layer, and is converted into the output data of the trained model LM1, i.e., the format of the estimated health condition, and is output.

That is, the input data input to the trained model LM1 is passed through each layer shown in FIG. 6 in the order shown, and an estimated health condition result is output. Note that the output format of the estimated result is not particularly limited. For example, the health condition may be indicated as text data.

<Health status estimation method flow>
7 is a flow diagram illustrating the flow of a health condition estimation method M1 performed by the information processing device 20. Note that some steps may be executed in parallel or in a different order.

(Step S101)
In step S101, the processor 21 receives detection values of the sensor group 11 via the communication IF 25. The received detection values are accumulated in the secondary memory 23 in chronological order.

(Step S102)
In step S102, the processor 21 determines whether the subject is wearing the sensor module 10 based on the detection value received from the sensor module 10. This determination is made, for example, by determining whether the temperature detected by the temperature and humidity sensor 112 is equal to or higher than a threshold. For example, if the temperature detected by the temperature and humidity sensor 112 is equal to or higher than the threshold, the processor 21 determines that the sensor module 10 is worn by the subject. On the other hand, if the temperature detected by the temperature and humidity sensor 112 is lower than the threshold, the processor 21 determines that the sensor module 10 is not worn by the subject. Note that the determination method in step S102 is not limited to this, and various other methods may be adopted.

If it is determined that the subject is wearing the sensor module 10 (step S102; YES), the processor 21 proceeds to the process of step S103. On the other hand, if it is determined that the subject is not wearing the sensor module 10 (step S102; NO), the processor 21 returns to the process of step S101.

(Step S103)
In step S103, the processor 21 inputs detection data representing a time series of detection values of the sensor group 11 to the learned model LM1, and obtains an estimated result of the health condition output from the learned model LM1, thereby estimating the health condition of the subject. Particularly in this embodiment, the processor 21 estimates the nutritional condition of the subject using the detection data and the learned model LM1. The information indicating the nutritional condition of the subject includes, for example, information indicating whether or not fiber (dietary fiber) is sufficient and/or information indicating whether or not water is sufficient.

(Step S104)
In step S104, the processor 21 outputs the information indicating the health condition. As an example, the processor 21 outputs the information indicating the health condition to a display or the like.

Figure 8 is a diagram illustrating an example of a screen SC1 displayed on the display. Screen SC1 displays information Img11 including the subject's name, identification information (ID), etc., as well as information Inf12 indicating the subject's health condition.

The caregiver who is caring for the subject can understand the subject's health condition by checking the information shown on the display. By understanding the subject's health condition, the caregiver can determine what medication to prescribe and reflect the subject's health condition in future meal plans.

The processor 21 may also identify information for improving the health condition of the subject, such as a menu recommended for the subject, based on the output of the trained model LM1. In this case, as an example, the processor 21 may select a menu corresponding to the output of the trained model LM1 from among a plurality of menus in a rule-based manner. As another example, the processor 21 may estimate a menu suitable for the subject using a trained model that derives a class indicating the type of menu from information indicating the nutritional state.

<Effects of this embodiment>
As described above, according to this embodiment, the sensor module 10 transmits the detection values of the gas sensor 111 and the temperature and humidity sensor 112 to the information processing device 20 by wireless communication. Since the detection values of the sensor group 11 are transmitted by wireless communication, there is no need to connect the sensor module 10 to other devices by wire, and this can reduce the burden on the subject in daily life. Furthermore, according to this embodiment, since the detection values of the temperature and humidity sensor 112 are transmitted to the information processing device 20, the health condition of the subject can be estimated with higher accuracy.

Furthermore, according to this embodiment, the information processing device 20 executes a process of estimating the subject's health condition using a learned model LM1 that has been machine-learned from detection data that represents a time series of detection values from the sensor group 11. The information processing device 20 estimates the subject's health condition by inputting the detection data that represents a time series of detection values from the sensor group 11 into the learned model LM1. This makes it possible to improve the accuracy of estimating the subject's health condition.

[Embodiment 2]
Other embodiments of the present invention will be described below. For ease of explanation, the same reference numerals are given to members having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

The health management system 1B of the second embodiment includes an information processing device 20B instead of the information processing device 20 of the first embodiment. The information processing device 20B has a configuration similar to that of the information processing device 20 shown in FIG. 5. In other words, the information processing device 20B includes a processor 21, a primary memory 22, a secondary memory 23, an input/output IF 24, and a communication IF 25. These components are similar to those of the first embodiment described above, and the description thereof will not be repeated.

The secondary memory 23 of the information processing device 20B stores the trained model generation program P2 and the training dataset LD1 in addition to the health state estimation program P1 and the trained model LM1. The processor 21 expands the trained model generation program P2 stored in the secondary memory 23 onto the primary memory 22. The processor 21 then executes each step included in the trained model generation method M2 according to the instructions included in the trained model generation program P2 expanded onto the primary memory 22. The training dataset LD1 is used when the processor 21 executes the trained model generation method M2. Note that the trained model generation program P2 being stored in the secondary memory 23 refers to the source code, or an executable file obtained by compiling the source code, being stored in the secondary memory 23.

(Flow of trained model generation method)
9 is a flow diagram illustrating a flow of a trained model generation method M2 performed by the information processing device 20B. Note that some steps may be executed in parallel or in a different order.

(Step S201)
In step S201, the processor 21 collects a learning dataset LD1. The learning dataset LD1 is a collection of teacher data used in the generation process of the trained model LM1. The teacher data is data that combines detection data indicating a time series of detection values of the sensor group 11 and a class indicating the health state of a subject.

The detection data included in the learning dataset LD1 is data showing a time series of detection values output by the sensor module 10 worn by the subject. The class showing the subject's health condition is, for example, the result of a medical professional examining the subject based on the results of the subject's blood test. For example, the class shows "good", "normal", or "needs observation".

The class also indicates, for example, the nutritional status of the subject. The class indicates, for example, whether one or more nutrients are sufficient. More specifically, the class indicates, for example, whether there is a fiber (dietary fiber) deficiency and whether there is a moisture deficiency. In this case, the class is, for example, one of four types: "fiber OK, moisture OK," "fiber OK, moisture NG," "fiber NG, moisture OK," and "fiber NG, moisture NG."

The processor 21 adds time information to the detection data and stores it in a specified storage device (e.g., secondary memory 23), and also adds time information to the class indicating the subject's health condition and stores it in the specified storage device. The processor 21 may collect detection data and classes for one subject, or may collect detection data and classes for multiple subjects.

(Step S202)
In step S202, the processor 21 executes a generation process for generating a trained model by machine learning the correlation between the detection data and the class, using the detection data and the class to which the detection data belongs. In this operation example, the processor 21 reads out a pair of the detection data and the class associated based on the time information from the storage device, and generates a trained model LM1 by machine learning the correlation between the read detection data and the class.

In addition, the information processing device 20B executes a health condition estimation process. The health condition estimation process executed by the information processing device 20B is similar to the health condition estimation process executed by the information processing device 20 according to the above-described first embodiment, and therefore a description thereof will be omitted here.

Effects of the embodiment
As described above, according to the present embodiment, the information processing device 20B generates the learned model LM1 by using the detection data and the class. By using the learned model LM1, the information processing device 20B can estimate the health condition of the subject with higher accuracy.

Furthermore, according to this embodiment, information regarding the excrement of a subject wearing the sensor module 10 is collected using the sensor module 10. Because the sensor module 10 transmits the detection values of the sensor group 11 by wireless communication, the sensor module 10 does not need to be connected to another device by wire, and is easy to wear. Therefore, for example, it can be worn by a care recipient with a high level of care needs who has difficulty moving to the toilet, and detection data can be collected from such a care recipient. In this way, according to this embodiment, detection data of various subjects, such as care recipients with a high level of care needs, can be easily collected without guiding them to the toilet.

[Embodiment 3]
Other embodiments of the present invention will be described below. For ease of explanation, the same reference numerals are given to members having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

Figure 10 is a diagram showing a schematic configuration of a health management system 1C according to embodiment 3. The health management system 1 includes a sensor module 10, an information processing device 20C, an information processing device 20D, an information processing device 30, and a user terminal 40.

The information processing device 20C has a function of receiving detection values of the sensor group 11 from the sensor module 10 via wireless communication. The information processing device 20C also has a function of displaying various information such as the subject's health condition. As an example, the information processing device 20C is installed at the entrance to the subject's room, such as a hospital room or nursery room. The information processing device 20D is a server device such as a cloud server, and has a function of estimating the subject's health condition based on the detection values output from the sensor module 10.

The information processing device 30 and the user terminal 40 are devices to which the estimated results of the subject's health condition are notified. As an example, the information processing device 30 is a terminal installed in a nurse's station in a hospital ward. As an example, the user terminal 40 is a terminal owned by a caregiver who cares for the subject or a childcare worker who looks after the subject.

The sensor module 10 and the information processing device 20C are connected via a communication path N1. The communication path N1 is a transmission path for wireless communication, for example, a wireless transmission path for Bluetooth (registered trademark). In order to prevent the drawing from becoming too cluttered, FIG. 1 illustrates one sensor module 10 and one information processing device 20, but multiple sensor modules 10 and/or multiple information processing devices 20 may be included in the health management system 1C.

The information processing device 20C, the information processing device 20D, the information processing device 30, and the user terminal 40 are connected via a network N2. The network N2 may be, for example, a wired LAN (Local Area Network), a wireless LAN, the Internet, or a combination of these.

Information processing device 20C and information processing device 20D have a configuration similar to information processing device 20 shown in FIG. 5. Information processing device 20C and information processing device 20D have a configuration similar to information processing device 20 shown in FIG. 5. In other words, information processing device 20C and information processing device 20D include a processor 21, a primary memory 22, a secondary memory 23, an input/output IF 24, and a communication IF 25. These components are similar to those in embodiment 1 described above, and the description thereof will not be repeated.

The secondary memory 23 of the information processing device 20D, like the information processing device 20B, stores the health state estimation program P1 and the trained model LM1, as well as the trained model generation program P2 and the learning dataset LD1.

The secondary memory 23 of the information processing device 20D further stores a trained model generation program P3 and a trained model LM2. The processor 21 expands the trained model generation program P3 and the trained model LM2 stored in the secondary memory 23 onto the primary memory 22. The processor 21 then executes each step included in the trained model generation method M3 according to the instructions included in the trained model generation program P3 expanded onto the primary memory 22. The trained model LM2 expanded onto the primary memory 22 is used when the processor 21 executes the trained model generation method M3. Note that the trained model generation program P3 being stored in the secondary memory 23 refers to the source code or the executable file obtained by compiling the source code being stored in the secondary memory 23. Also, the trained model LM2 being stored in the secondary memory 23 refers to the parameters defining the trained model LM2 being stored in the secondary memory 23.

The trained model LM2 derives output data indicating the type of excrement from detection data representing a time series of detection values of the sensor group 11. The trained model LM2 may be any machine learning model capable of generating data indicating the type of excrement based on the detection data. The trained model LM2 is a trained model constructed by supervised machine learning using a training dataset that is a set of combinations of detection data representing a time series of detection values of the sensor group and classes indicating the type of excrement. The trained model LM2 may be realized, for example, by a CNN, an RNN, an LSTM, a DNN, or a combination of these.

The input of the trained model LM2 is detection data representing a time series of detection values of the sensor group 11. The detection data may be, for example, time series data in which the detection values of the sensors (e.g., gas concentration and temperature) are arranged in a time series, or image data representing the signal waveforms of the detection values of the sensors.

The output data of the trained model LM2 is data indicating the type of excrement. The type of excrement includes, by way of example, data indicating feces, urine, or gas (farts). The data indicating the type of excrement may also include data indicating the characteristics of the excrement (hard/soft).

<Health status estimation method flow>
Fig. 11 is a sequence diagram illustrating the flow of a health condition estimation method M4 performed by the health management system 1C. The health condition estimation method M4 is an example of a method in which the sensor module 10 is attached to a subject's clothing, and at least one of detecting the subject's defecation and estimating the subject's health condition is performed based on the detection values of the sensor group 11. Note that some of the steps shown in Fig. 11 may be performed in parallel or in a different order.

(Step S301)
In step S301, the sensor module 10 wirelessly transmits the detection values of the sensor group 11. The information processing device 20C receives the detection values output by the sensor module 10.

(Step S302)
In step S302, the information processing device 20C transmits the detection value received from the sensor module 10 to the information processing device 20D via the network N2. The information processing device 20C may transmit the received detection value each time it receives a detection value from the sensor module 10, or may temporarily store the detection value received from the sensor module 10 in its own memory and transmit the stored detection value at a predetermined timing to the information processing device 20D. The information processing device 20D receives the detection value from the information processing device 20C.

(Step S303)
In step S303, the processor 21 of the information processing device 20D executes an estimation process to estimate a health state using detection data representing a time series of detection values of the sensor group 11 and the learned model LM1.

(Step S304)
In step S304, the processor 21 executes an estimation process to estimate the type of excrement using the detection data representing a time series of the detection values of the sensor group 11 and the learned model LM2.

(Step S305)
In step S305, the information processing device 20D transmits inference result data indicating the inference result of the health condition and the inference result of the type of excrement to the information processing device 30. The information processing device 30 receives the inference result data from the information processing device 20D. Note that, in addition to transmitting the inference result data to the information processing device 30, the information processing device 20D may transmit the inference result data to the information processing device 20C.

(Step S306)
In step S306, the information processing device 30 transmits the estimation result data to the user terminal 40. The user terminal 40 displays the estimation result indicated by the estimation result data on a display. In addition, when the information processing device 20D also transmits the estimation result data to the information processing device 20C, the information processing device 20C may display the estimation result on the display of its own device.

Figure 12 is a diagram illustrating an example of screen SC2 displayed on the display of user terminal 40. Screen SC2 displays information Img11 including the subject's name and identification information (ID), etc., as well as information Inf12 indicating the subject's health condition. Screen SC2 also displays information Inf13 indicating the type of excrement presumed to have been excreted by the subject. The screen of Figure 12 may also display the detection results of temperature and humidity sensor 112, i.e., the subject's body temperature.

According to this embodiment, a caregiver who cares for the subject can check the information displayed on the display to understand the subject's health condition and whether or not the subject has excreted. The caregiver can reflect the subject's health condition in the subject's future medication and future meal contents, and can understand whether or not the subject needs excretion care.

In addition, by displaying the subject's body temperature, caregivers and others can more easily grasp the subject's health condition. For example, during times when infectious diseases such as influenza are prevalent, the subject's body temperature and health condition can be managed simply by having the subject wear the sensor module 10, without having to separately measure the subject's body temperature.

In the above embodiment, the information processing device 20D transmits the estimation result data to the information processing device 30, and the information processing device 30 transmits the estimation result data to the user terminal 40. The information processing device 20D may transmit the estimation result to the user terminal 40 without passing through the information processing device 30.

[Embodiment 4]
Other embodiments of the present invention will be described below. For ease of explanation, the same reference numerals are given to members having the same functions as those described in the above embodiment, and the description thereof will not be repeated.

Figure 13 is a diagram showing a schematic configuration of a health management system 1D according to embodiment 4. The health management system 1D includes a sensor module 10, a relay device 20E, an information processing device 20D, an information processing device 30, and a user terminal 40. In other words, the health management system 1D includes a relay device 20E instead of the information processing device 20C of the health management system 1D according to embodiment 3 described above.

The relay device 20E is a device that relays data exchange between the sensor module 10 and the information processing device 20D. As an example, the relay device 20E is installed in the subject's room, such as a hospital room or a nursery room. The relay device 20E includes, for example, a processor, a primary memory, a secondary memory, a communication IF, and an input/output IF. The communication IF is a wireless module for wireless communication with the sensor module 10. In this embodiment, the detection value output by the sensor module 10 is transmitted to the information processing device 20D via the relay device 20E.

According to this embodiment, a caregiver who cares for the subject can check the information displayed on the display to understand the subject's health condition and whether or not the subject has excreted. The caregiver can reflect the subject's health condition in the subject's future medication and future meal contents, and can understand whether or not the subject needs excretion care.

[Additional Notes]
[Additional Note 1]
The data included in the input data input to the trained model LM1 is not limited to those shown in the above-mentioned embodiments. The input data may include other data. As an example, the input data may include dietary data representing the dietary content ingested by the subject. In this case, the trained model LM1 derives output data from the input data including the detection data and dietary data. In this case, the estimation process executed by the processor 21 is a process of estimating a health condition using the detection data, dietary data, and the trained model LM1.

The dietary data may, for example, include information indicating the types and amounts of nutrients ingested by the subject, and may also include information indicating the menu of the meal ingested by the subject. The dietary data may, for example, include data representing images of the menu of the meal ingested by the subject, or photographed images of the menu.

[Additional Note 2]
The input data may also include, for example, information indicating attributes of the subject, such as the subject's sex, the subject's room, age, level of care required, type of disease, or stage of disease.

The input data may also include, for example, information regarding administration of medication to the subject. The information regarding administration may include, for example, information indicating the type of medication, the amount administered, and the timing of administration. The information regarding administration may also include, for example, information indicating the timing of administration of a laxative.

[Additional Note 3]
In each of the above-described embodiments, the processor 21 may construct a trained model LM1 for each subject and estimate the health condition of each subject using the trained model LM1 corresponding to each subject. In this case, the transmitter 12 of the sensor module 10 transmits the detection value of the sensor group 11 and identification information for identifying the sensor module 10 via wireless communication.

When the processor 21 receives the detection values and identification information of the sensor group 11, it estimates the subject's health condition using the detection data representing the time series of the detection values of the sensor group 11 in the learned model LM1 corresponding to the received identification information. In the estimation phase of the learned model LM1, the processor 21 constructs a learned model LM1 for each subject using the detection data and classes accumulated for each subject.

[Additional Note 4]
In the above embodiment, the processor 21 estimates the health condition of the subject using one learned model LM1. The processor 21 may estimate the health condition of the subject using a plurality of learned models LM1.

As an example, the processor 21 may estimate the health condition of the subject using two learned models, a first learned model LM11 and a second learned model LM12. The inputs of the first learned model LM11 and the second learned model LM12 are both detection data representing a time series of detection values of the sensor group. Meanwhile, the output of the first learned model LM11 is a class indicating whether or not there is a fiber deficiency. Also, the output of the second learned model LM12 is a class indicating whether or not there is a moisture deficiency.

In this case, the processor 21 outputs, for example by displaying on a display, information indicating whether or not there is a fiber deficiency, which is obtained by inputting the detection data into the first trained model LM11, and information indicating whether or not there is a moisture deficiency, which is obtained by inputting the detection data into the second trained model LM12. By checking the information displayed on the display, the caregiver or the like can determine whether or not the subject is lacking in fiber and whether or not there is a moisture deficiency. By understanding the nutritional status of the subject, the caregiver or the like can determine what medication to prescribe to the subject and reflect the subject's nutritional status in future meal plans.

The processor 21 may also identify information for improving the health condition of the subject, such as a menu recommended for the subject, based on the output of the first trained model LM11 and the output of the second trained model LM12. In this case, as an example, the processor 21 may select, from among a plurality of menus, a menu that corresponds to a combination of the output of the first trained model LM11 and the output of the second trained model LM12 in a rule-based manner. As another example, the processor 21 may estimate a menu suitable for the subject using a trained model that derives a class indicating the type of menu from information indicating the nutritional state.

[Additional Note 5]
The functions of the information processing device 20 and the information processing device 20B in each of the above-mentioned embodiments may be realized by a plurality of devices, or may be realized by a plurality of devices working together. For example, a first device that executes an estimation process and a second device that executes a generation process for generating a learned model LM1 may be configured as separate devices, and the above-mentioned information processing device 20B may be realized by the first device and the second device working together. In this case, in other words, the second device (an example of an information processing device) includes a controller, a memory, and a communication unit, and the communication unit receives the detection values of the sensor group 11 output from the sensor module 10, and the controller executes a generation process for generating a learned model LM1 by machine learning the correlation between the detection data and the class using detection data representing a time series of the detection values of the sensor group 11 and a class representing the health state of the subject.

[Additional Note 5]
The sensors included in the sensor group 11 of the sensor module 10 are not limited to those shown in the above-mentioned embodiments. The sensor group 11 may include a temperature sensor and/or a humidity sensor instead of the temperature and humidity sensor 112, for example. The sensor group 11 may also include an acceleration sensor. The acceleration sensor is a sensor that detects acceleration, and as an example, detects angular acceleration with respect to an axis (body axis) from the toes to the head of the subject, or linear acceleration in the body axis direction. When the sensor group 11 includes an acceleration sensor, the gas sensor 111 may be disposed closer to the opening 161 than the acceleration sensor. The temperature and humidity sensor 112 may also be disposed closer to the opening 161 than the acceleration sensor.

[Software implementation example]
The control blocks of the information processing devices 20, 20B, 20C, 20D, the information processing device 30, and the user terminal 40 may be realized by a logic circuit (hardware) formed in an integrated circuit (IC chip) or the like, or may be realized by software.

In the latter case, the information processing devices 20, 20B, 20C, and 20D, the information processing device 30, and the user terminal 40 are provided with a computer that executes instructions of a program, which is software that realizes each function. This computer is provided with, for example, one or more processors, and a computer-readable recording medium that stores the program. In the computer, the processor reads the program from the recording medium and executes it, thereby achieving the object of the present invention. The processor can be, for example, a CPU (Central Processing Unit). The recording medium can be a "non-transient tangible medium", such as a ROM (Read Only Memory), as well as a tape, a disk, a card, a semiconductor memory, a programmable logic circuit, or the like. The computer may further include a RAM (Random Access Memory) that expands the program. The program may be supplied to the computer via any transmission medium (such as a communication network or a broadcast wave) that can transmit the program. Note that one aspect of the present invention can also be realized in the form of a data signal embedded in a carrier wave, in which the program is embodied by electronic transmission.

The present invention is not limited to the above-described embodiments, and various modifications are possible within the scope of the claims. The technical scope of the present invention also includes embodiments obtained by appropriately combining the technical means disclosed in different embodiments.

REFERENCE SIGNS LIST 1 Health management system 10 Sensor module 11 Sensor group 111 Gas sensor 112 Temperature and humidity sensor 12 Transmitter 13 Controller 14 Power supply 15 Board 16 Housing 20 Information processing device 21 Processor 22 Primary memory 23 Secondary memory 24 Input/output IF
25 Communication IF

Claims (7)

The device includes a controller, a memory, and a communication unit.
The communication unit is
a sensor module including a sensor group for detecting information related to excrement, the sensor group including a gas sensor for detecting gas and a temperature and humidity sensor for detecting temperature and humidity , and a transmitter for transmitting detection values of the sensor group , the sensor module being attached to a subject and used to receive detection values of the sensor group transmitted by the sensor module;
The memory includes:
storing a trained model that derives output data indicating a health condition of the subject, the health condition including at least one of a stage of a disease and whether or not one or more nutrients are deficient, from detection data indicating a time series of detection values of the sensor group;
The controller:
An estimation process is performed to estimate the health state by inputting the detection data into the trained model;
The trained model is constructed by supervised machine learning using a set of combinations of data representing a time series of detection values of the sensor group and a class representing the health state as a training data set.
23. An information processing apparatus comprising: The communication unit is
receiving dietary data representative of meals consumed by the subject;
The trained model derives the output data from input data including the detection data and the meal data;
The estimation process is a process of estimating the health state by inputting the detection data and the dietary data into the trained model.
The information processing device according to claim 1 . The controller:
A generation process is further performed to generate the trained model by machine learning the correlation between data representing a time series of the detection values of the sensor group and a class representing the health state of the subject.
3. The information processing device according to claim 1 or 2. The device includes a controller, a memory, and a communication unit.
The communication unit is
a sensor module including a sensor group for detecting information related to excrement, the sensor group including a gas sensor for detecting gas and a temperature and humidity sensor for detecting temperature and humidity, and a transmitter for transmitting detection values of the sensor group, the sensor module being attached to a subject and used to receive detection values of the sensor group transmitted by the sensor module;
The controller:
and executing a generation process for generating a trained model by machine learning a correlation between the detection data representing a time series of the detection values of the sensor group and a class representing the health state of the subject, the class being information indicating a stage of a disease and/or information indicating whether one or more nutrients are deficient, by using the detection data representing the time series of the detection values of the sensor group and the class.
23. An information processing apparatus comprising: The controller:
At least one processor that executes the estimation process according to a predetermined program;
At least one memory storing the program.
3. The information processing apparatus according to claim 1, wherein the information processing apparatus is a computer. 3. A program for controlling the information processing device according to claim 1, comprising: a program for causing the controller to execute the estimation process. The present invention comprises a sensor module that is worn by a subject and an information processing device that estimates a health condition of the subject,
The sensor module includes:
a sensor group for detecting information related to excrement, the sensor group including a gas sensor for detecting gas and a temperature and humidity sensor for detecting temperature and humidity;
A transmitter for transmitting detection values of the group of sensors;
Equipped with
The information processing device includes:
The device includes a controller, a memory, and a communication unit.
The communication unit is
receiving the detection values of the sensor group transmitted by the sensor module;
The memory includes:
storing a trained model that derives output data indicating a health condition of the subject, the health condition including at least one of a stage of a disease and whether or not one or more nutrients are deficient, from detection data indicating a time series of detection values of the group of sensors;
The controller:
An estimation process is performed to estimate the health state by inputting the detection data into the trained model;
The trained model is constructed by supervised machine learning using a set of combinations of data representing a time series of detection values of the sensor group and a class representing the health state as a training data set.
A system characterized by:

Images