Program, information processing method, and information processing apparatus

Abstract

To provide a program or the like for outputting excretion information of a subject.SOLUTION: The program causes a computer to execute a process of acquiring sensor information obtained through a bed sensor provided in relation to a subject and outputting excretion information on excretion of the subject by inputting the acquired sensor information to a learning model that outputs the excretion information on excretion of the subject when the sensor information is input. Preferably, the sensor information includes a bed-entering time, a bed-leaving time, a mid-arousal time, a heart rate, a respiration rate, a body movement amount, a sleep time, or a sleep quality.SELECTED DRAWING: Figure 1

Description

[Technical Field]

[0001] The present disclosure relates to a program, an information processing method, and an information processing device. [Background technology]

[0002] Patent Document 1 discloses a system that determines whether a person being watched over on a bed is lying down or sitting up. [Prior art documents] [Patent documents]

[0003] [Patent Document 1] Patent No. 6582305 Summary of the Invention [Problem to be solved by the invention]

[0004] However, the invention of Patent Document 1 has a problem in that it is not possible to output excretion information of the subject.

[0005] In one aspect, an object is to provide a program or the like that outputs excretion information of a subject. [Means for solving the problem]

[0006] A program according to one aspect causes a computer to execute a process of acquiring sensor information obtained through a bed sensor provided in association with a subject, and outputting excretion information of the subject based on the acquired sensor information. [Effects of the Invention]

[0007] In one aspect, it becomes possible to output information about the subject's excretion. [Brief explanation of the drawings]

[0008] [Figure 1] FIG. 1 is an explanatory diagram showing an overview of a notification system. [Figure 2] FIG. 2 is a block diagram illustrating an example of the configuration of a server. [Figure 3] FIG. 2 is an explanatory diagram illustrating an example of a record layout of a sensor information DB. [Figure 4] FIG. 1 is an explanatory diagram of a learning model. [Figure 5] FIG. 10 is an explanatory diagram showing an example of a record layout of a training DB. [Figure 6] FIG. 1 is a block diagram illustrating an example of the configuration of a computer. [Figure 7] FIG. 10 is an explanatory diagram showing an example of an alert display. [Figure 8] 10 is a flowchart showing a procedure for generating a learning model. [Figure 9A] 10 is a flowchart illustrating an example of a processing procedure of a notification system. [Figure 9B] 10 is a flowchart illustrating an example of a processing procedure of a notification system. [Figure 10] FIG. 10 is an explanatory diagram showing an overview of a notification system according to a second embodiment. [Figure 11] 10 is a flowchart illustrating an example of a processing procedure related to relearning. [Figure 12] FIG. 10 is an explanatory diagram showing an overview of a notification system according to a third embodiment. [Figure 13] FIG. 11 is a block diagram showing an example of the configuration of a server according to a third embodiment. [Figure 14] FIG. 10 is an explanatory diagram showing an example of a record layout of a subject DB. [Figure 15] FIG. 10 is an explanatory diagram showing an example of a record layout of an excretion date and time DB. [Figure 16] FIG. 10 is an explanatory diagram showing an example of a record layout of an excretion time period DB. [Figure 17] 10 is a flowchart showing an example of a processing procedure related to an excretion time period. [Figure 18] 11 is a flowchart illustrating an example of a processing procedure of a notification system according to a third embodiment. [Figure 19] FIG. 13 is an explanatory diagram illustrating an example of a record layout of a sensor information DB according to the fourth embodiment. [Figure 20] FIG. 10 is an explanatory diagram of a learning model according to the fourth embodiment. [Figure 21] FIG. 10 is an explanatory diagram showing an example of a record layout of a training DB according to the fourth embodiment. [Figure 22] FIG. 11 is an explanatory diagram showing an example of an alert display according to the fourth embodiment. [Figure 23] 10 is a flowchart showing a procedure for generating a learning model according to the fourth embodiment. [Figure 24] 10 is a flowchart showing an example of a processing procedure of a notification system according to a fourth embodiment. [Figure 25] FIG. 13 is an explanatory diagram showing an example of a record layout of a subject DB according to the fifth embodiment. [Figure 26] FIG. 13 is an explanatory diagram showing an example of an alert display according to the fifth embodiment. [Figure 27] 13 is a flowchart showing an example of a processing procedure of a notification system according to a fifth embodiment. DETAILED DESCRIPTION OF THE INVENTION

[0009] (Embodiment 1) FIG. 1 is an explanatory diagram showing an overview of the notification system. The notification system acquires sensor information about a subject, and outputs excretion information about the subject by inputting the acquired sensor information into a learning model. The notification system includes an information processing device 10, a sensor I / F (Interface) device 20, and an information processing device 30. The information processing device 10, the sensor I / F device 20, and the information processing device 30 are connected via a network N. The subject is, for example, an elderly person or a physically or mentally disabled person living in each room in a facility such as a nursing home or a medical facility.

[0010] The information processing device 10 is an information processing device that processes, stores, and transmits / receives various types of information. The information processing device 10 is, for example, a server device, a personal computer, or a general-purpose tablet PC (personal computer). The information processing device 10 may be a cloud server device that provides functions included in the information processing device 10 as a cloud service. In this embodiment, the information processing device 10 will be described as the server 10.

[0011] The information processing device 30 is, for example, a server device, a smartphone, a tablet, a personal computer, a general-purpose tablet PC (personal computer), etc. In this embodiment, the information processing device 30 will be described as a computer 30.

[0012] The sensor I / F device 20 is provided, for example, for each subject who uses the notification system. A sensor for detecting the condition of the subject is connected to the sensor I / F device 20. In the example shown in FIG. 1, a bed sensor 21 is connected to the sensor I / F device 20 by wire or wirelessly. The sensor I / F device 20 is an information processing device that analyzes data output from the bed sensor 21 for a predetermined period (e.g., 10 minutes) and transmits the analyzed data to the server 10 via the network N.

[0013] One bed sensor 21 is installed in each bed of a subject. The bed sensor 21 is, for example, a sheet-like sensor and is installed between the bed mattress and the sheet. The bed sensor 21 detects the subject's heart rate, breathing, body movement (acceleration), signals indicating a state of awakening during the night (a state of waking up in the middle of the night), or sleep depth (e.g., REM sleep, shallow non-REM sleep, intermediate-depth non-REM sleep, deepest non-REM sleep, etc.) of the subject lying in bed. The bed sensor 21 detects the subject's heart rate, breathing, body movement, signals indicating a state of awakening during the night, or sleep depth, etc., by detecting, for example, the rate of change in air pressure in an air mattress, the subject's snoring and breathing sounds, and vibrations caused by turning over in bed.

[0014] The bed sensor 21 may be composed of conductive fabric and multiple piezoelectric elements, and may be configured to detect periodic micro-vibrations to detect the subject's heart rate, breathing, body movement, signals indicating a state of awakening during sleep, or the depth of sleep, etc.

[0015] The bed sensor 21 outputs signals indicating the subject's detected heart rate, breathing, body movement, and awakening state, or sleep depth, etc., to the sensor I / F device 20. The sensor I / F device 20 analyzes the signals indicating the subject's heart rate, breathing, body movement, and awakening state, or sleep depth, output by the bed sensor 21 for a predetermined period (e.g., 10 minutes), to determine the heart rate, breathing rate, amount of body movement, the time period when the subject woke up (hereinafter referred to as the time of awakening), or the quality of sleep. The heart rate is, for example, the average heart rate over 10 minutes. The breathing rate is, for example, the average breathing rate over 10 minutes. The amount of body movement is, for example, the number of times the subject's body movement occurred over 10 minutes. The sensor I / F device 20 determines the amount of body movement by measuring the number of times that vibrations (body movement) detected from turning over, etc., exceed a predetermined threshold. The quality of sleep is, for example, the depth of sleep over a predetermined period. The sensor I / F device 20 determines the time of awakening and the quality of sleep based on the number of turns, breathing sounds, number of snores, etc. If the sensor I / F device 20 detects both REM sleep and non-REM sleep within a predetermined period, it determines the sleep quality based on the higher proportion within the predetermined period. The sensor I / F device 20 transmits the determined heart rate, breathing rate, amount of body movement, time of awakening, or sleep quality to the server 10 at predetermined time intervals (e.g., every 10 minutes).

[0016] FIG. 2 is a block diagram showing an example of the configuration of a server. The server 10 includes a control unit 11, a storage unit 12, a communication unit 13, a mass storage unit 14, and a reading unit 15. The above-mentioned units are connected to each other via a bus. The control unit 11 is configured using one or more processors, such as a central processing unit (CPU), a microprocessing unit (MPU), or a graphics processing unit (GPU). The storage unit 12 includes a random access memory (RAM) or a read-only memory (ROM). The storage unit 12 stores various data and the like required for a control program 12P (a program product) executed by the control unit 11. The storage unit 12 also temporarily stores data and the like generated when the control program 12P is executed. The control unit 11 executes the control program 12P stored in the storage unit 12 as needed to perform various information processing and control processing related to the server 10.

[0017] The communication unit 13 is a communication module that transmits and receives information between the sensor I / F device 20 and the computer 30 via the network N. The mass storage unit 14 includes a RAM, a ROM, or the like. The mass storage unit 14 stores a sensor information DB 141, a learning model 142, and a training DB 143. The sensor information DB 141, the learning model 142, and the training DB 143 will be described later.

[0018] In this embodiment, the storage unit 12 and the mass storage unit 14 may be configured as an integrated storage device. The mass storage unit 14 may be configured with multiple storage devices. The mass storage unit 14 may be an external storage device connected to the server 10.

[0019] The reading unit 15 reads information stored in the portable storage medium 1a. The portable storage medium 1a is, for example, a CD (Compact Disc), a DVD (Digital Versatile Disc), a USB (Universal Serial Bus) memory, or an SD (Secure Digital). The reading unit 15 reads the control program 12P from the portable storage medium 1a. The control unit 11 stores the read control program 12P in the storage unit 12. The control unit 11 may download the control program 12P from another computer via the network N. In this case, the control unit 11 stores the downloaded control program 12P in the storage unit 12. The control unit 11 may store the read control program 12P in the mass storage unit 14.

[0020] In this embodiment, the server 10 may be configured with multiple servers. The server 10 may be a virtual machine virtually constructed by software within a single device. The server 10 may be a local server installed within the facility where the server 10 is located. The server 10 may be a cloud server connected for communication via a network N. Furthermore, the control program 12P may be executed on a single server, or may be distributed and executed on multiple servers interconnected via the network N.

[0021] FIG. 3 is an explanatory diagram showing an example of a record layout of the sensor information DB. The sensor information DB 141 stores sensor information obtained through the bed sensor 21. The sensor information DB 141 includes a date and time column, a bed entry time column, a bed exit time column, a sleep awakening time column, a heart rate column, a respiration rate column, a body movement column, a sleep duration column, or a sleep quality column. The date and time column stores the date and time when the server 10 receives the subject's heart rate, respiration rate, body movement, wakeup time, sleep quality, etc. from the sensor I / F device 20. The bed entry time column stores the time when reception of the subject's heart rate, respiration rate, body movement, wakeup time, sleep quality, etc. from the sensor I / F device 20 started (hereinafter referred to as the bed entry time). The bed exit time column stores the time when reception of the subject's heart rate, respiration rate, body movement, wakeup time, sleep quality, etc. from the sensor I / F device 20 ended (hereinafter referred to as the bed exit time). The midnight awakening time column stores the midnight awakening time received from the sensor I / F device 20. The heart rate column stores the heart rate received from the sensor I / F device 20. The respiratory rate column stores the respiratory rate received from the sensor I / F device 20. The body movement amount column stores the body movement amount received from the sensor I / F device 20. The sleep duration column stores the elapsed time from the time of going to bed to the present time (hereinafter referred to as sleep duration). The sleep quality column stores the quality of sleep received from the sensor I / F device 20.

[0022] The record for the date and time "2024 / 3 / 2 / 3:20" in Figure 3 stores the time of going to bed "22:30", time of getting out of bed "-", time of awakening during the night "3:05-3:08", heart rate "54 beats / minute", pulse rate "19 beats / minute", amount of body movement "5 times", sleep time "4:50", and sleep quality "non-REM sleep".

[0023] The sensor information stored in the sensor information DB 141 is not limited to the above examples. The control unit 11 may store, in the sensor information DB 141, time-series data indicating changes over time in the heart rate, the respiratory rate, the amount of body movement, or the quality of sleep.

[0024] FIG. 4 is an explanatory diagram of a learning model. The learning model 142 is a trained model that has been trained to output excretion information related to the excretion of the subject when sensor information is input. For example, a convolution neural network (CNN) is used for the learning model 142. The learning model 142 includes an input layer that receives sensor information, an intermediate layer that performs calculations, and an output layer that outputs information related to the excretion of the subject. The excretion information related to the excretion of the subject is, for example, a probability indicating whether or not excretion will occur after a predetermined time, or the time until excretion. In this embodiment, the excretion information will be described as a probability indicating whether or not excretion will occur after a predetermined time (hereinafter referred to as 10 minutes later). When sensor information is input, the learning model 142 outputs a probability indicating whether or not excretion will occur 10 minutes later. The probability output from the output layer is, for example, a value between 0 and 1. The sum of the probabilities output from the output layer is 1.0 (100%).

[0025] In FIG. 4, when the control unit 11 inputs sensor information at a certain point in time (for example, 3:20 on 3 / 2 / 2024) to the learning model 142, the learning model 142 outputs "excretion in 10 minutes: 0.8" and "no excretion in 10 minutes: 0.2" as probabilities indicating whether or not excretion will occur in 10 minutes. The control unit 11 acquires the highest probability value from the multiple probability values ​​output from the output layer as the estimation result. In FIG. 4, the control unit 11 acquires "excretion in 10 minutes" as the estimation result.

[0026] The control unit 11 may acquire the estimation result using a predetermined threshold. In the example of Fig. 4, when the probability "0.8" of "excretion in 10 minutes" is equal to or greater than a predetermined threshold (e.g., 0.75), the control unit 11 may acquire "excretion in 10 minutes" as the estimation result. The predetermined threshold can be changed as appropriate according to the embodiment.

[0027] The control unit 11 may input sensor information at multiple consecutive points in time (for example, 2024 / 3 / 2 / 3:00, 2024 / 3 / 2 / 3:10, and 2024 / 3 / 2 / 3:20) into the learning model 142 instead of sensor information at a certain point in time.

[0028] The probabilities indicating whether or not there will be an excretion after 10 minutes output by the learning model 142 are not limited to the above example. The learning model 142 may output the probability of excreting feces after 10 minutes, the probability of excreting urine after 10 minutes, and the probability of not excreting after 10 minutes. In this case, the learning model 142 outputs, for example, "excreting feces after 10 minutes: 0.6", "excreting urine after 10 minutes: 0.3", and "not excreting after 10 minutes: 0.1".

[0029] The learning model 142 may output the time until excretion instead of the probability indicating whether or not excretion will occur after 10 minutes. When the control unit 11 inputs sensor information at a certain point in time to the learning model 142, the learning model 142 outputs the probability of excretion after several minutes and the probability of not excreting. The learning model 142 outputs, for example, "excretion after 3 minutes: 0.1," "excretion after 5 minutes: 0.2," "excretion after 10 minutes: 0.6," and "not excreting: 0.1." The control unit 11 acquires the highest probability value from the multiple probability values ​​output from the output layer as the estimation result. The time until excretion output by the learning model 142 is not limited to the above example and can be changed as appropriate.

[0030] The learning model 142 may output a score indicating the likelihood of excretion in 10 minutes (hereinafter referred to as the excretion score). The excretion score is, for example, a value between 0 and 1. The closer the excretion score is to 0, the lower the likelihood that the subject will excrete in 10 minutes, and the closer the excretion score is to 1, the higher the likelihood that the subject will excrete in 10 minutes. When the excretion score is equal to or greater than a predetermined threshold (for example, 0.75), the control unit 11 may acquire "excretion will occur in 10 minutes" as the estimated result.

[0031] FIG. 5 is an explanatory diagram showing an example of a record layout of the training DB. The training DB 143 stores training data used to generate the learning model 142. The training DB 143 includes a sensor information sequence and an excretion sequence. The sensor information sequence stores sensor information for training obtained through a plurality of bed sensors 21. The sensor information sequence includes a date and time sequence, a bed entry time sequence, a bed getting-out time sequence, a mid-awakening time sequence, a heart rate sequence, a respiratory rate sequence, a body movement sequence, a sleep duration sequence, or a sleep quality sequence. The information stored in the date and time sequence, bed entry time sequence, bed getting-out time sequence, mid-awakening time sequence, a heart rate sequence, a respiratory rate sequence, a body movement sequence, a sleep duration sequence, or a sleep quality sequence is the same as that in the sensor information DB 141. The excretion sequence stores information indicating whether or not excretion has occurred 10 minutes after the sensor information is acquired ("1": excretion has occurred 10 minutes later / "0": excretion has not occurred 10 minutes later). The information stored in the excretion sequence is not limited to the above example. The excretion sequence may store information indicating whether or not there is an excretion after 10 minutes ("large": feces excreted after 10 minutes, "small": urination after 10 minutes, "no": no excretion after 10 minutes). In the example of Fig. 5, sensor information (date and time "2023 / 8 / 2 / 2:30", time of getting into bed "22:00", time of getting out of bed "-", time of awakening during the night "1:02-1:05", heart rate "56 beats / min", respiratory rate "18 breaths / min", amount of body movement "5 times", sleep time "4:30", and sleep quality "REM sleep") and excretion information ("0": no excretion after 10 minutes) are stored in association with each other.

[0032] A method for generating the learning model 142 will be described. The control unit 11 reads out from the training DB 143 a plurality of training data in which sensor information is associated with information indicating whether or not excretion has occurred 10 minutes later. When training sensor information is input to the learning model 142, the control unit 11 trains the learning model 142 so that, when the training sensor information is input to the learning model 142, the output value from the node corresponding to the correct value included in the training data approaches 1 and the output values ​​from the other output nodes approach 0. The control unit 11 optimizes parameters used in the calculation process in the intermediate layer so that the output value from the output layer of the learning model 142 approaches the correct value. The parameters are, for example, the weight (coupling coefficient) of the node or the coefficient of the activation function used in each node. The parameter optimization method is, for example, the backpropagation method or the steepest descent method. The control unit 11 stores the learning model 142 generated by the above process in the mass storage unit 14.

[0033] The learning model 142 may be generated by an external computer (not shown) or the like. In this case, the generated learning model 142 is deployed to the server 10 from the external computer or the like via the network N.

[0034] The learning model 142 may use algorithms such as, for example, LSTM (Long Short Term Memory), RNN (Recurrent Neural Network), GPT (Generative Pre-trained Transformer), SVM (Support Vector Machine), BERT (Bidirectional Encoder Representations from Transformers), Seq2Seq (Sequence to Sequence), Random Forest, or Decision Tree.

[0035] FIG. 6 is a block diagram showing an example configuration of a computer. The computer 30 includes a control unit 31, a storage unit 32, a communication unit 33, an input unit 34, and a display unit 35. The above-mentioned units are connected to each other via a bus. The control unit 31 is configured using one or more processors such as a CPU, an MPU, or a GPU. The storage unit 32 includes a RAM or a ROM. The storage unit 32 stores various data and the like required for a control program 32P (program product) executed by the control unit 31. The storage unit 32 temporarily stores data and the like generated when executing the control program 32P. The control unit 31 executes the control program 32P stored in the storage unit 32 as needed to perform information processing and control processing related to the computer 30.

[0036] The communication unit 33 is a communication module that transmits and receives information to and from the server 10 via the network N. The input unit 34 accepts operational input from an administrator of the notification system. The input unit 34 includes, for example, a keyboard, a mouse, and a microphone for voice input. The display unit 35 is a liquid crystal display or an organic EL display. The display unit 35 outputs various information in accordance with instructions from the control unit 31. The input unit 34 and the display unit 35 may be a touch panel configured as an integrated unit.

[0037] The processing of the first embodiment will be described. The bed sensor 21 detects the heart rate, breathing, body movement, signals indicating the state of awakening during sleep, or the depth of sleep of a subject lying on a bed. The bed sensor 21 outputs the detected signals indicating the heart rate, breathing, body movement, the state of awakening during sleep, or the depth of sleep of the subject to the sensor I / F device 20. The sensor I / F device 20 acquires the heart rate, breathing rate, body movement, signals indicating the state of awakening during sleep, or the depth of sleep of the subject output by the bed sensor 21. The sensor I / F device 20 analyzes the signals indicating the heart rate, breathing, body movement, the state of awakening during sleep, or the depth of sleep of the subject output by the bed sensor 21 for a predetermined period (e.g., 10 minutes) to identify the heart rate, breathing rate, amount of body movement, time of awakening during sleep, or quality of sleep. The sensor I / F device 20 transmits the identified heart rate, respiratory rate, amount of body movement, time of midnight awakening, quality of sleep, etc. to the server 10 at predetermined time intervals (for example, every 10 minutes).

[0038] The control unit 11 receives the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, quality of sleep, etc. transmitted from the sensor I / F device 20. The control unit 11 identifies the date and time when the subject's heart rate, respiratory rate, amount of body movement, time of awakening during sleep, quality of sleep, etc. were received. The control unit 11 identifies the time of going into bed, time of getting out of bed, or sleep duration based on the subject's heart rate, respiratory rate, amount of body movement, time of awakening during sleep, quality of sleep, etc. The time of going into bed is identified, for example, when the subject's heart rate, respiratory rate, amount of body movement, time of awakening during sleep, quality of sleep, etc. are first received from the sensor I / F device 20. The time of getting out of bed is identified, for example, when the subject's heart rate, respiratory rate, body movement, time of awakening during sleep, quality of sleep, etc. are last received from the sensor I / F device 20. The sleep duration is identified, for example, using the identified date and time and the time of going into bed. The control unit 11 stores sensor information including date and time, time of getting into bed, time of getting out of bed, time of waking up during sleep, heart rate, respiratory rate, amount of body movement, sleep time, or quality of sleep in the sensor information DB 141.

[0039] The control unit 11 reads out the sensor information stored immediately before (for example, 3:20 on 3 / 2 / 2024) from the sensor information DB 141. The control unit 11 inputs the read sensor information into the learning model 142. The control unit 11 acquires an estimation result based on the probability output by the learning model 142. The control unit 11 determines whether the estimation result is "excretion will occur in 10 minutes." If the control unit 11 determines that the estimation result is not "excretion will occur in 10 minutes," it again receives the heart rate, respiratory rate, amount of body movement, time of awakening mid-sleep, or quality of sleep, etc., transmitted from the sensor I / F device 20 after a predetermined time has elapsed. If the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes," it generates an alert indicating the possibility of excretion by the subject. The control unit 11 transmits the generated alert to the computer 30.

[0040] The control unit 31 receives the alert sent from the server 10. The control unit 31 displays the received alert on the display unit 35. FIG. 7 is an explanatory diagram showing an example of how an alert is displayed. The screen d01 in FIG. 7 includes an alert field d11. The alert field d11 displays an alert generated by the server 10. The alert field d11 in FIG. 7 displays an icon indicating signs of excretion and an alert that includes the message "Excretion is possible."

[0041] In this embodiment, an example has been described in which the learning model 142 outputs a probability indicating whether or not excretion will occur 10 minutes later when sensor information including date and time, time of getting into bed, time of getting out of bed, time of waking up during sleep, heart rate, respiratory rate, amount of body movement, sleep duration, and sleep quality is input, but this is not limiting. The learning model 142 may also output a probability indicating whether or not excretion will occur 10 minutes later when multiple types of sensor information (e.g., 2 to 8 types) arbitrarily selected from the above 9 types of sensor information are input.

[0042] (Variation) In the first embodiment, an example has been described in which sensor information is input to the learning model 142 to output the excretion information of the subject, but this is not limited to this. The notification system may output the excretion information of the subject based on the comparison result between the measurement parameters and the reference parameters. The processing in this case will be described below. When a medical professional confirms the excretion of the subject, the medical professional inputs the date and time when the subject excreted (hereinafter referred to as excretion date and time) via the input unit 34. The control unit 31 acquires the excretion date and time input by the medical professional. The control unit 31 transmits the acquired excretion date and time to the server 10.

[0043] The control unit 11 receives the excretion date and time transmitted from the computer 30. The control unit 11 reads the heart rate, respiration rate, body movement amount, and sleep quality obtained through the bed sensor 21 for a predetermined period (hereinafter referred to as 30 minutes) before the excretion date and time from the sensor information DB 141. The predetermined period can be changed appropriately depending on the embodiment. Specifically, when the excretion date and time is "2024 / 3 / 2 / 3:30," the control unit 11 refers to the sensor information DB 141 and reads the heart rate, respiration rate, body movement amount, and sleep quality from the record storing "3:00," "3:10," "3:20," and "3:30" in the date and time column. Based on the read heart rate, respiration rate, body movement, and sleep quality, the control unit 11 acquires measurement parameters of data related to the heart rate, data related to the respiration rate, data related to the body movement amount, and data related to the sleep quality for the 30 minutes before the excretion date and time of the subject. In the following explanation, measurement parameters of data related to heart rate will be referred to as "heart rate measurement parameters," measurement parameters of data related to respiratory rate will be referred to as "respiratory rate measurement parameters," measurement parameters of data related to body movement will be referred to as "body movement measurement parameters," and measurement parameters of data related to sleep quality will be referred to as "sleep quality measurement parameters."

[0044] The measurement parameters of the heart rate include the average heart rate, the standard deviation of the heart rate, or the maximum / minimum difference of the heart rate. The measurement parameters of the heart rate may be at least one of the average heart rate, the standard deviation of the heart rate, or the maximum / minimum difference of the heart rate. The average heart rate is, for example, the average heart rate for 30 minutes before the date and time of excretion. If the heart rate for the date and time series "3:00" is "54 beats / minute," the heart rate for the date and time series "3:10" is "53 beats / minute," the heart rate for the date and time series "3:20" is "54 beats / minute," and the heart rate for the date and time series "3:30" is "55 beats / minute," the average heart rate is "54 beats / minute." The standard deviation of the heart rate is, for example, the standard deviation of the heart rate for 30 minutes before the date and time of excretion. In the above example, the standard deviation of the heart rate is "0.70...". The maximum and minimum difference in heart rate is, for example, the difference between the maximum and minimum average heart rate in the 30 minutes before the excretion date and time. In the example above, the maximum heart rate is 55 beats / minute and the minimum heart rate is 53 beats / minute. Therefore, the maximum and minimum difference in heart rate is 2 beats / minute.

[0045] The measurement parameters of the respiratory rate include the average respiratory rate, the standard deviation of the respiratory rate, or the maximum-minimum difference of the respiratory rate. The measurement parameters of the respiratory rate may be at least one of the average respiratory rate, the standard deviation of the respiratory rate, or the maximum-minimum difference of the respiratory rate. The average respiratory rate is, for example, the average value of the respiratory rate for 30 minutes before the date and time of excretion. The standard deviation of the respiratory rate is, for example, the standard deviation of the respiratory rate for 30 minutes before the date and time of excretion. The maximum-minimum difference of the respiratory rate is, for example, the difference between the maximum average respiratory rate and the minimum average respiratory rate for 30 minutes before the date and time of excretion. The measurement parameters of the body movement include the average body movement value, the standard deviation of the body movement values, or the maximum-minimum difference of the body movement. The measurement parameters of the body movement may be at least one of the average body movement value, the standard deviation of the body movement values, or the maximum-minimum difference of the body movement. The average body movement is, for example, the average value of the body movement for 30 minutes before the date and time of excretion. The standard deviation of the body movement is, for example, the standard deviation of the body movement for 30 minutes before the date and time of excretion. The maximum-minimum difference in the amount of body movement is, for example, the difference between the maximum amount of body movement and the minimum amount of body movement in the 30 minutes before the excretion date and time.

[0046] The sleep quality measurement parameters include the average sleep quality, the standard deviation of sleep quality, or the maximum / minimum difference of sleep quality. The sleep quality measurement parameters may be at least one of the average sleep quality, the standard deviation of sleep quality, or the maximum / minimum difference of sleep quality. Sleep quality includes REM sleep, light NREM sleep, medium-depth NREM sleep, and deepest NREM sleep. The read sleep quality is converted into a score. For example, REM sleep is converted into "1," light NREM sleep into "2," medium-depth NREM sleep into "3," and deepest NREM sleep into "4." The average sleep quality is, for example, the average value of score-converted sleep quality for 30 minutes before the excretion date and time. The standard deviation of sleep quality is, for example, the standard deviation of score-converted sleep quality for 30 minutes before the excretion date and time. The maximum / minimum difference of sleep quality is, for example, the difference between the maximum and minimum values ​​of score-converted sleep quality for 30 minutes before the excretion date and time.

[0047] The control unit 11 stores the acquired measurement parameters in the mass storage unit 14. The control unit 11 determines whether the number of times that measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is a predetermined number or more (hereinafter referred to as five or more times). The predetermined number can be changed appropriately depending on the embodiment. If the number of times that measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is not five or more, the control unit 11 repeats the above-mentioned process. If the number of times that measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is five or more, the control unit 11 reads out multiple measurement parameters from the mass storage unit 14. The control unit 11 calculates reference parameters based on the average of the multiple read measurement parameters. Specifically, the control unit 11 calculates reference parameters for data related to heart rate obtained by averaging multiple heart rate measurement parameters (hereinafter referred to as heart rate reference parameters) and reference parameters for data related to respiratory rate obtained by averaging multiple respiratory rate measurement parameters (hereinafter referred to as respiratory rate reference parameters). Similarly, the control unit 11 calculates reference parameters for data relating to the amount of body movement (hereinafter referred to as reference parameters for body movement) obtained by averaging multiple measurement parameters for the amount of body movement, and reference parameters for data relating to the quality of sleep (hereinafter referred to as reference parameters for sleep quality) obtained by averaging multiple measurement parameters for the quality of sleep. The control unit 11 stores the acquired reference parameters in the mass storage unit 14.

[0048] The reference parameters for each subject may be updated as appropriate. In this case, the control unit 11 recalculates the reference parameters by averaging the reference parameters and measurement parameters for the same subject acquired at different dates and times. The control unit 11 stores the recalculated reference parameters in the mass storage unit 14.

[0049] When the control unit 11 newly acquires the heart rate, respiration rate, amount of body movement, and quality of sleep through the bed sensor 21, it reads out the heart rate, respiration rate, amount of body movement, and quality of sleep for a predetermined period (e.g., 30 minutes) before the acquisition date and time of the new heart rate, etc. from the sensor information DB 141. The control unit 11 reacquires measurement parameters based on the newly acquired heart rate, respiration rate, amount of body movement, and quality of sleep and the read-out heart rate, respiration rate, amount of body movement, and quality of sleep. Specifically, the control unit 11 calculates the average of the newly acquired heart rate, respiration rate, amount of body movement, and quality of sleep and the read-out heart rate, respiration rate, amount of body movement, and quality of sleep, thereby reacquiring the measurement parameters of the heart rate, respiration rate, amount of body movement, and quality of sleep.

[0050] The control unit 11 may read the heart rate, respiration rate, body movement amount, and sleep quality from the sensor information DB 141 at predetermined intervals (hereinafter referred to as 30-minute intervals). The predetermined intervals can be changed as appropriate depending on the embodiment. In this case, the control unit 11 calculates the average of the heart rate, respiration rate, body movement amount, and sleep quality over the 30-minute intervals to reacquire the measurement parameters of the heart rate, respiration rate, body movement amount, and sleep quality.

[0051] The control unit 11 compares the reacquired measurement parameters with the reference parameters. A method for comparing the measurement parameters with the reference parameters includes, for example, calculating cosine similarity. The following describes the process of calculating the cosine similarity between the reacquired measurement parameters and the reference parameters. The control unit 11 calculates the cosine similarity between the heart rate measurement parameters and the heart rate reference parameters, the cosine similarity between the respiration rate measurement parameters and the respiration rate reference parameters, the cosine similarity between the body movement measurement parameters and the body movement reference parameters, and the cosine similarity between the sleep quality measurement parameters and the sleep quality reference parameters. The four types of cosine similarity are −1 to +1, respectively. The control unit 11 sums the four calculated cosine similarities. The sum of the cosine similarities is −4 to +4. The closer the sum of the cosine similarities is to +4, the higher the similarity between the reacquired parameters and the reference parameters, and the closer the sum of the cosine similarities is to −4, the lower the similarity between the reacquired parameters and the reference parameters.

[0052] The control unit 11 outputs a probability indicating whether or not there will be excretion after a predetermined time (hereinafter referred to as 30 minutes) based on the comparison result (total value of cosine similarity). The probability indicating whether or not there will be excretion after 30 minutes is output as a value between 0 and 1. The probability indicating whether or not there will be excretion after 30 minutes is "0" when the total value of cosine similarity is "-4". Furthermore, the probability indicating whether or not there will be excretion after 30 minutes is "1" when the total value of cosine similarity is "+4". The control unit 11 determines whether or not the probability indicating whether or not there will be excretion after 30 minutes exceeds a predetermined threshold (e.g., 0.75). If the control unit 11 determines that the probability indicating whether or not there will be excretion after 30 minutes does not exceed the predetermined threshold, the control unit 11 ends the processing as is. If the control unit 11 determines that the calculated probability exceeds the predetermined threshold, the control unit 11 generates an alert. The subsequent processing is the same as when the learning model 142 is used, and therefore a description thereof will be omitted.

[0053] In the modified example, the cosine similarities between the four types of reference parameters and the four types of measurement parameters are calculated, and the calculated cosine similarities are summed up, but the present invention is not limited to this. The cosine similarity between one type of reference parameter obtained by combining four types of reference parameters and one type of measurement parameter obtained by combining four types of measurement parameters may be calculated. In this case, the calculated cosine similarity will be between 0 and 1. The control unit 11 can output a probability indicating the presence or absence of excretion 30 minutes later based on the calculated cosine similarity.

[0054] In the modified example, cosine similarity is used as a method for comparing the reacquired measurement parameters with the reference parameters, but this is not limiting. The difference between the reacquired measurement parameters and the reference parameters may also be used. The processing in this case will be described below. The control unit 11 normalizes each value of the reacquired measurement parameters and each value of the reference parameters. Specifically, the control unit 11 normalizes each value of the reacquired heart rate measurement parameter and each value of the reference heart rate parameter by dividing each value of the reacquired heart rate measurement parameter and each value of the reference heart rate parameter by each value of the reference heart rate parameter. In this case, each value of the reference heart rate parameter becomes "1." The control unit 11 also performs similar normalization on each value of the other measurement parameters and each value of the reference parameters. The control unit 11 sums each value of the normalized measurement parameters. Specifically, the control unit 11 sums each value of the normalized heart rate measurement parameter, the normalized respiration rate measurement parameter, the normalized body movement measurement parameter, and the normalized sleep quality measurement parameter. The control unit 11 sums the values ​​of the normalized reference parameters. Specifically, the control unit 11 sums the values ​​of the normalized reference parameter for the heart rate, the normalized reference parameter for the respiration rate, the normalized reference parameter for the amount of body movement, and the normalized reference parameter for the quality of sleep. The control unit 11 calculates the difference between the sum of the normalized measurement parameters and the sum of the normalized reference parameters. If the absolute value of the calculated difference is close to 0, the similarity between the reacquired measurement parameters and the reference parameters is high, and if the absolute value of the calculated difference is not close to 0, the similarity between the reacquired measurement parameters and the reference parameters is low. The control unit 11 can output a probability indicating whether or not excretion will occur 30 minutes later based on the comparison result (the difference between the measurement parameters and the reference parameters).

[0055] The control unit 11 may acquire measurement parameters and reference parameters for each subject. The processing in this case will be described. When a medical professional confirms the excretion of a subject, the medical professional inputs the excretion date and time as well as the subject ID, etc. (see FIG. 14). The control unit 31 acquires the excretion date and time, the subject ID, etc. input by the medical professional. The control unit 31 transmits the acquired excretion date and time, the subject ID, etc. to the server 10. The control unit 11 receives the excretion date and time, the subject ID, etc. transmitted from the computer 30. The control unit 11 reads out the heart rate, respiratory rate, body movement amount, and sleep quality obtained via the bed sensor 21 for 30 minutes before the excretion date and time from the sensor information DB 141 for each subject. The control unit 11 acquires measurement parameters for each subject using the read heart rate, respiratory rate, body movement amount, and sleep quality. The control unit 11 acquires reference parameters for each subject based on the average of multiple measurement parameters. When the control unit 11 newly acquires the heart rate, respiratory rate, amount of body movement, and quality of sleep through the bed sensor 21, it acquires the measurement parameters again based on the acquired heart rate, respiratory rate, amount of body movement, and quality of sleep and the read-out heart rate, respiratory rate, amount of body movement, and quality of sleep. The control unit 11 can output, for each subject, a probability indicating whether or not there will be excretion 30 minutes later based on a comparison result between the reacquired measurement parameters and the reference parameters.

[0056] In the modified example, the probability indicating whether or not excretion will occur after 30 minutes is output using four types of reference parameters and four types of measurement parameters, but this is not limiting. At least two types of reference parameters (e.g., the heart rate reference parameter and the respiratory rate reference parameter) out of the four types of reference parameters and at least two types of measurement parameters (e.g., the heart rate measurement parameter and the respiratory rate measurement parameter) out of the four types of measurement parameters may be used to output the probability indicating whether or not excretion will occur after 30 minutes.

[0057] 8 is a flowchart showing the procedure for generating a learning model. The control unit 11 reads out from the training DB 143 a large amount of training data in which sensor information is associated with information indicating whether or not excretion will occur 10 minutes later (step S101). The control unit 11 uses the read out training data to generate a learning model 142 that takes the sensor information as input and outputs a probability indicating whether or not excretion will occur 10 minutes later (step S102). The control unit 11 stores the generated learning model 142 in the mass storage unit 14 (step S103).

[0058] 9A is a flowchart showing an example of a processing procedure of the notification system. FIG. 9A is a flowchart when the learning model 142 is used. The control unit 11 determines whether or not the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has been received from the bed sensor 21 connected to the sensor I / F device 20 via the sensor I / F device 20 (step S201). If the control unit 11 determines that the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has not been received from the bed sensor 21 (step S201: NO), the control unit 11 waits until the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has been received. If the control unit 11 determines that the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has been received from the bed sensor 21 (step S201: YES), the control unit 11 identifies the date and time when the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep was received (step S202). The control unit 11 identifies the time of getting into bed, the time of getting out of bed, or the sleeping duration based on the subject's heart rate, respiratory rate, amount of body movement, time of waking up during sleep, quality of sleep, etc. (Step S203). The control unit 11 stores the sensor information including the date and time, time of getting into bed, time of getting out of bed, time of waking up during sleep, heart rate, respiratory rate, amount of body movement, sleeping duration, quality of sleep, etc. in the sensor information DB 141 (Step S204).

[0059] The control unit 11 reads out the most recently stored sensor information from the sensor information DB 141 (step S205). The control unit 11 inputs the read sensor information into the learning model 142 (step S206). The control unit 11 acquires an estimation result based on the probability output by the learning model 142 (step S207). The control unit 11 determines whether the estimation result is "excretion will occur in 10 minutes" (step S208). If the control unit 11 determines that the estimation result is not "excretion will occur in 10 minutes" (step S208: NO), the control unit 11 returns the process to step S201. If the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes" (step S208: YES), the control unit 11 generates an alert indicating the possibility of excretion by the subject (step S209). The control unit 11 transmits the generated alert to the computer 30 (step S210).

[0060] The control unit 31 receives an alert transmitted from the server 10 (step S301). The control unit 31 displays the received alert on the display unit 35 (step S302).

[0061] FIG. 9B is a flowchart showing an example of the processing procedure of the notification system. FIG. 9B is a flowchart when measurement parameters and reference parameters are used. In the flowchart shown in FIG. 9B, steps S201 to S208 in the processing shown in FIG. 9A are replaced with steps S401 to S413. Descriptions of steps similar to those in FIG. 9A will be omitted. The control unit 11 determines whether or not the excretion date and time and the subject ID have been received from the computer 30 (step S401). If the control unit 11 determines that the excretion date and time and the subject ID have not been received from the computer 30 (step S401: NO), the control unit 11 waits until they are received. If the control unit 11 determines that the excretion date and time and the subject ID have been received from the computer 30 (step S401: YES), the control unit 11 reads out the heart rate, respiratory rate, amount of body movement, and quality of sleep for a predetermined period before the excretion date and time from the sensor information DB 141 (step S402). The control unit 11 acquires measurement parameters from the read sensor information (step S403). The control unit 11 stores the acquired measurement parameters of each subject in the mass storage unit 14 in association with the subject ID (step S404).

[0062] The control unit 11 determines whether the number of times that the measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is equal to or greater than a predetermined number (step S405). If the number of times that the measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is not equal to or greater than the predetermined number (step S405: NO), the control unit 11 returns the process to step S401. If the number of times that the measurement parameters acquired at different dates and times have been stored in the mass storage unit 14 is equal to or greater than the predetermined number (step S405: YES), the control unit 11 reads out multiple measurement parameters from the mass storage unit 14 (step S406). The control unit 11 calculates reference parameters based on the average of the multiple read measurement parameters (step S407). The control unit 11 associates the acquired reference parameters of each subject with the subject ID and stores them in the mass storage unit 14 (step S408).

[0063] The control unit 11 determines whether the heart rate, respiratory rate, body movement, and sleep quality have been newly acquired through the bed sensor (step S409). If the control unit 11 determines that the heart rate, respiratory rate, body movement, and sleep quality have not been newly acquired (step S409: NO), it waits until they are acquired. If the control unit 11 determines that the heart rate, respiratory rate, body movement, and sleep quality have been newly acquired (step S409: YES), it reads out the heart rate, respiratory rate, body movement, and sleep quality for a predetermined period before the acquisition date and time of the new heart rate, etc. from the sensor information DB 141 (step S410). The control unit 11 reacquires measurement parameters based on the newly acquired heart rate, respiratory rate, body movement, and sleep quality and the read-out heart rate, respiratory rate, body movement, and sleep quality (step S411). The control unit 11 outputs a probability indicating whether or not excretion will occur 30 minutes later based on the reacquired measurement parameters and the comparison result with the reference parameters (step S412). The control unit 11 determines whether the probability indicating whether or not there will be excretion after 30 minutes exceeds a predetermined threshold (step S413). If the control unit 11 determines that the probability indicating whether or not there will be excretion after 30 minutes does not exceed the predetermined threshold (step S413: NO), the control unit 11 ends the process. If the control unit 11 determines that the calculated probability exceeds the predetermined threshold (step S413: YES), the control unit 11 proceeds to step S209.

[0064] According to the first embodiment, the notification system can acquire sensor information obtained through the bed sensor and output excretion information of the subject based on the acquired sensor information.

[0065] According to the first embodiment, the notification system can output excretion information of the subject by inputting the acquired sensor information into a learning model that outputs excretion information related to the excretion of the subject when sensor information is input.

[0066] According to the first embodiment, the notification system can output the excretion information of the subject based on the result of comparing the reacquired measurement parameters with the reference parameters.

[0067] According to the first embodiment, the notification system can output the excretion information of the subject to the administrator of the notification system.

[0068] (Embodiment 2) In the second embodiment, a method is described in which a notification system retrains a learning model using training data including information on the excretion status of a subject and sensor information before excretion. The information on the excretion status of a subject is, for example, the date and time of excretion, or whether or not the subject has excreted. In the second embodiment, the information on the excretion status of a subject is described as the date and time of excretion.

[0069] Fig. 10 is an explanatory diagram showing an overview of a notification system according to a second embodiment. In addition to a bed sensor 21, an excretion sensor 22 is connected to the sensor I / F device 20 in Fig. 10 by wire or wirelessly. The excretion sensor 22 is a sensor installed in the diaper of the subject, and detects whether or not the subject has excreted. The excretion sensor 22 detects whether or not the subject has excreted by measuring, for example, the temperature and humidity inside the diaper of the subject.

[0070] The processing of the second embodiment will be described. When the detected temperature and humidity are not within a predetermined range, the excretion sensor 22 outputs a detection signal indicating that the subject has excreted to the sensor I / F device 20. The sensor I / F device 20 analyzes the detection signal output by the excretion sensor 22 to identify the excretion date and time. The sensor I / F device 20 transmits the identified excretion date and time to the server 10.

[0071] The control unit 11 receives the excretion date and time transmitted from the sensor I / F device 20. The control unit 11 identifies sensor information from the sensor information DB 141 several minutes before (e.g., about 10 minutes before) the received excretion date and time. For example, if the excretion date and time is "2024 / 3 / 2 / 3:18", the control unit 11 reads out sensor information in which "2024 / 3 / 2 / 3:10" is stored in the date and time column from the sensor information DB 141. The control unit 11 stores training data including the read sensor information and the excretion date and time in the mass storage unit 14. The control unit 11 repeatedly performs the above-mentioned process for a predetermined period (e.g., six months). After the predetermined period has elapsed, the control unit 11 reads out a large amount of training data from the mass storage unit 14. The control unit 11 re-trains the learning model 142 using the read training data. The control unit 11 stores the re-trained learning model 142 in the mass storage unit 14.

[0072] In the second embodiment, the control unit 11 acquires the excretion date and time using the excretion sensor 22, but this is not limiting. The control unit 11 may acquire the excretion date and time manually input by the caregiver of the subject or upon request from the subject himself / herself. In this case, the control unit 11 identifies, from the sensor information DB 141, the sensor information that is closest to approximately 10 minutes before the acquired excretion date and time. The control unit 11 stores the identified sensor information and training data including the excretion date and time in the mass storage unit 14.

[0073] 11 is a flowchart showing an example of a processing procedure for relearning. The control unit 11 determines whether or not the excretion date and time has been received from the sensor I / F device 20 connected to the excretion sensor 22 (step S501). When the control unit 11 determines that the excretion date and time has not been received from the sensor I / F device 20 (step S501: NO), the control unit 11 waits until the excretion date and time is received. When the control unit 11 determines that the excretion date and time has been received from the sensor I / F device 20 (step S501: YES), the control unit 11 reads out sensor information from the sensor information DB 141 that is several minutes before the received excretion date and time (step S502). The control unit 11 then The control unit 11 stores training data including the sensor information and the excretion date and time in the mass storage unit 14 (step S503). The control unit 11 determines whether a predetermined period has elapsed (step S504). If the control unit 11 determines that the predetermined period has not elapsed (step S504: NO), the control unit 11 returns the process to step S501. If the control unit 11 determines that the predetermined period has elapsed (step S504: YES), the control unit 11 reads out a large amount of training data from the mass storage unit 14 (step S505). The control unit 11 retrains the learning model 142 using the read training data (step S506). The control unit 11 stores the retrained learning model 142 in the mass storage unit 14 (step S507).

[0074] According to the second embodiment, the notification system can re-train the learning model and update the learning model.

[0075] (Embodiment 3) In the third embodiment, a method is described in which the notification system generates an alert when the predicted excretion time based on the excretion information is included in the time period related to the excretion of the subject.

[0076] Fig. 12 is an explanatory diagram showing an overview of a notification system according to a third embodiment. The notification system includes a plurality of sensor I / F devices 20. A server 10, the plurality of sensor I / F devices 20, and a computer 30 are connected via a network N. In this embodiment, a sensor I / F device 20, a bed sensor 21, and an excretion sensor 22 are provided for each subject. In Fig. 12, each subject (subject X and subject Y) is associated with the sensor I / F device 20, the bed sensor 21, and the excretion sensor 22 via a sensor ID or the like.

[0077] 13 is a block diagram showing an example of the configuration of a server according to embodiment 3. The mass storage unit 14 includes a subject DB 144, an excretion date and time DB 145, and an excretion time zone DB 146. The subject DB 144 stores information about the subject. The excretion date and time DB 145 stores the excretion date and time of the subject. The excretion time zone DB 146 stores the time zone related to the excretion of the subject.

[0078] FIG. 14 is an explanatory diagram showing an example of a record layout of a subject DB. The subject DB 144 includes a subject ID column, a sensor ID column, a name column, an age column, a medical history column, a gender column, and a room number column. The subject ID column stores a subject ID for identifying the subject. The sensor ID column stores a sensor ID for identifying each sensor associated with the subject. The name column stores the name of the subject. The age column stores the age of the subject. The medical history column stores the medical history of the subject. Examples of medical history include "cerebrovascular disease," "heart disease," "fracture / fall," "cancer," "respiratory disease," "joint disease," "diabetes," "dementia," "Parkinson's disease," "spinal cord injury," "visual / hearing impairment," and "speech disorder." If the subject does not have any of the above medical histories, "none" is stored in the medical history column. The gender column stores the gender of the subject. The room number column stores the room number in which the subject is located. 14, information corresponding to each of subject ID "T001" (subject X) and subject ID "T002" (subject Y) is stored. In the record of subject ID T001 in FIG. 14, sensor ID "S001", name "X", age "65", medical history "dementia", gender "male", and room number "310" are stored.

[0079] FIG. 15 is an explanatory diagram showing an example of a record layout of an excretion date and time DB. The excretion date and time DB 145 includes a sensor ID column and an excretion date and time column. The sensor ID column stores the same sensor ID as the subject DB 144. The excretion date and time column stores the excretion date and time of the subject corresponding to the sensor ID. In FIG. 15, multiple excretion dates and times corresponding to the sensor ID "S001" (subject X) and the sensor ID "S002" (subject Y) are stored. Multiple records in FIG. 15 with the sensor ID S001 store multiple excretion dates and times including the excretion date and time "2024 / 3 / 1 / 2:24".

[0080] FIG. 16 is an explanatory diagram showing an example of the record layout of an excretion time period DB. The excretion time period DB 146 includes a subject ID column and an excretion time period column. The subject ID column stores the same subject ID as in the subject DB 144. The excretion time period column stores a time period related to the excretion of the subject (hereinafter referred to as an excretion time period). The excretion time period is, for example, a time period during which excretion by a sleeping subject frequently occurs. In FIG. 16, excretion time periods corresponding to subject ID "T001" (subject X) and subject ID "T002" (subject Y) are stored. In the record of subject ID "T001" in FIG. 16, the excretion time period "2:00-4:00" is stored.

[0081] A method for identifying an excretion time period will be described. The control unit 11 receives the sensor ID and excretion date and time transmitted by the sensor I / F device 20. The control unit 11 stores the received sensor ID and excretion date and time in the excretion date and time DB 145. The control unit 11 repeats the above-mentioned process for a predetermined period (for example, six months). The control unit 11 reads out a large number of excretion dates and times stored in the mass storage unit 14 for each sensor ID. The control unit 11 identifies the subject ID (subject) corresponding to the read sensor ID from the subject DB 144. The control unit 11 identifies to which time segment of a predetermined time unit each of the read out excretion dates and times belongs. The time segments in predetermined time units are, for example, time segments in one-hour increments from bedtime (10 p.m.) to wake-up time (7 a.m.), such as "22:00 to 23:00," "23:00 to 24:00," ..., "5:00 to 6:00," and "6:00 to 7:00." The excretion date and time "2:24" belongs to the time segment "2:00 to 3:00." The control unit 11 counts the number of excretion dates and times that belong to each of the above time segments. The control unit 11 identifies a time segment whose counted number is a predetermined number (e.g., three) or more as an excretion time slot. Specifically, the control unit 11 identifies the excretion time slot "2:00 to 4:00" when the many excretion dates and times that it has read out include "2:11," "2:24," "2:49," "3:01," "3:17," and "3:59," etc. The control unit 11 repeatedly performs the same process to identify the excretion time period for each subject ID (subject). The control unit 11 stores the identified excretion time period in the excretion time period DB 146 in association with the subject ID.

[0082] The control unit 11 may identify a time segment whose total number is at or above a predetermined rank (for example, second place) as the excretion time slot. Specifically, the control unit 11 identifies the excretion time slot "2:00-4:00" when the total number of the time segments "2:00-3:00" and "3:00-4:00" is "9" and the total number of the other time segments is less than "9".

[0083] The processing of the third embodiment will be described. The control unit 11 inputs the sensor information read from the sensor information DB 141 to the learning model 142. The control unit 11 acquires an estimation result based on the probability output by the learning model 142. The control unit 11 determines whether the estimation result is "excretion will occur in 10 minutes." If the control unit 11 determines that the estimation result is not "excretion will occur in 10 minutes," it again receives the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep transmitted from the sensor I / F device 20 after a predetermined time has elapsed. If the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes," it identifies the predicted excretion time of the subject. Specifically, if the control unit 11 acquires the estimation result "excretion will occur in 10 minutes" from sensor information that includes "2024 / 3 / 2 / 2:10" in the date and time column, it identifies the predicted excretion time as "2:20." The control unit 11 identifies the subject ID corresponding to the sensor ID received from the sensor I / F device 20 from the subject DB 144. Specifically, when the sensor ID received from the sensor I / F device 20 is "S001", the control unit 11 identifies the subject ID "T001" from the subject DB 144. The control unit 11 reads out the excretion time period corresponding to the identified subject ID from the excretion time period DB 146. Specifically, the control unit 11 reads out the excretion time period "2:00-4:00" corresponding to the subject ID "T001" from the excretion time period DB 146. The control unit 11 determines whether the identified predicted excretion time "2:20" is included in the excretion time period "2:00-4:00". If the predicted excretion time is not included in the excretion time period, the control unit 11 ends the process as is. If the predicted excretion time is included in the excretion time period, the control unit 11 generates an alert indicating the possibility of the subject excreting. The subsequent processing is the same as in the first embodiment, and therefore a description thereof will be omitted.

[0084] (Variation 1) In the third embodiment, an example has been described in which all sensor information acquired by the control unit 11 is input to the learning model 142, but this is not limiting. The control unit 11 may input only sensor information included in the excretion time slot to the learning model 142. The processing in this case will be described below. The control unit 11 stores sensor information including date and time, time of getting into bed, time of getting out of bed, time of awakening during the night, heart rate, respiratory rate, amount of body movement, sleep time, or sleep quality in the sensor information DB 141. The control unit 11 identifies a subject ID corresponding to the sensor ID received from the sensor I / F device 20 from the subject DB 144. The control unit 11 reads out an excretion time slot corresponding to the identified subject ID from the excretion time slot DB 146. The control unit 11 refers to the sensor information DB 141 and determines whether the date and time of the sensor information most recently stored is included in the excretion time slot. If the control unit 11 determines that the date and time of the sensor information most recently stored is not included in the excretion time slot, it ends the processing. When the control unit 11 determines that the date and time of the most recently stored sensor information is included in the excretion time period, it reads out the sensor information from the sensor information DB 141. The control unit 11 inputs the read sensor information into the learning model. The subsequent processing is the same as in the first embodiment, and therefore a description thereof will be omitted. By inputting only the sensor information included in the excretion time period into the learning model 142, the control unit 11 can efficiently predict the possibility that the subject will excrete.

[0085] (Variation 2) When the control unit 11 inputs sensor information included in the excretion time period to the learning model 142, the control unit 11 may change the method for obtaining the estimation result. The processing in this case will be described below. An example in which the learning model 142 outputs an excretion score (0 to 1) will be described below. When the control unit 11 inputs sensor information included in the excretion time period to the learning model 142, the learning model 142 outputs an excretion score (0.7). The control unit 11 corrects the excretion score output by the learning model 142. Specifically, the control unit 11 adds a predetermined value (e.g., 0.2) to the excretion score output by the learning model 142. When the corrected excretion score (0.9) is equal to or greater than a predetermined threshold (e.g., 0.75), the control unit 11 obtains "excretion will occur in 10 minutes" as the estimation result. The subsequent processing is the same as in the first embodiment, and therefore description thereof will be omitted.

[0086] Instead of correcting the excretion score, the control unit 11 may correct a predetermined threshold used for comparing the excretion scores. Specifically, the control unit 11 lowers the predetermined threshold (e.g., 0.75) used for comparing the excretion scores to "0.65." If the excretion score (0.7) output by the learning model is equal to or greater than the corrected predetermined threshold (e.g., 0.65), the control unit 11 acquires "excretion will occur in 10 minutes" as the estimated result. The subsequent processing is the same as in the first embodiment, and therefore a description thereof will be omitted. The control unit 11 can alert medical personnel caring for the subject during the excretion period by changing the method for acquiring the estimated result by correcting the excretion score or the predetermined threshold used for comparing the excretion scores.

[0087] 17 is a flowchart showing an example of a processing procedure related to an excretion time period. The control unit 11 determines whether or not a sensor ID and an excretion date and time have been received from the sensor I / F device 20 connected to the excretion sensor 22 (step S601). If the control unit 11 determines that the sensor ID and the excretion date and time have not been received from the sensor I / F device 20 (step S601: NO), the control unit 11 waits until they are received. If the control unit 11 determines that the excretion date and time have been received from the sensor I / F device 20 (step S601: YES), the control unit 11 stores the sensor ID and the excretion date and time in the excretion date and time DB 145 (step S602). The control unit 11 determines whether or not a predetermined period has elapsed (step S603). If the control unit 11 determines that the predetermined period has not elapsed (step S603: NO), the control unit 11 returns to the process at step S601. When the control unit 11 determines that the predetermined period has elapsed (step S603: YES), it reads out a large number of excretion dates and times stored in the excretion date and time DB 145 for each sensor ID (step S604). The control unit 11 identifies the subject ID corresponding to the read sensor ID (step S605). The control unit 11 identifies the one-hour time segment to which the read excretion date and time belongs (step S606). The control unit 11 repeats step S606 for a large number of excretion dates and times. The control unit 11 tallies the number of excretion dates and times belonging to the one-hour time segment (step S607). The control unit 11 identifies the time segment in which the tallied number occurs a predetermined number of times (e.g., three times) or more as an excretion time slot (step S608). The control unit 11 stores the identified excretion time slot in the excretion time slot DB 146 in association with the subject ID (step S609). The control unit 11 determines whether the above-described processing has been completed for all target person IDs (step S610). If the control unit 11 determines that the above-described processing has not been completed for all target person IDs (step S610: NO), the control unit 11 returns the processing to step S604. If the control unit 11 determines that the above-described processing has been completed for all target person IDs (step S610: YES), the control unit 11 ends the processing.

[0088] FIG. 18 is a flowchart showing an example of a processing procedure of the notification system according to the third embodiment. In the flowchart shown in FIG. 18, steps S701 to S704 are added between step S208 and step S209 in the processing shown in FIG. 9A. Descriptions of steps similar to those in FIG. 9A will be omitted. When the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes" (step S208: YES), it identifies the predicted excretion time of the subject (step S701). The control unit 11 identifies the subject ID corresponding to the sensor ID received from the sensor I / F device 20 from the subject DB 144 (step S702). The control unit 11 reads out the excretion time period corresponding to the identified subject ID from the excretion time period DB 146 (step S703). The control unit 11 determines whether the identified predicted excretion time is included in the excretion time period (step S704). When the predicted excretion time is not included in the excretion time period (step S704: NO), the control unit 11 ends the processing. If the predicted excretion time is included in the excretion time slot (step S704: YES), the control unit 11 advances the process to step S209.

[0089] In the third embodiment, when the predicted excretion time is included in the excretion time period, the notification system: Can output alerts.

[0090] According to the third embodiment, the notification system can output an alert when the possibility of the subject having an excretion is particularly high.

[0091] (Embodiment 4) In the fourth embodiment, the notification system acquires additional data and outputs excretion information based on the acquired additional data and sensor information. The additional data includes the subject's age or medical history.

[0092] FIG. 19 is an explanatory diagram showing an example of a record layout of a sensor information DB according to the fourth embodiment. The sensor information DB 141 stores sensor information obtained through each bed sensor 21 provided in association with a plurality of subjects. The sensor information DB 141 includes a first table 141a and a second table 141b. The first table 141a and the second table 141b include a subject ID, a date and time column, a bed entry time column, a bed exit time column, a mid-awakening time column, a heart rate column, a respiratory rate column, a body movement amount column, a sleep duration column, and a sleep quality column. The data stored in the subject ID column is the same as that in the subject DB 144. The data stored from the date and time column to the sleep quality column is the same as that in the sensor information DB 141. The first table 141a in FIG. 19 stores sensor information obtained through the bed sensor 21 of a subject ID "T001" (subject X). The second table 141b in FIG. 19 stores sensor information obtained through the bed sensor 21 of the subject ID "T002" (subject Y).

[0093] FIG. 20 is an explanatory diagram of a learning model according to the fourth embodiment. The learning model 142 is a trained model that has been trained to output excretion information related to the excretion of a subject when additional data and sensor information are input. For example, a CNN is used for the learning model 142. The learning model 142 includes an input layer that receives sensor information, an intermediate layer that performs calculations, and an output layer that outputs excretion information related to the excretion of the subject. The excretion information related to the excretion of the subject is, for example, a probability indicating whether or not excretion will occur 10 minutes later, or the time until excretion. In this embodiment, the excretion information will be described as a probability indicating whether or not excretion will occur 10 minutes later. When additional data and sensor information are input, the learning model 142 outputs a probability indicating whether or not excretion will occur 10 minutes later. The probability output from the output layer is, for example, a value between 0 and 1. The sum of the probabilities output from the output layer is 1.0 (100%).

[0094] In FIG. 20, when the control unit 11 inputs additional data (e.g., the age and medical history of subject X) and sensor information at a certain point in time (e.g., 2024 / 3 / 2 / 3:20) into the learning model 142, the learning model 142 outputs "excretion after 10 minutes: 0.8" and "no excretion after 10 minutes: 0.2" as probabilities indicating whether or not excretion will occur after 10 minutes. The control unit 11 acquires the highest probability value from the multiple probability values ​​output from the output layer as the estimation result. In FIG. 20, the control unit 11 acquires "excretion after 10 minutes" as the estimation result.

[0095] FIG. 21 is an explanatory diagram showing an example of a record layout of a training DB according to the fourth embodiment. The training DB 143 includes a plurality of sheets 1431. The sheet 1431 stores, for each subject, training data used to generate the learning model 142 of FIG. 20. The sheet 1431 of FIG. 21 stores training data acquired from subject X. The sheet 1431 includes a sensor information sequence, an excretion sequence, and an additional data column R1. The information stored in the sensor information sequence and the excretion sequence is the same as in FIG. 5. The additional data column R1 stores additional data of the subject whose sensor information has been acquired. The additional data column R1 of FIG. 21 stores the age and medical history of subject X. In the example of Figure 21, additional data (age ``65 years old'' and medical history ``dementia''), sensor information (date and time ``2023 / 8 / 2 / 2:30'', time of going to bed ``22:00'', time of getting out of bed ``-'', time of waking up during the night ``1:02-1:05'', heart rate ``56 beats / minute'', respiratory rate ``18 breaths / minute'', amount of body movement ``5 times'', sleep time ``4:30'' and sleep quality ``REM sleep''), and information indicating whether or not there was an excretion 10 minutes later (``0'': no ​​excretion after 10 minutes) are stored in association with each other.

[0096] A method for generating the learning model 142 will be described. The control unit 11 reads out from the training DB 143 a large amount of training data in which additional data, sensor information, and information indicating whether or not excretion has occurred 10 minutes later are associated with each other. When the additional training data and sensor information are input to the learning model 142, the control unit 11 trains the learning model 142 so that, when the additional training data and sensor information are input to the learning model 142, the output value from the node corresponding to the correct value included in the training data approaches 1 and the output values ​​from the other output nodes approach 0. The control unit 11 optimizes parameters used in the calculation process in the intermediate layer so that the output value from the output layer of the learning model 142 approaches the correct value. The parameters are, for example, the weight (coupling coefficient) of the node or the coefficient of the activation function used in each node. The parameter optimization method is, for example, the backpropagation method or the steepest descent method. The control unit 11 stores the learning model 142 generated by the above process in the mass storage unit 14.

[0097] The learning model 142 may be generated by an external computer (not shown) or the like. In this case, the generated learning model 142 is deployed to the server 10 from the external computer or the like via the network N.

[0098] The learning model 142 may use algorithms such as, for example, LSTM (Long Short Term Memory), RNN (Recurrent Neural Network), GPT (Generative Pre-trained Transformer), SVM (Support Vector Machine), BERT (Bidirectional Encoder Representations from Transformers), Seq2Seq (Sequence to Sequence), Random Forest, or Decision Tree.

[0099] The processing of the fourth embodiment will be described. The sensor I / F device 20 identifies the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep. The sensor I / F device 20 transmits the sensor ID, heart rate, respiratory rate, amount of body movement, time of awakening during sleep, quality of sleep, etc. to the server 10 at predetermined time intervals (for example, every 10 minutes).

[0100] The control unit 11 receives the sensor ID, heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep transmitted from the sensor I / F device 20. The control unit 11 identifies the subject ID corresponding to the received sensor ID from the subject DB 144. The control unit 11 identifies the date and time when the subject's heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep was received. The control unit 11 identifies the time of getting into bed, time of getting out of bed, time of awakening during sleep, amount of body movement, or duration of sleep. The control unit 11 stores the sensor information including the date and time, time of getting into bed, time of getting out of bed, time of awakening during sleep, heart rate, respiratory rate, amount of body movement, duration of sleep, or quality of sleep in association with the subject ID in the sensor information DB 141. Specifically, when the subject ID is "T001", the control unit 11 stores the sensor information in the first table 141a.

[0101] The control unit 11 reads the most recently stored sensor information from the sensor information DB 141. The control unit 11 reads additional data including the age or medical history of the subject corresponding to the subject ID from the subject DB 144. Specifically, when the subject ID is "T001", the control unit 11 reads additional data including the age "65" or the medical history "dementia" from the subject DB 144. The control unit 11 inputs the read additional data and the sensor information into the learning model 142. The control unit 11 acquires an estimation result based on the probability output by the learning model 142. The control unit 11 determines whether the estimation result is "excretion will occur in 10 minutes". If the control unit 11 determines that the estimation result is not "excretion will occur in 10 minutes", the control unit 11 again receives the heart rate, respiratory rate, amount of body movement, time of awakening mid-sleep, quality of sleep, etc. transmitted from the sensor I / F device 20 after a predetermined time has elapsed. When the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes," it reads out information about the subject corresponding to the subject ID from the subject DB 144. Specifically, when the subject ID is "T001," the control unit 11 reads out information about the subject including the name "X," the gender "male," and the room number "310" from the subject DB 144. The control unit 11 generates an alert indicating the possibility of the subject excreting. The control unit 11 transmits the generated alert and the subject information to the computer 30.

[0102] The control unit 31 receives the alert and information about the target person sent from the server 10. The control unit 31 displays the received alert and information about the target person on the display unit 35. FIG. 22 is an explanatory diagram showing an example of the display of an alert according to the fourth embodiment. The screen d01 in FIG. 22 includes an alert column d11 and a target person column d12. The display format of the alert column d11 is the same as in FIG. 7. The target person column d12 displays information about the target person, "Mr. X (male), Room 310."

[0103] In this embodiment, an example has been described in which the learning model 142 outputs a probability indicating whether or not excretion will occur 10 minutes later when additional data including the subject's age or medical history and sensor information are input, but this is not limiting. The learning model 142 may output a probability indicating whether or not excretion will occur 10 minutes later when one type of additional data selected from the above two types of additional data (the subject's age or medical history) and sensor information are input. The additional data may also include the subject's gender.

[0104] 23 is a flowchart showing the procedure for generating a learning model according to the fourth embodiment. The control unit 11 reads out from the training DB 143 a large amount of training data in which additional data, sensor information, and information indicating whether or not excretion will occur 10 minutes later are associated with each other (step S801). The control unit 11 uses the read-out training data to generate a learning model 142 that receives the additional data and the sensor information as input and outputs a probability indicating whether or not excretion will occur 10 minutes later (step S802). The control unit 11 stores the generated learning model 142 in the mass storage unit 14 (step S803).

[0105] 24 is a flowchart showing an example of a processing procedure of the notification system according to the fourth embodiment. The control unit 11 determines whether or not a sensor ID, heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has been received from the bed sensor 21 connected to the sensor I / F device 20 via the sensor I / F device 20 (step S901). If the control unit 11 determines that the sensor ID, heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has not been received from the bed sensor 21 (step S901: NO), the control unit 11 waits until the information is received. If the control unit 11 determines that the sensor ID, heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep has been received from the bed sensor 21 (step S901: YES), the control unit 11 identifies the subject ID corresponding to the received sensor ID from the subject DB 144 (step S902). The control unit 11 identifies the date and time when the heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep was received (step S903). The control unit 11 identifies the time of getting into bed, time of getting out of bed, time of awakening during sleep, amount of body movement, or sleeping duration based on the subject's heart rate, respiratory rate, amount of body movement, time of awakening during sleep, or quality of sleep (step S904). The control unit 11 stores the sensor information including the date and time, time of getting into bed, time of getting out of bed, time of awakening during sleep, heart rate, respiratory rate, amount of body movement, sleeping duration, or quality of sleep in the sensor information DB 141 (step S905).

[0106] The control unit 11 reads out the most recently stored sensor information from the sensor information DB 141 (step S906). The control unit 11 reads out additional data including the age and medical history of the subject corresponding to the subject ID from the subject DB 144 (step S907). The control unit 11 inputs the read out additional data and the sensor information into the learning model 142 (step S908). The control unit 11 acquires an estimation result based on the probability output by the learning model 142 (step S909). The control unit 11 determines whether the estimation result is "excretion will occur in 10 minutes" (step S910). If the control unit 11 determines that the estimation result is not "excretion will occur in 10 minutes" (step S910: NO), the control unit 11 returns the process to step S901. When the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes" (step S910: YES), it reads out information about the subject corresponding to the subject ID from the subject DB 144 (step S911). The control unit 11 generates an alert indicating the possibility of excretion by the subject (step S912). The control unit 11 transmits the generated alert and information about the subject to the computer 30 (step S913).

[0107] The control unit 31 receives the alert and the information of the target person transmitted from the server 10 (step S1001). The control unit 31 displays the received alert and the information of the target person on the display unit 35 (step S1002).

[0108] According to the fourth embodiment, the notification system can acquire additional data and sensor information, and output excretion information of the subject based on the acquired additional data and sensor information.

[0109] According to the fourth embodiment, the notification system can output an alert based on the excretion information of the subject and information about the subject.

[0110] (Embodiment 5) In the fifth embodiment, a method will be described in which the notification system outputs a message indicating what to do next based on information about the subject's independence level and excretion information about the subject.

[0111] FIG. 25 is an explanatory diagram showing an example of a record layout of a subject DB according to the fifth embodiment. The subject DB 144 includes an independence degree column. The independence degree column stores information about the subject's independence degree. The information about the subject's independence degree is, for example, the Katz Index, the Barthel Index, or the DASC-21. In this embodiment, the information about the subject's independence degree will be described using the Katz Index. The independence degree column classifies the subject's independence degree into seven levels from A to G. The subject's highest independence degree is "A," and the subject's lowest independence degree is "G." The record with subject ID T001 in FIG. 25 stores sensor ID "S001," name "X," age "65," medical history "dementia," gender "male," room number "310," and independence degree "A."

[0112] The mass storage unit 14 also stores messages (hereinafter referred to as content messages) indicating what to do next, corresponding to each of the subject's independence levels (A to G). The content messages are, for example, messages intended for medical professionals caring for the subject. The content message for an independence level of "A" is "Please wake up the subject and have them use the toilet." The content message for an independence level of "B" is "Please wake up the subject and have them use the toilet. Please make sure the subject has returned to bed after a specified time has passed." The content message for an independence level of "C" is "Please wake up the subject and have them use the toilet. Please accompany them from the bed to the toilet." The content message for an independence level of "D" is "Please wake up the subject and have them use the toilet. Please accompany them from the bed to the toilet. Also, please assist the subject when getting up from the bed." If the independence level is "E", the content message is "Please wake the person up to use the toilet. Please accompany the person from the bed to the toilet. Please also assist the person when getting out of bed and when using the toilet." If the independence level is "F", the content message is "Please wake the person up to use the toilet. Please accompany the person from the bed to the toilet. Please also assist the person when getting out of bed, when using the toilet, and when transferring from the wheelchair." If the independence level is "G", the content message is "Please wake the person up to use the toilet. Please accompany the person from the bed to the toilet. Please also assist the person when getting out of bed, when using the toilet, and when transferring from the wheelchair. In addition, two caregivers will be required."

[0113] The processing of the fifth embodiment will be described. When the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes" using the same procedure as in the fourth embodiment, it identifies information about the subject corresponding to the subject ID (e.g., name, gender, and room number) and the subject's independence level from the subject DB 144. Specifically, when the subject ID is "T002," the control unit 11 identifies the subject's information, "70 years old" and "none," and the subject's independence level, "D," from the subject DB 144. The control unit 11 reads a content message corresponding to the identified subject's independence level from the mass storage unit 14. Specifically, when the subject's independence level is "D," the control unit 11 reads a content message from the mass storage unit 14, "Please wake the subject up and allow them to excrete. Please accompany the subject from the bed to the toilet. Also, please assist the subject when getting up from the bed." The control unit 11 generates an alert indicating the possibility of the subject excreting. The control unit 11 transmits the generated alert, the subject's information, and the content message to the computer 30.

[0114] The control unit 31 receives the alert, the subject's information, and the content message sent from the server 10. The control unit 31 displays the received alert, the subject's information, and the content message on the display unit 35. Figure 26 is an explanatory diagram showing an example of the display of an alert according to the fifth embodiment. The screen d01 in Figure 26 includes an alert field d11, a subject field d12, and a content message field d13. The display format of the alert field d11 and the subject field d12 is the same as in Figure 22. The content message field d13 displays the content message. The content message field d13 in Figure 26 displays, "Please wake up the subject and allow them to defecate. Please accompany the subject from the bed to the toilet. Also, please assist the subject when getting up from the bed."

[0115] FIG. 27 is a flowchart showing an example of a processing procedure of the notification system according to the fifth embodiment. In the flowchart shown in FIG. 27, step S910 and subsequent steps in the processing shown in FIG. 24 are replaced with steps S1101 to S1104 and steps S1201 to S1202. The same steps as in FIG. 24 will not be described again. When the control unit 11 determines that the estimation result is "excretion will occur in 10 minutes" (step S910: YES), the control unit 11 identifies information about the subject corresponding to the subject ID (e.g., name, gender, and room number) and the subject's independence level from the subject DB 144 (step S1101). The control unit 11 reads a content message corresponding to the identified subject's independence level from the mass storage unit 14 (step S1102). The control unit 11 generates an alert indicating the possibility of the subject excreting (step S1103). The control unit 11 transmits the generated alert, the subject's information, and the content message to the computer 30 (step S1104).

[0116] The control unit 31 receives the alert, the target person's information, and the content message transmitted from the server 10 (step S1201). The control unit 31 displays the received alert, the target person's information, and the content message on the display unit 35 (step S1202).

[0117] From the fifth embodiment, the notification system can output a content message based on information on the subject's independence level and the subject's excretion information.

[0118] From the fifth embodiment, the notification system can change the content of the message to be output in accordance with the subject's level of independence.

[0119] The features described in each of the above embodiments can be combined with each other. Furthermore, the independent claims and dependent claims described in the claims can be combined with each other in any and all combinations, regardless of the reference format. Furthermore, the claims use a format in which a claim references two or more other claims (multiple claim format), but this is not limited to this. Multiple claims (multi-multi claims) that reference at least one other multiple claim may also be used.

[0120] The embodiments disclosed herein are illustrative in all respects and should not be considered limiting. The scope of the present invention is defined by the claims, not by the above meaning, and is intended to include all modifications within the meaning and scope of the claims. [Explanation of symbols]

[0121] 10 Information processing device (server) 11 Control section 12 Storage section 12P control program 13 Communications Department 14 Mass storage 141 Sensor Information DB 141a Table 1 141b Table 2 142 Learning Model 143 Training DB 1431 seats 144 Target Person DB 145 Excretion date and time DB 146 Excretion Time Period DB 15 Reading unit 1a Portable storage media 20 Sensor I / F device 21 Bed Sensor 22 Excretion sensor 30 Information processing equipment (computer) 31 Control Unit 32 Storage section 32P control program 33 Communications Department 34 Input section 35 Display section

Claims

1. acquiring sensor information obtained through a bed sensor provided in relation to the subject; Outputting excretion information of the subject based on the acquired sensor information A program that causes a computer to perform a process.

2. When sensor information is input, the acquired sensor information is input to a learning model that outputs excretion information regarding the excretion of a subject, and the excretion information of the subject is output. The program according to claim 1.

3. The sensor information includes time of getting into bed, time of getting out of bed, time of waking up during sleep, heart rate, respiratory rate, amount of body movement, sleep duration, or quality of sleep. The program according to claim 1 or 2.

4. The excretion information is a probability indicating whether or not an excretion will occur after a predetermined time, or a time until an excretion occurs. The program according to claim 1 or 2.

5. Acquire training data including information on the excretion status of the subject and sensor information before excretion; Retraining the learning model based on the acquired training data The program according to claim 2.

6. aggregating information on the plurality of excretion situations; Based on the collected results, the time period for excretion of the subject is identified. The program according to claim 5.

7. determining whether a predicted excretion time of the subject based on the excretion information is included in the time period; When the predicted excretion time is included in the time period, an alert indicating the possibility of excretion by the subject is output. The program according to claim 6.

8. Obtaining additional data, including the subject's age or medical history; When additional data and sensor information are input, a learning model outputs excretion information related to the excretion of a subject by inputting the acquired additional data and the sensor information. The program according to claim 1.

9. The information of the subject is output in association with an alert indicating the possibility of excretion by the subject. The program according to claim 8.

10. acquiring information about the subject's independence level corresponding to the sensor information; A message indicating what to do next is output based on the acquired information on the independence level and the excretion information. The program according to claim 1 or 2.

11. Acquire measurement parameters including data on heart rate, data on respiratory rate, data on amount of body movement, and data on quality of sleep during a predetermined period before the date and time of excretion through the bed sensor; calculating a reference parameter based on an average of the plurality of measured parameters; reacquiring the measurement parameters for the predetermined period through the bed sensor; Outputting excretion information of the subject based on a comparison result between the reacquired measurement parameters and the reference parameters. The program according to claim 1.

12. acquiring sensor information obtained through a bed sensor provided in relation to the subject; When sensor information is input, the acquired sensor information is input to a learning model that outputs excretion information regarding the excretion of a subject, and thereby the excretion information of the subject is output. Information processing methods.

13. An information processing device having a control unit, The control unit acquiring sensor information obtained through a bed sensor provided in relation to the subject; When sensor information is input, the acquired sensor information is input to a learning model that outputs excretion information regarding the excretion of a subject, and thereby the excretion information of the subject is output. Information processing device.

Images