Monitoring system, management device, monitoring method, and monitoring program

Abstract

To make it easier to start using the monitoring system for people within monitoring areas such as living spaces. [Solution] The monitoring system comprises at least one radar-type sensor set in the monitoring area, a biometric information detection device that detects the biometric information of a target person in the monitoring area based on sensor data output from the sensor, and a management device. The management device has a trained model that has been trained using multiple training data sets, with medical examination information based on the results of a person's medical examination as input data and biometric condition determination information for determining the degree of the person's biological state as the correct answer data. The medical examination information based on the results of the target person's medical examination is input into the trained model to obtain biometric condition determination information of the target person from the trained model, and using the biometric condition determination information of the target person, biometric condition information including the degree of the target person is generated from the biometric information of the target person output from the biometric information detection device.

Description

Technical Field

[0001] The present disclosure relates to a monitoring system, a management device, a monitoring method, and a monitoring program.

Background Art

[0002] Conventionally, a monitoring system for monitoring a person present in a room has been known. In Patent Document 1, a care support system includes an action detection unit that detects the action or posture of a target person in a room and the position thereof, a time measurement unit that measures the duration of the above action or posture, an action information storage unit that stores past action information of the target person in a past predetermined period and an allowable time set according to the past action information, and a transmission control unit that controls the transmission of an alarm to a terminal possessed by a staff based on the information stored in the action information storage unit and the information detected by the action detection unit and the time measurement unit about the current action or posture of the target person. The transmission control unit is disclosed to transmit an alarm to the terminal when the duration of the current action or posture of the target person in the room is longer than the allowable time.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Patent Document 1 includes an action information storage unit that stores past action information of a target person in a past predetermined period and an allowable time set according to the past action information, and transmits an alarm to a terminal when the duration of the current action or posture of the target person in the room is longer than the allowable time. That is, Patent Document 1 needs to observe the target person for a predetermined period or more in order to specify an allowable time for distinguishing normal and abnormal for the target person (that is, for distinguishing whether or not to transmit an alarm), and it is difficult for a user to start using it immediately.

[0005] The purpose of this disclosure is to make it easier to start using systems or devices for monitoring people within monitoring areas such as living spaces. [Means for solving the problem]

[0006] One aspect of the present disclosure provides a monitoring system comprising: at least one radar-type sensor set in a monitoring area; a biometric information detection device that detects biometric information of a target person in the monitoring area based on sensor data output from the sensor; and a management device, wherein the management device has a trained model that has been trained using a plurality of training data sets that take medical examination information based on the results of a person's medical examination as input data and biometric condition determination information for determining the degree of the person's biological state as correct answer data; medical examination information based on the results of the person's medical examination as input to the trained model to obtain biometric condition determination information of the target person from the trained model; and using the biometric condition determination information of the target person to generate biometric condition information including the degree of the target person from the biometric information output from the biometric information detection device.

[0007] One aspect of the present disclosure provides a management device comprising a processor and memory, wherein a trained model is stored in the memory, which is trained using a plurality of training data, with medical examination information based on the results of a person's medical examination as input data and biological condition determination information for determining the degree of the person's biological state as correct answer data; the processor acquires biological information of a target person in the monitoring area detected based on sensor data output from at least one radar-type sensor set in the monitoring area; medical examination information based on the results of the target person's medical examination is input to the trained model to acquire biological condition determination information of the target person from the trained model; and the biological condition determination information of the target person is used to generate biological condition information, including the degree of the target person, from the biological information of the target person.

[0008] One aspect of the present disclosure provides a monitoring method for monitoring a target person within a monitoring area, comprising: acquiring biometric information of the target person within the monitoring area based on sensor data output from at least one radar-type sensor set in the monitoring area; inputting the medical examination information based on the examination results of the target person into a trained model that has been trained using a plurality of training data sets, the training model being trained using medical examination information based on the results of the person's medical examination as input data and biometric condition determination information for determining the degree of the person's biological state as ground truth data; acquiring biometric condition determination information of the target person from the trained model; and generating biometric condition information, including the degree of the target person, from the target person's biometric information using the biometric condition determination information of the target person.

[0009] One aspect of this disclosure provides a monitoring program for monitoring a target person within a monitoring area, which causes a computer to perform the following actions: acquire biometric information of the target person within the monitoring area based on sensor data output from at least one radar-type sensor set in the monitoring area; input the medical examination information based on the examination results of the target person into a trained model that has been trained using multiple training data sets, the training data set being input data for medical examinations based on the results of the person's medical examination and correct answer data for determining the degree of the person's biological condition; acquire the biometric condition determination information of the target person from the trained model; and use the biometric condition determination information of the target person to generate biometric condition information, including the degree of the target person, from the target person's biometric information.

[0010] These comprehensive or specific embodiments may be implemented as systems, devices, methods, integrated circuits, computer programs, or recording media, or as any combination of systems, devices, methods, integrated circuits, computer programs, and recording media. [Effects of the Invention]

[0011] According to this disclosure, it will become easier to start using systems or devices for monitoring people within monitoring areas such as living spaces. [Brief explanation of the drawing]

[0012] [Figure 1] Block diagram showing an example configuration of the monitoring system according to Embodiment 1. [Figure 2] This figure shows an example of the configuration of sensor data according to Embodiment 1. [Figure 3] A diagram showing an example of the configuration of biological information according to Embodiment 1. [Figure 4] A diagram showing an example of medical information related to Embodiment 1. [Figure 5] This figure shows an example of the setting screen for biological condition determination information according to Embodiment 1. [Figure 6] A diagram illustrating a method for automatically generating biological condition determination information according to Embodiment 1. [Figure 7] Sequence diagram showing an example of display and modification processing of biological condition determination information according to Embodiment 1. [Figure 8] A diagram showing an example of the input screen for medical information according to Embodiment 1. [Figure 9] A diagram showing an example of the setting screen for action determination information according to Embodiment 1. [Figure 10] A diagram showing an example of the room information setting screen according to Embodiment 1. [Figure 11] This figure shows an example of the setting screen for alarm condition information according to Embodiment 1. [Figure 12] Sequence diagram showing an example of monitoring processing according to Embodiment 1 [Figure 13] A diagram showing an example of a monitoring screen according to Embodiment 1. [Figure 14] A diagram showing an example of a detailed screen of a room according to Embodiment 1. [Figure 15] This diagram shows the hardware configuration of a computer that implements the functions of each device related to this disclosure using a computer program. [Modes for carrying out the invention]

[0013] Hereinafter, embodiments of the present disclosure will be described in detail with appropriate reference to the drawings. However, a more detailed description may be omitted as necessary. For example, detailed descriptions of well-known matters and redundant descriptions of substantially identical configurations may be omitted. This is to avoid making the following description unnecessarily redundant and to facilitate understanding by those skilled in the art. Note that the accompanying drawings and the following description are provided to enable those skilled in the art to fully understand the present disclosure, and are not intended to limit the subject matter described in the claims. Even if the functions of one configuration shown in this embodiment are realized by two or more physical configurations, or the functions of two or more configurations are realized by, for example, one physical configuration, it does not matter.

[0014] (Embodiment 1) <System Configuration> FIG. 1 is a block diagram showing a configuration example of a monitoring system 10 according to Embodiment 1. FIG. 2 is a diagram showing a configuration example of sensor data 20 according to Embodiment 1. FIG. 3 is a diagram showing a configuration example of biological information 21 according to Embodiment 1. FIG. 4 is a diagram showing an example of examination information 30 according to Embodiment 1.

[0015] The monitoring system 得 10 is a system that monitors a person (e.g., an elderly person, a care recipient, a patient, an inmate, etc.) existing in a room 8 which is an example of a monitoring area.

[0016] As shown in FIG.得 1, the monitoring system 10 includes a millimeter-wave sensor 11, a biological information detection device 12, a management device 13, a terminal device 14, a position detection device 15, and a reporting device 16. The millimeter-wave sensor 11, the biological information detection device 12, the management device 13, the terminal device 14, and the position detection device 15 may be able to transmit and receive data or information through a predetermined communication network N. Examples of the communication network N include a wired Local Area Network (LAN), a wireless LAN, the Internet, and a mobile communication network.

[0017] At least one millimeter-wave sensor 11 is placed in room 8 (monitoring area). Although Figure 1 shows only one room 8, there may be multiple rooms 8, and a millimeter-wave sensor 11 is installed in each room 8.

[0018] The millimeter-wave sensor 11 is a radar-type sensor that emits radio waves in a predetermined frequency band (e.g., millimeter-wave band) and receives reflected waves reflected from an object (e.g., person 1) within the sensing range of the millimeter-wave sensor 11. The millimeter-wave sensor 11 compares the emitted radio waves with the received reflected waves to generate the sensor data 20 shown in Figure 2. The millimeter-wave sensor 11 may be a millimeter-wave radar-type sensor. Examples of millimeter-wave radar systems include the FMCW (Frequency Modulated Continuous Wave) system or the Step ICW (Interrupted Continuous Wave) system. However, millimeter waves are just an example, and radio waves in frequency bands other than the millimeter-wave band may be used.

[0019] As shown in Figure 2, the sensor data 20 associates timestamps, detection information, and signal information.

[0020] A timestamp is information indicating when the sensor data 20 was obtained. Examples of timestamps include time information or frame numbers.

[0021] The detection information includes, for example, information indicating the position in three-dimensional space where a reflected wave of a predetermined intensity or higher was obtained (hereinafter referred to as the reflection point), information indicating the distance from the millimeter-wave sensor 11 to the reflection point, and information indicating the reflection intensity at the reflection point.

[0022] The signal information includes, for example, IQ data for each receiving antenna.

[0023] In other words, the sensor data 20 is data that shows the position, shape, and temporal changes of objects in three-dimensional space within the sensing range, and does not have the resolution of video data. Therefore, the monitoring system 10 of this embodiment cannot identify each individual, and data acquisition is possible while protecting the privacy of individuals using the millimeter-wave sensor 11. For this reason, even in places where monitoring with video data is difficult from the standpoint of protecting privacy, such as toilets or bathrooms in room 8, millimeter-wave sensing can be performed.

[0024] The millimeter-wave sensor 11 transmits the generated sensor data 20 to the biological information detection device 12.

[0025] At least one location detection device 15 is placed in room 8. The location detection device 15 detects the location of equipment 2 (e.g., wheelchair, walker, cane, etc.) used by person 1 in room 8. For example, if equipment 2 is equipped with a wireless tag 3, the location detection device 15 detects the location of the wireless tag 3 via wireless communication (e.g., WiFi or Bluetooth Low Energy (BLE)) and generates location data 29 indicating its location. The location detection device 15 transmits the generated location data 29 to the management device 13. Note that if equipment 2 is not equipped with a wireless tag 3, the location detection device 15 may be omitted.

[0026] The biological information detection device 12 generates biological information 21 shown in Figure 3 based on sensor data 20 received from the millimeter-wave sensor 11.

[0027] In the biometric information 21, as shown in Figure 3, the timestamp, detected person information, vital sign information, and confidence level are associated.

[0028] A timestamp is information that indicates the timing when the biometric information 21 was generated. An example of a timestamp is time information. The management device 13 uses the time when the biometric information 21 was generated as the timestamp.

[0029] The detected person information is information about the detected person 1. The detected person information includes the person ID, the person's location, and the person's posture. The person ID is an ID used to identify the detected person 1. The person's location indicates the position of person 1 in three-dimensional space within the monitoring area 8, as indicated by the person ID. The person's posture indicates the posture of person 1, as indicated by the person ID. The biometric information detection device 12 identifies the person's location and posture based on the detection information from the sensor data 20.

[0030] Vital sign information refers to vital signs that represent the activity status of the detected person 1. Examples of vital signs include respiratory rate, depth of respiration, intensity of respiration (intensity of abdominal movement due to respiration), heart rate, and blood pressure. The biometric information detection device 12 calculates the vital signs of the detected person 1 (e.g., respiratory rate, depth of respiration, intensity of respiration, heart rate, blood pressure, etc.) based on the time changes in the detection information of the sensor data 20.

[0031] Reliability information indicates the reliability (likelihood) of vital sign information. The biological information detection device 12 calculates the reliability of the calculated vital sign information based on the detection information and signal information of the sensor data 20. The reliability information can be generated based on the noise level ratio of the radar signal, which is the source of the vital sign information, or the result of comparing the reliability information of multiple different vital sign information candidates.

[0032] The biological information detection device 12 transmits the generated biological information 21 to the management device 13.

[0033] The management device 13 maintains medical information 30 for a large number of individuals. Medical information 30 is information that records the results of an individual's medical examination. As shown in Figure 4, medical information 30 includes the individual's name, gender, age, height, and weight. Furthermore, medical information 30 may include the individual's systolic blood pressure, diastolic blood pressure, pulse rate, heart rate, and blood test results. Furthermore, medical information 30 may include information on the individual's medications taken as needed and medical history. Furthermore, medical information 30 may include information indicating what kind of equipment (e.g., wheelchair, walker, cane, etc.) the individual is using, or not using any equipment. The method for inputting medical information 30 will be described later (see Figure 8).

[0034] The management device 13 has a trained model 40 that has been trained (e.g., deep learning) using multiple training data sets, with the person's medical examination information 30 as input data and the biological condition determination information 23 appropriately set for that person as ground truth data. The biological condition determination information 23 includes information for determining the degree from vital signs. When the trained model 40 receives the medical examination information 30 as input, it can output the biological condition determination information 23 that corresponds to (has a correlation with) the medical examination information 30. The management device 13 may also perform additional training (retraining) of the trained model 40 using the medical examination information 30 obtained from subsequent periodic diagnoses. The management device 13 can also perform pattern matching using the medical examination information 30 obtained from periodic diagnoses as input data and the previously detected biological condition information 22 or behavioral determination information 31 as training data. The training and additional training (retraining) of the trained model 40 may be performed by the management device 13 or by another device.

[0035] The management device 13 inputs the medical examination information 30 of the person being monitored 1 into the trained model 40 and obtains the corresponding biological condition assessment information 23. As a result, the management device 13 can automatically generate the biological condition assessment information 23.

[0036] The management device 13 displays the generated biological condition assessment information 23 on the display device 1005 (see Figure 15), and the user (administrator) can modify the displayed biological condition assessment information 23. This allows the user to easily modify the automatically generated biological condition assessment information 23. Details of the biological condition assessment information 23 will be described later (see Figure 5).

[0037] The management device 13 uses the biological condition determination information 23 corresponding to person 1 to generate biological condition information 22 for person 1 from the biological information 21 received from the biological information detection device 12. The biological condition information 22 is information indicating the degree (level) of vital values, which are the activity level of person 1.

[0038] The management device 13 uses the behavior determination information 31 to estimate the behavior or state of person 1 from the person's position and posture included in the person's biometric information 21 and the person's biometric status information 22. The behavior determination information 31 is information that associates the person's position and posture with the biometric status information 22 and the person's estimated behavior at that time. Details of the behavior determination information 31 will be described later (see Figure 9).

[0039] In this way, the monitoring system 10 uses a radar-type millimeter-wave sensor 11 that senses people using radio waves to detect the biological state of person 1. This makes it possible to detect the biological state of person 1 while protecting the privacy of person 1.

[0040] <Settings for Biological Condition Assessment Information> Figure 5 shows an example of the setting screen 100A for the biological condition determination information 23 according to Embodiment 1. Next, the case of manually setting the biological condition determination information 23 will be explained with reference to Figure 5.

[0041] As shown in Figure 5, the setting screen 100 for the biological condition assessment information 23 includes a threshold setting area 101 for vital values, a threshold setting area 102 for the rate of change of vital values ​​over time, a threshold setting area 103 for the depth of respiration, and a setting area 104 for the number of people to be measured.

[0042] The user adjusts the threshold bar 120 included in the vital value threshold setting area 101 to set thresholds for determining the degree of vitality related to the level of vital signs. The degree of vitality related to the level of vital signs is referred to as the vital sign level. For example, if three levels of vital sign level are set, "low," "normal," and "high," the user sets threshold 1 to distinguish between "low" and "normal," and threshold 2 to distinguish between "normal" and "high," as shown in Figure 5. For example, the user may set these thresholds lower (i.e., stricter) for elderly individuals who are being monitored. The set thresholds 1 and 2 may be displayed as numerical values. Note that the number of vital sign levels is not limited to three; there may be two levels, four or more levels. Also, the name of each level of vital sign level may be changed by the user. Furthermore, each level of vital sign level may be read as a vital sign level.

[0043] The user adjusts the threshold bar 120 included in the threshold setting area 102 for the time-dependent change in vital signs to set thresholds for determining the degree of biological change related to the time-dependent change in vital signs. The degree of biological change related to the time-dependent change in vital signs includes both an upward and downward trend. The upward trend is referred to as the degree of vital sign increase. The downward trend is referred to as the degree of vital sign decrease. For example, if three stages are set for the degree of vital sign increase: "no change," "increase," and "rapid increase," then, as shown in Figure 5, threshold 1 is set to distinguish between "no change" and "increase," and threshold 2 is set to distinguish between "increase" and "rapid increase." For example, the user may set these thresholds lower (i.e., stricter) for elderly individuals who are being monitored. The set thresholds 1 and 2 may be displayed as numerical values. Note that each stage of the degree of vital sign increase may be read as a vital sign increase level. For example, if three levels are set for the degree of decline in vital signs—"no change," "decline," and "rapid decline"—threshold 1 is set to distinguish between "no change" and "decline," and threshold 2 is set to distinguish between "decline" and "rapid decline," as shown in Figure 5. For example, the user may set these thresholds lower (i.e., stricter) for elderly individuals being monitored. The set thresholds 1 and 2 may be displayed as numerical values. Note that each level of the degree of decline in vital signs may be read as a vital sign decline level.

[0044] The user adjusts the threshold bar 120 included in the respiratory depth threshold setting area 103 to set a threshold for determining the biological degree related to the respiratory depth of vital signs from the respiratory depth of vital signs. The biological degree related to the respiratory depth of vital signs is referred to as the respiratory depth degree. For example, if three levels of respiratory depth degree are set: "shallow," "normal," and "deep," the user sets threshold 1 to distinguish between "shallow" and "normal," and threshold 2 to distinguish between "normal" and "deep," as shown in Figure 5. For example, the user may set these thresholds lower (i.e., stricter) for elderly people who are being monitored. The set thresholds 1 and 2 may be displayed as numerical values. Note that each level of respiratory depth degree may be read as a respiratory depth level.

[0045] The user sets the maximum number of people to be measured in the measurement person setting area 104. Hereinafter, the set maximum number will be referred to as the maximum measurement person. For example, if the elderly person being monitored lives alone in room 8, the user may set "1" in the measurement person setting area 104. Typically, reducing the maximum measurement person improves the reliability of the biometric information.

[0046] The management device 13 generates biological condition determination information 23, which includes each threshold and the maximum number of people to be measured set by the user, and stores it in memory 1002 or storage 1003 (see Figure 15). This allows the management device 13 to utilize the biological condition determination information 23 in the monitoring process shown in Figure 12.

[0047] <Automatic generation method for determining biological condition> Figure 6 is a diagram illustrating a method for automatically generating biological condition determination information 23 according to Embodiment 1.

[0048] As explained in Figure 5, manually setting each threshold of the biological condition determination information 23 is cumbersome for the user, and it is also difficult for the user to determine whether the settings are appropriate. Therefore, the management device 13 may automatically generate the biological condition determination information 23 corresponding to the person 1 being monitored using the trained model 40.

[0049] As described above, the trained model 40 was trained (deep learning) using multiple training datasets, with the input data being the medical examination information 30 of a large number of people, and the ground truth data being the biomedical condition assessment information 23 appropriately set for those people.

[0050] The management device 13 inputs the medical examination information 30 of the person 1 to be newly monitored into the trained model 40 and obtains the biological condition assessment information 23 output from the trained model 40. As a result, the management device 13 can automatically generate appropriate biological condition assessment information 23 for person 1.

[0051] The management device 13 may display the biological condition determination information 23 obtained from the trained model 40 as a correction screen on the display device 1005. The user may manually correct the contents of the biological condition determination information 23 from the correction screen. This allows the user to easily correct the biological condition determination information 23 even if the biological condition determination information 23 output from the trained model 40 is not appropriate for person 1.

[0052] Furthermore, the management device 13 may retrain the trained model 40 using the medical examination information 30 of the newly designated person 1 as input data, and the corrected biological condition assessment information 23 provided by the user via the correction screen as the correct data. This improves the accuracy of the trained model 40. However, correction of the biological condition assessment information for person 1 is not mandatory, and the management device 13 may save the biological condition assessment information 23 obtained from the trained model 40 without making any corrections.

[0053] <Correction process for biological condition assessment information> Figure 7 is a sequence diagram showing an example of the display and modification process of the biological condition determination information 23 according to Embodiment 1. Figure 8 is a diagram showing an example of the input screen for the medical examination information 30 according to Embodiment 1.

[0054] The terminal device 14 displays an input screen 200 for medical examination information 30, as shown in Figure 8, for example (S201).

[0055] The user (administrator) enters the medical information 30 of the person 1 to be newly monitored from the input screen 200 (S202). For example, the user enters the name, gender, age, height, weight, systolic blood pressure, diastolic blood pressure, pulse rate, heart rate, blood test results, medications taken as needed, medical history, equipment being used, etc.

[0056] The terminal device 14 transmits the entered medical examination information 30 to the management device 13 (S203).

[0057] The management device 13 receives the medical examination information 30 from the terminal device 14. The management device 13 then inputs the medical examination information 30 into the trained model 40 and obtains the corresponding biological condition determination information 23 from the trained model 40 (S204).

[0058] The management device 13 transmits the biological condition determination information 23 to the terminal device 14 (S205).

[0059] The terminal device 14 displays a correction screen for the received biological condition determination information 23 (S206). The correction screen may be a screen like the one shown in Figure 5.

[0060] The user looks at the modification screen and decides whether or not to modify the biological condition assessment information 23 (S207). If the user decides not to modify the biological condition assessment information 23 and inputs this to the terminal device 14 (S207: NO), the terminal device 14 proceeds to step S209.

[0061] If the user decides to modify the biological condition assessment information 23 (S207: YES), they modify each threshold displayed on the modification screen (S208). After the modification is complete, the terminal device 14 proceeds to step S209.

[0062] If the biological condition determination information 23 is not modified, the terminal device 14 sends the biological condition determination information 23 as is to the management device 13. If the biological condition determination information 23 is modified, the terminal device 14 sends the modified biological condition determination information 23 to the management device 13 (S209).

[0063] In step S209, the management device 13 receives the biological condition assessment information 23 transmitted from the terminal device 14. The management device 13 then stores the biological condition assessment information 23 in association with person 1 from step S202 (S210). This allows the management device 13 to obtain appropriate biological condition assessment information 23 for person 1 to be newly monitored.

[0064] The management device 13 uses the medical examination information 30 from step S202 as input data and the biological condition determination information 23 received in step S210 as correct data to train or retrain the trained model 40 (S211). This improves the accuracy of the trained model 40. Therefore, when medical examination information of a person similar to the medical examination information of person 1 is input to the trained model 40, the accuracy of the biological condition determination information output from the trained model 40 improves. In this way, as the retraining of the trained model 40 progresses and the accuracy of the biological condition information improves, the opportunities for the user to manually correct the information in steps S207 and S208 are reduced.

[0065] According to the above process, the user only needs to modify the biological condition assessment information 23 that the management device 13 automatically generates based on the medical examination information 30 of person 1, as needed (and does not need to modify it if it is not necessary). This reduces the user's workload compared to manually setting all of the biological condition assessment information 23.

[0066] <Settings screen for action judgment information> Figure 9 shows an example of the setting screen for the action determination information 31 according to Embodiment 1.

[0067] The behavior determination information 31 is information that includes the determination conditions used when estimating a person's behavior. The user can input or modify the contents of the behavior determination information 31 from the settings screen 210 shown in Figure 9.

[0068] As shown in Figure 9, the setting screen 210 for the action determination information 31 has a UI for setting the alarm flag 211, action 212, location 213, posture 214, and duration 215 in association with each other.

[0069] For example, as shown in the first line 221 of the settings screen 210, the user sets the following associations for action 212 "washing face": alarm flag 211 "OFF", location 213 "bathroom", posture 214 "standing", and duration 215 "120 seconds". The management device 13 generates action determination information 31 that includes the set determination conditions. In this case, if the person's location detected using the biometric information 21 is "bathroom" and the person's posture is "standing" for "120 seconds" or longer, then these satisfy the above determination conditions in the action determination information 31, and the management device 13 estimates that the person's action is "washing face". Also, since the alarm flag 211 is "OFF", when the management device 13 makes this estimation, it does not need to notify the terminal device 14 that the person's biometric state is abnormal (i.e., issue an alarm instruction).

[0070] For example, as shown in line 6, 226, the user sets the following associations for action 212 "fall": alarm flag 211 "ON", location 213 "toilet", posture 214 "lying down", and duration 215 "180 seconds". The management device 13 generates action determination information 31 that includes the set determination conditions. In this case, if the person's location detected using the biometric information 21 is "toilet" and the person's posture is "lying down" for 180 seconds or more, these satisfy the above determination conditions in the action determination information 31, and the management device 13 estimates that the person's action is "fall". Also, since the alarm flag 211 is "ON", if the management device 13 makes this estimation, it may notify the terminal device 14 that the person's biometric state is abnormal (i.e., issue an alarm instruction).

[0071] Furthermore, the behavior determination information 31 may be configured to indicate whether or not device 2 should be used for a given action. For example, the behavior determination information 31 may be configured to indicate that device "cane" or "walker" should be used for action 212 "washing face". In this case, if the management device 13 estimates that person 1's action is "washing face" and determines that person 1 is not using device 2, it may notify the terminal device 14 that the device is not being used (i.e., issue an alert).

[0072] <Room Information Settings Screen> Figure 10 shows an example of the setting screen 250 for room information 32 according to Embodiment 1.

[0073] Room information 32 includes information such as the maximum number of people that can be detected in room 8. Users can input or modify the contents of room information 32 from the settings screen 250 shown in Figure 10.

[0074] As shown in Figure 10, the setting screen 250 for room information 32 has a UI for setting the usage flag 251, the room name 252, and the maximum number of people to be detected 253 in association with each other.

[0075] For example, as shown in line 261, the user sets the usage flag 251 "ON", the room name 252 "ROOM1", and the maximum number of people to be detected 253 "1" for room 1. In this case, the management device 13 will monitor room 1 (because the usage flag 251 is "ON"), set the room name to "ROOM1", and set the maximum number of people that can be in room 1 (i.e., the maximum number of people to be detected) to "1", and then perform the monitoring process. Setting the maximum number of people to be detected in a room improves the accuracy of person detection using millimeter waves.

[0076] <Setting screen for alarm activation conditions> Figure 11 shows an example of the setting screen 300 for the alarm condition information 33 according to Embodiment 1.

[0077] The alarm activation condition information 33 includes information that determines whether or not to issue an alarm in response to the estimated (detected) behavior or state of person 1.

[0078] The setting screen 300 for the alarm condition information 33 has a UI for setting the usage flag 301, the action 302, and the duration 303 in association with each other, as shown in Figure 11. Furthermore, the setting screen 300 for the alarm condition information 33 has a UI for setting the lower limit 304 and upper limit 305 of the respiratory rate per minute, as shown in Figure 11. Furthermore, the setting screen 300 for the alarm condition information 33 has a UI for setting the lower limit 306 and upper limit 307 of the heart rate per minute, as shown in Figure 11.

[0079] For example, as shown in line 311, the user sets the usage flag 301 "ON" and the duration 303 "120 seconds" in association with action 302 "respiratory arrest". In this case, if the biological condition information 22 of person 1 shows that the state of "respiratory arrest" continues for "120 seconds" or more, the management device 13 will find that the conditions of the alarm activation condition information 33 are met, and will send a notification (i.e., an alarm activation instruction) to the terminal device 14 indicating that an abnormality has occurred in the biological condition of person 1.

[0080] For example, the user sets a lower limit of 304 "5" and an upper limit of 305 "25" for the respiratory rate. In this case, if the respiratory rate per minute indicated by the biometric information 21 of person 1 is less than "5" or greater than "25", the management device 13 will send a notification (i.e., an alarm command) to the terminal device 14 indicating that an abnormality has occurred in the biometric condition of person 1, as this satisfies the conditions of the alarm activation condition information 33.

[0081] For example, a user sets a lower limit of 306 "30" and an upper limit of 307 "200" for the heart rate. In this case, if the heart rate per minute indicated by the biometric information 21 of person 1 is less than "30" or greater than "200", the management device 13 sends a notification (i.e., an alert command) to the terminal device 14 indicating that an abnormality has occurred in the biometric condition of person 1.

[0082] <Automatic generation of action judgment information, room information, and alarm activation condition information> The trained model 40 may be configured to output at least one of the following in addition to the biological condition determination information 23: behavior determination information 31, room information 32, and alarm condition information 33. For example, the trained model 40 may be trained using the medical examination information 30 as input data and at least one of the following as ground truth data: behavior determination information 31, room information 32, and alarm condition information 33. The management device 13 may input the medical examination information 30 into the trained model 40 and display the contents of the behavior determination information 31 obtained from the trained model 40 on a setting screen 210 for behavior determination information 31, as shown in Figure 9. The user may modify the behavior determination information 31 from the setting screen 210 as needed. This allows the user to set the behavior determination information 31 more efficiently. The management device 13 may also input the room number along with the medical examination information 30 into the trained model 40 and display the contents of the room information 32 obtained from the trained model 40 on a setting screen 250 for room information 32, as shown in Figure 10. The user may modify the room information 32 from the settings screen 250 as needed. This allows the user to configure the room information 32 more efficiently. The management device 13 may also input the medical examination information 30 into the learned model 40 and display the contents of the alarm condition information 33 obtained from the learned model 40 on the alarm condition information 33 settings screen 300 as shown in Figure 11. The user may modify the alarm condition information 33 from the settings screen 300 as needed. This allows the user to configure the alarm condition information more efficiently.

[0083] <Monitoring and handling> Figure 12 is a sequence diagram showing an example of the monitoring process according to Embodiment 1. Next, the monitoring process performed by the monitoring system 10 will be described with reference to Figure 12.

[0084] The millimeter-wave sensor 11 senses the room 8 and generates sensor data 20 (S301). The millimeter-wave sensor 11 transmits the sensor data 20 to the biological information detection device 12 (S302).

[0085] The biological information detection device 12 generates biological information 21 based on the sensor data 20 received from the millimeter-wave sensor 11 (S303). The biological information detection device 12 transmits the generated biological information 21 to the management device 13 (S304).

[0086] If the device 2 is equipped with an acceleration sensor 4, the acceleration sensor 4 transmits the measured acceleration data 28 to the management device 13 (S305). The acceleration data 28 is used to determine whether or not person 1 is using the device 2, as will be described later. If the device 2 is not equipped with an acceleration sensor 4, step S305 may be omitted.

[0087] The position detection device 15 detects the position of the device 2 based on wireless communication with the wireless tag 3 equipped on the device 2 and generates position data 29 (S306). The position detection device 15 transmits the generated position data 29 to the management device 13 (S307). The position data 29 is used to determine whether or not person 1 is using the device 2, as will be described later. If the device 2 is not equipped with a wireless tag 3, steps S306 to S307 may be omitted.

[0088] The management device 13 receives biological information 21 from the biological information detection device 12. Furthermore, the management device 13 may also receive acceleration data 28 and / or position data 29.

[0089] The management device 13 uses the biometric assessment information 23 to generate biometric assessment information 22 for person 1 from the biometric information 21 of person 1 (S308). This biometric assessment information 23 may be generated to fit person 1 using the trained model 40 as described above. This generates biometric assessment information 22 that is appropriate for each person.

[0090] The control device 13 identifies the location and posture of person 1 from the biometric information 21 of person 1 (S309).

[0091] The control device 13 uses the behavior determination information 31 to estimate the behavior or state of person 1 from the person's position, posture, and biological condition information 22 (S310).

[0092] The control device 13 refers to the behavior determination information 31 and determines whether the behavior or state of person 1 estimated in step S310 is an behavior or state in which the device 2 should be used (S311). For example, if person 1's behavior or state is "going to sleep", the control device 13 determines that it is not an behavior or state in which the device 2 should be used, and if person 1's behavior or state is "washing their face", the control device 13 determines that it is an behavior or state in which the device 2 should be used.

[0093] If the control device 13 determines that the action or condition does not warrant the use of the device 2 (S311: NO), it proceeds to step S315.

[0094] If the control device 13 determines that the action or state requires the use of device 2 (S311: YES), it determines whether person 1 is using device 2 (S312). The method for determining whether person 1 is using device 2 will be described later.

[0095] If the management device 13 determines that person 1 is not using equipment 2 (S312: NO), it sends a notification to the terminal device 14 that person 1 is not using the equipment (S313). The terminal device 14 receives the notification and issues an alert that person 1 is not using equipment 2 (S314) (see Figure 13). Then the process proceeds to step S315.

[0096] If the control device 13 determines that person 1 is using equipment 2 (S312: YES), it proceeds to step S315.

[0097] The control device 13 determines whether the biological condition information 22 of person 1 satisfies the conditions included in the alarm activation condition information 33 (S315).

[0098] If the management device determines that the biological condition information 22 of person 1 satisfies the conditions included in the alarm condition information 33 (S315: YES), it sends a notification to the terminal device 14 that an abnormality has occurred in person 1's biological condition (S316). The terminal device 14 receives the notification and issues an alarm that an abnormality has occurred in person 1's biological condition (S317) (see Figure 13). Then, the process returns to step S301.

[0099] If the control device 13 determines that the biological condition information 22 of person 1 does not meet the conditions included in the alarm condition information 33 (S315: NO), it returns the process to step S301.

[0100] The above-described process may be performed for each room 8. This allows monitoring of the actions or state of individuals in each room 8. Furthermore, the alarm allows the user (e.g., administrator) to immediately recognize that an abnormality has occurred in the biological state of person 1 in a certain room 8, or that person 1 in a certain room 8 is not using device 2 when it should be used.

[0101] <Method for determining whether a person is using an instrument or not> First, we will explain how to determine whether or not person 1 is using device 2 when device 2 is equipped with an acceleration sensor 4.

[0102] The control device 13 continuously receives acceleration data 28 from the acceleration sensor 4 of the device 2. The control device 13 determines whether the direction of movement of person 1, detected from the time change of person position included in the biometric information 21, and the direction of movement of device 2, detected from the time change of acceleration data 28, are the same direction. Furthermore, the control device 13 may also determine whether the speed of movement of person 1, detected from the time change of person position included in the biometric information 21, and the speed of movement of device 2, detected from the time change of acceleration data 28, are roughly the same (i.e., the difference in speed of movement is within a predetermined range). If the control device 13 determines that the direction of movement is the same (or if the direction of movement is the same and the speed of movement is roughly the same), it determines that person 1 is using device 2. If the control device 13 determines that the direction of movement is not the same (or if the direction of movement is not the same or the speed of movement is not the same), it determines that person 1 is not using device 2.

[0103] Next, we will explain a method for determining whether person 1 is using device 2 when device 2 is equipped with a wireless tag 3.

[0104] The management device 13 receives location data 29 detected by the location detection device 15 via wireless communication with the wireless tag 3 of the device 2. The management device 13 determines whether the distance between the person's location included in the biometric information 21 and the location data 29 is less than a predetermined threshold. If the management device 13 determines that the distance is less than the predetermined threshold, it determines that person 1 is using the device 2. If the distance is greater than or equal to the predetermined threshold, it determines that person 1 is not using the device 2.

[0105] The above-described determination method is just one example, and the control device 13 may determine whether or not person 1 is using the device 2 by other means.

[0106] <Monitoring screen> Figure 13 shows an example of a monitoring screen 350 according to Embodiment 1.

[0107] The terminal device may display a monitoring screen 350 as shown in Figure 13. The monitoring screen 350 displays location information 351 indicating where person 1 is currently located within room 8, and behavioral status information 352 indicating person 1's current actions or state. Furthermore, the monitoring screen 350 displays whether or not an alarm has been triggered for each room 8.

[0108] For example, if the management device 13 determines that a person in room "1" is washing their face in the washroom, the terminal device 14 displays "washroom" as location information 351 and "washing face" as action status information 352, as shown in image 361.

[0109] For example, if the management device 13 determines that a person in room "2" has fallen in the toilet, the terminal device 14 displays "toilet" as location information 351 and "fall" as behavioral status information 352, as shown in image 362. Furthermore, if the management device 13 determines that the alarm condition information 33 is met in this case (corresponding to S315: YES in Figure 12), the terminal device 14 displays an alarm 371 indicating that an abnormality has occurred in the person's physical condition, as shown in image 362 (corresponding to S317 in Figure 12). In addition, the terminal device 14 uses the alarm device 16 to issue an alarm indicating that an abnormality has occurred in the physical condition of the person in room "2". For example, if the alarm device 16 is a speaker, the terminal device 14 outputs an audio message from the speaker informing the public that an abnormality has occurred in the physical condition of the person in room "2". For example, if the alarm device 16 is a warning light, the terminal device 14 lights up the warning light.

[0110] This allows the user (administrator) to quickly notice any abnormalities in the biological status of the eight individuals in each room.

[0111] For example, if the control device 13 determines that the person in room "4" is not using the device 2 (corresponding to step S312: NO in Figure 12), the terminal device 14 displays an alert 372 indicating that the device is not in use, as shown in image 363 (corresponding to step S314 in Figure 12).

[0112] For example, if the control device 13 determines that a person in room "5" is using the device 2 (corresponding to step S312: YES in Figure 12), the terminal device 14 may display a message 373 indicating that the device is in use, as shown in image 364.

[0113] This allows the user (administrator) to easily confirm whether the person who owns equipment 2 in each room is using equipment 2.

[0114] <Room details screen> Figure 14 shows an example of a detailed screen 400 of room 8 according to Embodiment 1.

[0115] When a user selects one of the room images on the monitoring screen 350, the terminal device 14 displays the details screen 400 of the selected room, as shown in Figure 14.

[0116] As shown in Figure 14, the detail screen 400 may display the room layout 401, a pictogram 402 of a person showing where and in what posture the person is in room 8, the person's current respiratory level 403, the person's current heart rate level 404, a graph 405 showing the time change (history) of the person's respiratory level, a graph 406 showing the time change (history) of the person's heart rate level, a graph 407 showing the person's estimated behavioral history, and a graph 408 showing the history of the person's locations.

[0117] This allows the user (administrator) to view the details screen 400 and learn more about the actions or status of the people in room 8.

[0118] <Hardware Configuration> As described above, embodiments relating to this disclosure have been described in detail with reference to the drawings, but the functions of the biometric information detection device 12, management device 13, and terminal device 14 described above can be realized by a computer program.

[0119] Figure 15 is a diagram showing the hardware configuration of a computer that implements the functions of each device related to this disclosure using a computer program.

[0120] The computer 1000 comprises a processor 1001, memory 1002, storage 1003, input device 1004, display device 1005, communication device 1006, and bus 1007. The processor 1001, memory 1002, storage 1003, input device 1004, display device 1005, and communication device 1006 are connected to the bus 1007 and can send and receive data bidirectionally via the bus 1007.

[0121] The processor 1001 is a device that executes a computer program stored in memory 1002 and realizes the functional blocks described above. The processor 1001 may be read as a Central Processing Unit (CPU), controller, or control device, etc. The processor 1001 may include a Graphics Processing Unit (GPU) and / or a Neural Network Processing Unit (NPU). The GPU and / or NPU may be used for training and utilizing the pre-trained model 40 described above.

[0122] Memory 1002 is composed of a volatile storage medium and is a device for storing computer programs and data handled by the computer 1000. However, at least a portion of memory 1002 may be composed of a non-volatile storage medium.

[0123] Storage 1003 is a device composed of a non-volatile storage medium that stores computer programs and data handled by computer 1000. Examples of storage 1003 include a Hard Disk Drive (HDD) or a Solid State Drive (SSD).

[0124] The input device 1004 is a device that receives data to be input to the processor 1001. Examples of the input device 1004 include a keyboard, mouse, touchpad, or microphone.

[0125] The display device 1005 is a device that displays images, etc., generated by the processor 1001. Examples of the display device 1005 include liquid crystal displays or organic EL displays.

[0126] The communication device 1006 is a device for sending and receiving data with other devices via a communication network N. The communication device 1006 may support either wired communication or wireless communication. An example of wired communication is Ethernet®. Examples of wireless communication include IEEE 802.11, Wi-Fi®, Bluetooth®, and mobile communication networks (LTE, 4G, 5G).

[0127] (Summary of this disclosure) Based on the description of Embodiment 1 above, the following technology is disclosed. <Technology 1> The monitoring system (10) according to Embodiment 1 includes at least one radar-type sensor (e.g., millimeter-wave sensor 11) set in a monitoring area (8), a biometric information detection device (12) that detects biometric information of a target person in the monitoring area based on sensor data (20) output from the sensor, and a management device (13). The management device has a trained model (40) that has been trained using multiple training data sets, with medical examination information based on the results of a person's medical examination as input data and biometric condition determination information (23) for determining the degree of the person's biological state as correct answer data. The medical examination information based on the results of the target person's medical examination is input to the trained model to obtain biometric condition determination information of the target person from the trained model, and using the biometric condition determination information of the target person, biometric condition information (22) including the degree of the target person is generated from the biometric information of the target person output from the biometric information detection device. This allows for the automatic generation of biometric assessment information based on a person's medical examination data. Therefore, the effort required for users to manually prepare biometric assessment information is reduced, making it easier to start using the monitoring system.

[0128] <Technology 2> In the monitoring system described in Technology 1, if the biological condition assessment information of the subject person is corrected, the management device uses the subject person's medical examination information as input data and the corrected biological condition assessment information of the subject person as ground truth data to retrain the trained model. This allows the modifications to the biological severity assessment information to be reflected in the trained model, improving the accuracy of the biological severity assessment information output by the trained model.

[0129] <Technology 3> In the monitoring system described in Technology 1 or 2, the medical information includes information indicating whether the person in question is a user of the device, and the management device determines whether the person in question is using the device if the medical information indicates that the person in question is a user of the device, and if it determines that the person in question is not using the device, it notifies the administrator accordingly. This allows administrators to quickly notice if the person in question is not using the device.

[0130] <Technology 4> In the monitoring system described in Technology 3, the device is equipped with an acceleration sensor (4), the management device detects the movement of the target person from the biometric information, detects the movement of the device based on the acceleration data measured by the acceleration sensor of the device, and determines whether the target person is using the device based on whether the detected direction of movement of the target person and the direction of movement of the device are in the same direction. This makes it possible to determine whether or not the person in question is using the device.

[0131] <Technology 5> The monitoring system described in Technology 3 further comprises a position detection device that communicates wirelessly with the device to generate position data of the device, and the management device detects the position of the target person from the biometric information and determines whether the target person is using the device based on whether the distance between the detected position of the target person and the position of the device indicated by the position data is less than a predetermined threshold. This makes it possible to determine whether or not the person in question is using the device.

[0132] <Technology 6> In the monitoring system described in any one of the technologies 1 to 5, the management device determines whether or not an abnormality in the biological state of the subject person has occurred based on the subject person's biological status information, and if it determines that an abnormality in the biological state of the subject person has occurred, it notifies the administrator accordingly. This allows administrators to quickly notice if there is an abnormality in the subject's biological condition.

[0133] <Technology 7> In the monitoring system described in any one of technologies 1 to 6, the biological condition determination information includes a threshold for determining the degree of high or low vital values ​​included in the biological information, and the trained model outputs the biological condition determination information with the threshold set, corresponding to the input medical examination information. This makes it possible to automatically generate biological condition assessment information, including thresholds for determining the degree of high or low levels of vital values ​​included in biological information.

[0134] <Technology 8> In the monitoring system described in any one of technologies 1 to 7, the biological condition determination information includes a threshold for determining the degree of change over time of vital values ​​included in the biological information, and the trained model outputs the biological condition determination information with the threshold set, corresponding to the input medical examination information. This enables the automatic generation of biological condition assessment information, including thresholds for determining the degree of change in vital values ​​over time, as contained in biological information.

[0135] <Technology 9> In the monitoring system described in any one of technologies 1 to 8, the biological condition determination information includes a threshold for determining the degree of the depth or intensity of respiration contained in the biological information, and the trained model outputs the biological condition determination information with the threshold set, corresponding to the input medical examination information. This makes it possible to automatically generate biological condition assessment information, including thresholds for determining the degree of respiration depth or intensity contained in biological information.

[0136] <Technology 10> In the monitoring system described in any one of technologies 1 to 9, the sensor is a millimeter-wave radar system. This allows us to obtain sensor data of a person using millimeter waves.

[0137] <Technology 11> In the monitoring system described in any one of technologies 1 to 10, the trained model is trained using the plurality of training data, which further include behavioral judgment information including judgment conditions for estimating a person's behavior as the ground truth data, and the management device inputs medical information based on the medical examination results of the target person into the trained model and obtains behavioral judgment information of the target person from the trained model. This allows for the automatic generation of behavioral assessment information based on a person's medical examination data. Therefore, the effort required for users to manually prepare behavioral assessment information is reduced, making it easier to start using the monitoring system.

[0138] <Technology 12> The management device (13) according to Embodiment 1 includes a processor (1001) and a memory (1002), and stores a trained model in the memory that has been trained using a plurality of training data, in which examination information based on the results of a person's medical examination is used as input data and biological condition determination information for determining the degree of the person's biological state is used as correct answer data. The processor acquires biological information of a target person in the monitoring area detected based on sensor data output from at least one radar-type sensor set in the monitoring area, inputs examination information based on the results of the person's medical examination into the trained model, acquires biological condition determination information of the target person from the trained model, and uses the biological condition determination information of the target person to generate biological condition information including the degree of the target person from the biological information of the target person. This allows for the automatic generation of biometric assessment information based on a person's medical examination data. Therefore, the effort required for users to manually prepare biometric assessment information is reduced, making it easier to start using the monitoring system.

[0139] <Technology 13> The first embodiment of the monitoring method for monitoring a target person within a monitoring area involves acquiring biometric information of the target person within the monitoring area based on sensor data output from at least one radar-type sensor set in the monitoring area, inputting the medical examination information based on the target person's examination results into a trained model that has been trained using multiple training data sets, the model being trained using medical examination information based on the person's medical examination results as input data and biometric condition determination information for determining the degree of the person's biological state as the correct answer data, acquiring biometric condition determination information of the target person from the trained model, and generating biometric condition information including the degree of the target person from the target person's biometric information using the biometric condition determination information of the target person. This allows for the automatic generation of biometric assessment information based on a person's medical examination data. Therefore, the effort required for users to manually prepare biometric assessment information is reduced, making it easier to start using the monitoring method.

[0140] <Technology 14> The monitoring program for monitoring a target person within a monitoring area according to Embodiment 1 involves the computer (1000) performing the following actions: acquiring biometric information of the target person within the monitoring area based on sensor data output from at least one radar-type sensor set up in the monitoring area; inputting the medical examination information based on the person's medical examination results as input data and biometric condition determination information for determining the degree of the person's biological state as correct answer data into a trained model that has been trained using multiple training data sets; acquiring biometric condition determination information of the target person from the trained model; and generating biometric condition information, including the degree of the target person, from the target person's biometric information using the biometric condition determination information of the target person. This allows for the automatic generation of biometric assessment information based on a person's medical examination data. Therefore, the effort required for users to manually prepare biometric assessment information is reduced, making it easier to start using the monitoring program.

[0141] While embodiments have been described above with reference to the attached drawings, this disclosure is not limited to such examples. It is clear to those skilled in the art that various modifications, alterations, substitutions, additions, deletions, and equivalents can be conceived within the scope of the claims, and these are also understood to fall within the technical scope of this disclosure. Furthermore, the components of the embodiments described above can be combined in any way without departing from the spirit of the invention. [Industrial applicability]

[0142] The technology disclosed herein is useful for systems, devices, or methods for monitoring individuals while protecting their privacy. [Explanation of symbols]

[0143] 1 person 2. Equipment 3 Wireless tags 4. Accelerometer 8. Monitoring Area 10. Monitoring System 11 millimeter wave sensor 12. Biological Information Detection Device 13 Management device 14 Terminal devices 15 Position detection device 16 Alarm device 20 Sensor data 21 Biometric Information 22. Biological status information 23. Biological condition assessment information 28 Acceleration Data 29 Location data 30. Medical Information 31 Action Decision Information 32 Room Information 33. Alarm Activation Condition Information 40 pre-trained models 1000 computers 1001 Processor 1002 memory 1003 Storage 1004 Input device 1005 Display device 1006 Communication device 1007 Bus

Claims

1. At least one radar-type sensor set up in the monitoring area, A biometric information detection device that detects the biometric information of a target person within the monitoring area based on sensor data output from the aforementioned sensor, Equipped with a management device, The aforementioned control device is The system has a pre-trained model that uses multiple training datasets, where the input data is medical examination information based on the results of a person's medical examination, and the ground truth data is biological condition assessment information used to determine the degree of the person's biological state. The medical information based on the examination results of the subject person is input into the trained model, and the biological condition assessment information of the subject person is obtained from the trained model. Using the biological condition determination information of the subject person, biological condition information including the degree of the subject person is generated from the biological information of the subject person output from the biological information detection device. A monitoring system.

2. The aforementioned control device is If the biological condition assessment information of the subject person is corrected, the trained model is retrained using the subject person's medical examination information as input data and the corrected biological condition assessment information of the subject person as ground truth data. The monitoring system according to claim 1.

3. The aforementioned medical information includes information indicating whether or not the person in question is a user of the device. The aforementioned control device is If the aforementioned medical information indicates that the subject is a person who uses the device, then it is determined whether or not the subject is using the device. If it is determined that the person in question is not using the equipment, the administrator will be notified accordingly. The monitoring system according to claim 1.

4. The aforementioned device is equipped with an acceleration sensor, The aforementioned control device is The movement of the target person is detected from the aforementioned biometric information, Based on the acceleration data measured by the acceleration sensor of the aforementioned device, the movement of the device is detected. Based on whether the direction of movement of the detected person and the direction of movement of the device are the same, it is determined whether the person is using the device. The monitoring system according to claim 3.

5. The system further comprises a position detection device that performs wireless communication with the aforementioned device to generate position data of the aforementioned device, The aforementioned control device is The location of the target person is detected from the aforementioned biometric information, Based on whether the distance between the detected location of the target person and the location of the device indicated by the location data is less than a predetermined threshold, it is determined whether the target person is using the device. The monitoring system according to claim 3.

6. The aforementioned control device is Based on the biological condition information of the subject person, it is determined whether or not an abnormality in the biological condition of the subject person has occurred. If it is determined that an abnormality in the biological condition of the aforementioned person is occurring, the administrator will be notified accordingly. A monitoring system according to any one of claims 1 to 5.

7. The biological condition determination information includes a threshold for determining the degree of high or low vital values ​​included in the biological information, The trained model outputs the biological condition determination information for which the threshold has been set, corresponding to the input medical examination information. The monitoring system according to claim 1.

8. The biological condition determination information includes a threshold for determining the degree of change over time of vital values ​​included in the biological information, The trained model outputs the biological condition determination information for which the threshold has been set, corresponding to the input medical examination information. The monitoring system according to claim 1.

9. The biological condition determination information includes a threshold for determining the degree of the depth or intensity of respiration contained in the biological information, The trained model outputs the biological condition determination information for which the threshold has been set, corresponding to the input medical examination information. The monitoring system according to claim 1.

10. The aforementioned sensor is a millimeter-wave radar system. The monitoring system according to claim 1.

11. The aforementioned trained model is trained using the aforementioned multiple training data sets, which further include behavioral judgment information, including judgment conditions for estimating human behavior, as the ground truth data. The aforementioned control device is The medical information based on the examination results of the subject is input into the trained model, and behavioral judgment information of the subject is obtained from the trained model. The monitoring system according to claim 1.

12. A management device comprising a processor and memory, A trained model is stored in the memory, using multiple training data sets that use health check information based on the results of a person's health check as input data and biological condition determination information for determining the degree of the person's biological condition as ground truth data. The aforementioned processor, Based on sensor data output from at least one radar-type sensor set up in the monitoring area, biometric information of the target person within the monitoring area is acquired. The medical information based on the examination results of the subject person is input into the trained model, and the biological condition assessment information of the subject person is obtained from the trained model. Using the biological condition determination information of the subject person, biological condition information including the subject person's condition is generated from the subject person's biological information. Management device.

13. A monitoring method for keeping an eye on a target person within a monitoring area, Based on sensor data output from at least one radar-type sensor set up in the monitoring area, biometric information of the target person within the monitoring area is acquired. The medical examination information based on the examination results of the subject person is input to a trained model that has been trained using multiple training data sets, in which the medical examination information based on the results of the person's medical examination is input as input data and the biological condition determination information for determining the degree of the person's biological state is used as ground truth data, and the medical examination information based on the examination results of the subject person is obtained from the trained model. Using the biological condition determination information of the subject person, biological condition information including the subject person's condition is generated from the subject person's biological information. Methods of monitoring.

14. This is a monitoring program that monitors individuals within a designated monitoring area. Based on sensor data output from at least one radar-type sensor set up in the monitoring area, biometric information of the target person within the monitoring area is acquired. The medical examination information based on the examination results of the subject person is input to a trained model that has been trained using multiple training data sets, in which the medical examination information based on the results of the person's medical examination is input as input data and the biological condition determination information for determining the degree of the person's biological state is used as ground truth data, and the medical examination information based on the examination results of the subject person is obtained from the trained model. Using the biological condition determination information of the subject person, biological condition information including the subject person's condition is generated from the subject person's biological information. A monitoring program that has a computer perform certain actions.

Images