Risk prediction device, risk prediction method, and program

Abstract

Provide feedback to users about the risk of decreased sleep quality. [Solution] The risk prediction device acquires a life log, including sleep quality. Based on the life log, the risk prediction device creates a life map representing lifestyle trends. Based on the subject's most recent life log, the risk prediction device estimates the subject's current state. Based on the life map and the subject's current state, the risk prediction device predicts the risk of a decline in the subject's sleep quality. If the risk exceeds a threshold, the risk prediction device provides feedback to the subject.

Description

Technical Field

[0001] The present disclosure relates to a technique for predicting the risk of reduced sleep quality.

Background Art

[0002] Sleep is an essential rest activity for promoting and maintaining health in all age groups, including children, adults, and the elderly. However, the average sleep time of Japanese people is the shortest among 33 countries around the world, mainly in developed countries, and ensuring good-quality sleep is an important health issue for the nation. Also, "ensuring appropriate sleep time" and "improving the sense of sleep rest" are important issues that all citizens should address and are considered significant for extending the healthy life expectancy of our country. Patent Document 1 describes a terminal device that displays advice for improving sleep quality based on the user's biological information and subjective evaluation regarding sleep.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] It is widely known that terminal devices and applications can measure sleep time and sleep quality. However, it has been difficult for terminal devices and applications to make users aware of the risk of reduced sleep quality.

[0005] One of the objectives of the present disclosure is to provide feedback to users on the risk of reduced sleep quality.

Means for Solving the Problems

[0006] To solve the above problems, from one aspect of the present disclosure, a risk prediction device A means of acquiring life logs, including sleep quality, A life map creation method that creates a life map representing lifestyle trends based on the aforementioned life log, A state estimation means that estimates the subject's recent state based on the subject's recent life log, A risk prediction means that predicts the risk of a decline in the sleep quality of the subject based on the life map and the subject's recent condition, The system includes a feedback means for providing feedback to the subject if the risk is above a threshold.

[0007] In other aspects of this disclosure, the risk prediction method performed by the risk prediction device is: By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. If the aforementioned risk exceeds a threshold, feedback will be provided to the person concerned.

[0008] In yet another aspect of this disclosure, a program executed by a computer-based risk prediction device is: By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. If the aforementioned risk exceeds a threshold, the computer is instructed to perform a process to provide feedback to the subject. [Effects of the Invention]

[0009] According to the present disclosure, it is possible to provide feedback to a user regarding the risk of deterioration in the quality of sleep.

Brief Description of the Drawings

[0010] [Figure 1] An example of the schematic configuration of the risk prediction system according to the present disclosure is shown. [Figure 2] It is a block diagram showing an example of the hardware configuration of a server and a subject terminal. [Figure 3] An example of the data stored in the personal information DB and the life log DB is shown. [Figure 4] An example of an input screen for the quality of sleep is shown. [Figure 5] It is a block diagram showing an example of the functional configuration of a server. [Figure 6] An example of a life map with average sleep time and average number of steps as axes is shown. [Figure 7] An example of converting the life map into another graph is shown. [Figure 8] An example of converting and enlarging a part of the life map into another graph is shown. [Figure 9] It is an example of a diagram for explaining deviation detection. [Figure 10] It is a flowchart showing an example of the processing by the feedback unit. [Figure 11] An example of a visualization map with average sleep time and average number of steps as axes is shown. [Figure 12] It is a diagram for explaining action proposals based on the visualization map. [Figure 13] It is a flowchart showing an example of the processing by the action proposal unit. [Figure 14] It is a diagram for explaining a plurality of routes regarding actions to reach a goal. [Figure 15] It is a flowchart showing an example of feedback processing. [Figure 16] An example of a life map for September and the converted graph is shown. [Figure 17] It is a block diagram showing an example of the functional configuration of the risk prediction device. [Figure 18] This flowchart shows an example of processing performed by a risk prediction device. [Modes for carrying out the invention]

[0011] Embodiments of this disclosure will be described below with reference to the drawings. [First Embodiment] (Overall structure) Figure 1 shows an example of the schematic configuration of a risk prediction system 100 to which the risk prediction device of this disclosure is applied. The risk prediction system 100 predicts the risk of decreased sleep quality (hereinafter also simply referred to as "risk") by estimating lifestyle trends based on the subject's life log, and if the risk is high, it is a system that provides feedback to the subject, such as alerts and actions to improve sleep quality (hereinafter also referred to as "recommended actions").

[0012] In the risk prediction system 100 shown in Figure 1, the server 1 and the target user terminal 2 are connected via a network 5 such as the Internet. The risk prediction system 100 provides a service to multiple users that predicts risk, outputs an alert when the risk is high, and suggests recommended actions. Through this service, users can improve the quality of their sleep and promote their health. The target user is one of the users who benefit from the services provided by the risk prediction system 100. In Figure 1, for convenience, the server 1 is shown connected to one target user terminal 2, but in reality, it is connected to multiple terminal devices used by multiple users.

[0013] Target device 2 is a smartphone, smartwatch, or other device used by the target person, which acquires the target person's life log and sends it to server 1, and receives alerts and recommended actions from server 1. Target device 2 is an example of a device used by the target person of this disclosure.

[0014] Server 1 is an information processing device that processes, stores, and transmits various types of data, and is connected to a personal information database (hereinafter referred to as "DB") 31 and a life log DB 32. Server 1 receives the life log of a subject from the subject terminal 2, predicts risks based on the life log, and sends alerts and recommended actions to the subject terminal 2 if the risk is high. Server 1 may be a virtual server located in a cloud environment. Server 1 is an example of a risk prediction device as disclosed herein.

[0015] (Hardware configuration) Figure 2(a) is a block diagram showing an example of the hardware configuration of Server 1. As shown in the figure, Server 1 comprises an interface 11, a processor 12, memory 13, a recording medium 14, a display unit 15, and an input unit 16. These components, along with the personal information database 31 and the life log database 32, are interconnected via a bus.

[0016] Interface 11 exchanges data with the target device 2. Interface 11 is used to receive life logs from the target device 2 and to send alerts and recommended actions to the target device 2 when the risk is high.

[0017] Processor 12 is a computer such as a CPU (Central Processing Unit) that controls the entire server 1 by executing pre-prepared programs. Processor 12 can be a CPU, GPU (Graphics Processing Unit), DSP (Digital Signal Processor), MPU (Micro Processing Unit), FPU (Floating Point Number Processing Unit), PPU (Physics Processing Unit), TPU (Tensor Processing Unit), quantum processor, microcontroller, or a combination of these.

[0018] Memory 13 consists of ROM (Read Only Memory), RAM (Random Access Memory), and other components. Memory 13 stores programs executed by the processor 12. Memory 13 is also used as working memory while the processor 12 is executing various processes.

[0019] The recording medium 14 is a non-volatile, non-temporary recording medium such as a disk-shaped recording medium or semiconductor memory, and is configured to be detachable from the server 1. The recording medium 14 stores various programs that the processor 12 executes. When the server 1 performs feedback processing, the programs stored in the recording medium 14 are loaded into the memory 13 and executed by the processor 12.

[0020] The display unit 15 displays a predetermined image, for example, using an LCD (Liquid Crystal Display). The input unit 16 is used by the operator managing server 1, and can be a keyboard, mouse, touch panel, etc.

[0021] Figure 3(a) is a schematic example of the data stored in the personal information database 31. For example, the personal information database 31 stores the name, gender, date of birth, height, and weight of each user registered during service user registration, linked to an identification ID that identifies the user, as personal information.

[0022] Figure 3(b) shows a schematic example of the data stored in the Life Log DB32. The Life Log DB32 stores, for example, each user's heart rate, pulse, blood pressure, body temperature, steps taken, sleep duration, sleep quality, and meal details, along with time information, linked to an identification ID and stored as a life log. In this way, the life log includes information about parameters that represent the lifestyle tendencies of each user, such as steps taken and sleep duration. The parameters that the life log has are not limited to the example in Figure 3(b), and can be arbitrarily set as long as they represent the lifestyle tendencies of the user.

[0023] Heart rate, pulse, blood pressure, body temperature, steps, and sleep duration data are measured and acquired by a designated application installed on the subject's terminal 2 using sensors. The time information is the date and time when heart rate, pulse, blood pressure, body temperature, steps, and sleep duration were measured. Sleep quality and meal content are acquired by the subject entering the information on an input screen displayed by a designated application installed on the subject's terminal 2. The time information for sleep quality is the date and time when sleep was taken, and the time information for meal content is the date and time when meals were taken. The subject's terminal 2 sends the acquired life logs to the server 1 as needed, and the server 1 stores the life logs in the life log DB 32 when it receives them from the subject's terminal 2.

[0024] Figure 4 shows an example of a sleep quality input screen. As shown in Figure 4, the input screen has a message that reads, "How was your sleep quality today (October 1, 2024)? Please enter a number," and options to rate sleep quality: "1 Very Good," "2 Good," "3 Average," "4 Poor," and "5 Very Poor." For example, when a participant wakes up, they will enter a numerical value representing their subjective evaluation of sleep quality on the input screen displayed on participant terminal 2 by performing a predetermined operation. This allows participant terminal 2 to acquire sleep quality data as a life log.

[0025] Note that the input screen in Figure 4 is just one example; the configuration and method of obtaining subjective evaluations of sleep quality by users are not limited to this example and can be set arbitrarily.

[0026] Figure 2(b) is a block diagram showing an example of the hardware configuration of the target terminal 2. As shown in the figure, the target terminal 2 includes an interface 21, a processor 22, a memory 23, a recording medium 24, a display unit 25, and an input unit 26.

[0027] Interface 21 exchanges data with Server 1 via Network 5. Interface 21 is used to send the subject's life log to Server 1 and to receive alerts and recommended actions from Server 1.

[0028] The processor 22 is a computer such as a CPU, and controls the entire target terminal by executing a pre-prepared program. The processor 22 can be a CPU, GPU, DSP, MPU, FPU, PPU, TPU, quantum processor, microcontroller, or a combination thereof. For example, if the target terminal 2 is a smartwatch, it can be worn by the target and a predetermined program can be executed to obtain the target's heart rate, pulse, blood pressure, body temperature, etc.

[0029] Memory 23 is composed of ROM, RAM, etc. Memory 23 stores programs executed by the processor 22. Memory 23 is also used as working memory while the processor 22 is executing various processes.

[0030] The recording medium 24 is a non-volatile, non-temporary recording medium such as a disk-shaped recording medium or semiconductor memory, and is configured to be detachable from the user terminal 2. The recording medium 24 stores various programs executed by the processor 22. The display unit 25 displays a predetermined image, for example, an LCD. The input unit 26 is a touch panel or the like, and is used when the user performs a predetermined operation.

[0031] (Functional Configuration) Figure 5 is a block diagram showing an example of the functional configuration of Server 1. Functionally, Server 1 includes a personal information acquisition unit 41, a life log acquisition unit 42, a life map creation unit 43, a state estimation unit 44, a deviation detection unit 45, a risk prediction unit 46, a feedback unit 47, a visualization map creation unit 48, an action recording unit 49, and an execution determination unit 50.

[0032] The personal information acquisition unit 41, life log acquisition unit 42, life map creation unit 43, state estimation unit 44, deviation detection unit 45, risk prediction unit 46, feedback unit 47, visualization map creation unit 48, behavior recording unit 49, and execution determination unit 50 are realized by the processor 12 executing a program.

[0033] The personal information acquisition unit 41 uses the subject's identification ID to acquire, for example, the subject's gender and date of birth from the personal information database 31.

[0034] The life log acquisition unit 42 uses the subject's identification ID to acquire, for example, the subject's sleep quality, sleep duration, and step count from the life log DB 32, along with the corresponding time information. The life log acquisition unit 42 also acquires the user's life log from the life log DB 32, which is necessary for creating a life map.

[0035] The life map creation unit 43 creates a life map that represents lifestyle trends based on life logs. Specifically, the life map creation unit 43 creates a life map that represents one user as one point, using parameters included in the life log as axes. Alternatively, a period of time for one user (for example, a week) may be represented as one point. Figure 6 shows an example of a life map with average steps and average sleep time as axes. The life map 40 shown in Figure 6 is a "life map of a man in his 60s in February," and is based on statistical information from the life logs of several male users in their 60s who are the same age and gender as the subject, with the horizontal axis (x-axis) representing average steps and the vertical axis (y-axis) representing average sleep time. Life map 40 is a density map that represents one man in his 60s as one point based on average steps and average sleep time. The life map makes it possible to estimate the lifestyle trends of users of the same age and gender as the subject. The life map also makes it possible to visualize the subject's own state in relation to the average state of other users.

[0036] Furthermore, the life map is not limited to being created based on the life logs of multiple users of the same age and gender as the subject; it may also be created based on the subject's own past life logs. In this case, the life map creation unit 43, for example, uses the subject's life logs for the past several years from February 1st to February 28th, with the horizontal axis (x-axis) representing steps taken and the vertical axis (y-axis) representing sleep time, and creates a life map in which the subject for each date is represented by a single point based on the steps taken and sleep time for that date. This allows for the estimation of the subject's own lifestyle trends through the life map.

[0037] Alternatively, the life map may be created by extracting only life logs with a good subjective evaluation of sleep quality and using the extracted life logs as a basis. In this case, the life map creation unit 43, for example, extracts only the life logs of users who entered a subjective evaluation of "1 Very Good" or "2 Good" from the life logs of users of the same age and gender as the target person for February, and creates a life map based on the statistical information of the extracted life logs. According to this, the life map makes it possible to estimate the lifestyle trends of users of the same age and gender as the target person who feel that they have high quality sleep.

[0038] Furthermore, the life map creation unit 43 extracts only the life logs of the subject for the past several years from the subject's life logs for the period from February 1st to February 28th, specifically those for which the subjective evaluation was entered as "1 Very Good" or "2 Good," and creates a life map based on the extracted life logs. This allows the life map to estimate the lifestyle tendencies of the subject when they feel they have high quality sleep.

[0039] Figure 6 shows a life map 40 created as an example, with average steps and average sleep time as the x and y axes, respectively. However, this disclosure is not limited to this, and a life map can be created using any parameter included in the life log as the axis.

[0040] The state estimation unit 44 estimates the subject's current state based on the subject's recent life log. Specifically, the state estimation unit 44 estimates the subject's current state from their position on the life map based on the parameters included in the subject's recent life log. For example, if the subject's recent sleep time and step count are 430 minutes and 4500 steps, respectively, the state estimation unit 44 estimates the subject's current state from the position of star 35 on the life map shown in Figure 6.

[0041] The deviation detection unit 45 detects when the subject's recent state deviates from the normal state in terms of lifestyle trends and acquires deviation information. The normal state may be, for example, the average state of multiple users of the same age and gender, or a state in which there are many users of the same age and gender. If the life map is created based on the subject's own past life logs, the normal state may be, for example, the average state of the subject on each date, or a state in which there are many subjects on each date.

[0042] Figure 7(a) is an example of converting Life Map 40 into a three-dimensional graph with the density of points representing users as the Z-axis. Figure 7(b) is an example of converting Life Map 40 into a graph where the density of points representing users is represented by contour lines. In the graphs shown in Figures 7(a) and 7(b), the higher the peak, the more users there are with the same average sleep time and average number of steps. Figure 8 is an example of converting the area enclosed by the thick line 36 of Life Map 40 into a graph where the density of points representing users is represented by density. In the graph 60 shown in Figure 8, the higher the density, the more users there are with the same average sleep time and average number of steps.

[0043] Figures 7(a), 7(b), and 8 are all examples of converting the life map 40 into a different graph. However, for the sake of explanation, we will use a graph like Figure 8, which represents the density of points indicating users as a density, to explain deviation detection. Also, Figures 7(a), 7(b), and 8 can be considered life maps that represent the lifestyle trends of users.

[0044] Figure 9 is an example of a diagram illustrating deviation detection. As shown in Figure 9, in graph 60, the area below the thick line 61 has a high density, indicating that there are many users with the same average sleep time and average number of steps, i.e., it is within the normal range. On the other hand, the area above the thick line 61 has a low density, indicating that there are few users with the same average sleep time and average number of steps, i.e., it is outside the normal range. If the subject's most recent state is at position 35 stars, the deviation detection unit 45 determines that it is within the normal range because it is in the area below the thick line 61, and therefore does not detect a deviation. On the other hand, if the subject's most recent state is at position 62 stars, the deviation detection unit 45 determines that it is outside the normal range because it is in the area above the thick line 61, and therefore detects a deviation. When a deviation is detected, the deviation detection unit 45 acquires deviation information in graph 60, such as how much the position of star 62, which indicates the most recent state, deviates from the normal range.

[0045] In graph 60, the thick line 61 is, for example, a boundary line that distinguishes between densities above and below a threshold, and can be set arbitrarily. In other words, the boundary line on the life map used for deviation detection can be set arbitrarily.

[0046] The risk prediction unit 46 predicts the risk of a decline in the subject's sleep quality based on lifestyle trends and the subject's recent condition. Specifically, the risk prediction unit 46 predicts risk based on deviation information. For example, server 1 predicts a high risk if the subject's recent condition deviates from the normal state. Server 1 may also predict risk by considering subjective evaluations of sleep quality and the magnitude of the deviation. Risk may be expressed in any stage or as a numerical value.

[0047] The feedback unit 47 provides feedback to the subject if the risk of a decline in the subject's sleep quality exceeds a threshold. Figure 10 is a flowchart of an example of processing by the feedback unit 47. As shown in Figure 10, the feedback unit 47 obtains a subjective evaluation of sleep quality from the subject's most recent life log (step S101). Next, the feedback unit 47 determines whether the sleep quality is high or low (step S102). High sleep quality means, for example, that the subjective evaluation is "1 Very good" or "2 Good". On the other hand, low sleep quality means, for example, that the subjective evaluation is "3 Average", "4 Poor", or "5 Very Poor".

[0048] If the sleep quality is determined to be poor (Step S102; No), the feedback unit 47 determines whether there is a high risk of further deterioration of sleep quality (Step S103). If the risk is above the threshold and the risk is determined to be high (Step S103; Yes), the feedback unit 47 outputs an alert to inform the subject that there is a high possibility of further deterioration of sleep quality and outputs advice suggesting recommended actions (Step S104). On the other hand, if the risk is below the threshold and the risk is determined to be low (Step S103; No), the feedback unit 47 does not output an alert and outputs advice to improve sleep quality (Step S105).

[0049] If the sleep quality is determined to be high (Step S102; Yes), the feedback unit 47 determines whether there is a high risk of a decline in sleep quality (Step S106). If the risk is above the threshold and the risk is determined to be high (Step S106; Yes), the feedback unit 47 outputs an alert to inform the subject that there is a high possibility of a decline in sleep quality and outputs advice suggesting recommended actions (Step S107). On the other hand, if the risk is below the threshold and the risk is determined to be low (Step S106; No), the feedback unit 47 does not output an alert and outputs advice recommending that the current state be maintained.

[0050] As shown in Figure 5, the feedback unit 47 includes an alert output unit 51 and an action suggestion unit 52. The alert output unit 51 outputs an alert to the subject terminal 2 to inform the subject that there is a high possibility of poor sleep quality if the risk is above a threshold. The action suggestion unit 52 proposes recommended actions based on a visualization map (described later) and the subject's recent condition if the risk is above a threshold or if the subject's subjective evaluation of sleep quality is low.

[0051] The visualization map creation unit 48 creates a visualization map that visualizes the relationship between lifestyle trends and subjective evaluations of sleep quality, based on life logs. Specifically, the visualization map creation unit 48 creates a visualization map using life logs with similar subjective evaluations of sleep quality, with an arbitrary parameter as the axis and the density of points represented by concentration. According to the visualization map, it is possible to visualize what kind of life logs result in good subjective evaluations of sleep quality, and what kind of life logs result in poor subjective evaluations of sleep quality. If the relationship between one's life logs and sleep quality is understood from the visualization map, server 1 can take proactive measures, such as suggesting recommended actions to the subject if the subject's sleep quality is likely to decline.

[0052] Figure 11 shows an example of a visualization map based on average steps and average sleep time. Visualization map 65 shown in Figure 11 is a graph that visualizes the probability of low sleep quality in February for men in their 60s, who are the same age and gender as the target group.

[0053] Figure 11 is a visualization map created based on the life logs of users who rated their sleep quality as low. However, the visualization map is not limited to this and may be created based on the life logs of users who rated their sleep quality as high. In this case, the visualization map is a graph that visualizes the probability of high sleep quality in February for men in their 60s, who are the same age and gender as the subject. Alternatively, the visualization map may be a graph that visualizes the probability of low or high sleep quality for the subject in February, based on the subject's own past life logs.

[0054] The action suggestion unit 52 proposes recommended actions based on the visualization map and the subject's current state. Specifically, the action suggestion unit 52 proposes which actions related to which parameters, which are the axes of the visualization map, should be performed and in what order, and then transmits the proposed content to the subject's terminal 2 for display.

[0055] Figure 12 is a diagram illustrating action suggestions based on a visualization map. For example, the action suggestion unit 52 sets the target to 55 stars, which is the position with the fewest subjective evaluations of poor sleep quality in the visualization map 65 shown in Figure 12, and proposes actions to reach the target of 55 stars, starting from 53 stars, which represents the subject's most recent state. As shown in Figure 12, in order to reach 55 stars from 53 stars, it is necessary to increase the subject's daily steps and decrease their daily sleep time so that the average number of steps and average sleep time are the same as those indicated by the 55-star position.

[0056] Furthermore, if the visualization map is a graph that visualizes the probability of high sleep quality, the action suggestion unit 52 targets the position where subjective evaluations of high sleep quality are most frequent, and proposes actions to reach the target position from the position representing the subject's most recent state.

[0057] The behavior recording unit 49, after suggesting recommended behaviors, temporarily records the actions actually performed by the subject based on the subject's life log within a predetermined period, for example, in memory 13.

[0058] The execution determination unit 50 compares the action performed by the subject with the action proposed to the subject and determines whether or not the subject performed the proposed action.

[0059] The action suggestion unit 52 proposes the next action to improve sleep quality based on the result of determining whether the subject has performed the proposed action, the visualization map, and the subject's recent state. Figure 13 is a flowchart showing an example of processing by the action suggestion unit 52. As shown in Figure 13, the action suggestion unit 52 first proposes recommended actions for the subject based on the visualization map and the subject's recent state (step S201).

[0060] Figure 14 is a diagram illustrating multiple routes for actions to reach a goal. For example, the action suggestion unit 52, using the visualization map 65, proposes Route A, as shown by the black arrow in Figure 14(a), to reduce sleep time first and then increase steps, as actions to reach the target star 55 from star 53, which indicates the subject's current state. Each arrow corresponds to an action taken in a day. In the example in Figure 14(a), the action suggestion unit 52 proposes reducing daily sleep time for two days (days 1 and 2 of Route A) and increasing daily steps for three days (days 3 through 5 of Route A).

[0061] The action suggestion unit 52 may, for example, send and display to the subject terminal 2 the advice to "reduce sleep time" on day 1, or it may send and display to the subject terminal 2 information regarding the visualization map 65, stars 53 and 55, and arrows in addition to the advice. Furthermore, the advice suggesting action may only be for the next day, i.e., day 1 of route A, or it may be for all actions from day 1 to day 5 of route A.

[0062] In step S201, when the action suggestion unit 52 suggests the first action, the execution determination unit 50 determines whether the subject's action was as suggested (step S202). Specifically, the execution determination unit 50 compares the subject's actual action recorded by the action recording unit 49 with the first action suggested by the action suggestion unit 52 and determines whether the subject performed the suggested action.

[0063] For example, if the actual behavior of the subject recorded by the behavior recording unit 49 was "increase the number of steps" as indicated by the white arrow in Figure 14(a), the first behavior proposed by the behavior suggestion unit 52 was "reduce sleep time," and therefore the execution determination unit 50 determines that the subject's behavior was not as suggested (Step S202; No).

[0064] If the subject's actions are not as suggested (Step S202; No), the action suggestion unit 52 re-searches for a route and proposes a new action (Step S203). As shown in Figure 14(b), the action suggestion unit 52 re-searches for a route based on the subject's actual action of "increasing the number of steps" and proposes Route B, which involves increasing the number of steps for the first three days and reducing sleep time for the following two days, as the action to reach the target of 55 stars.

[0065] The action suggestion unit 52 may, for example, send and display advice such as "increase your steps" to the user's terminal 2, or it may send and display information related to the visualization map 65, stars 53 and 55, and arrows to the user's terminal 2 in addition to the advice. Furthermore, the advice suggesting an action may only cover the actions for the next day, i.e., the second day of Route B, or it may cover all the actions from the third to the fifth day of Route B.

[0066] In step S203, when the action suggestion unit 52 suggests the next action, the execution determination unit 50 determines whether the subject's actions were as suggested (step S204). If the subject's actions were not as suggested (step S204; No), the action suggestion unit 52 re-searches the route and suggests a new action (step S205). Specifically, the action suggestion unit 52 suggests a new action for route D based on the subject's actual actions. On the other hand, if the subject's actions were as suggested (step S204; Yes), the action suggestion unit 52 suggests the next action for route B (step S206). As shown in Figure 14(b), if the subject's actual actions were as suggested, "increase the number of steps", the action suggestion unit 52 suggests the action for day 3 of route B, which is "increase the number of steps even further".

[0067] In step S202, if the subject's actions are as suggested (step S202; Yes), the action suggestion unit 52 suggests the next action in route A (step S207). As shown in Figure 14(a), if the subject's actual actions are as suggested, "reduce sleep time", the action suggestion unit 52 suggests the action for the second day of route A, which is "further reduce sleep time".

[0068] In step S207, when the action suggestion unit 52 suggests the next action on route A, the execution determination unit 50 determines whether the subject's actions were as suggested (step S208). If the subject's actions were not as suggested (step S208; No), the action suggestion unit 52 re-searches the route and suggests a new action (step S209). Specifically, the action suggestion unit 52 suggests a new action on route C based on the subject's actual actions. On the other hand, if the subject's actions were as suggested (step S208; Yes), the action suggestion unit 52 suggests the next action on route A (step S210). As shown in Figure 14(a), if the subject's actual actions were as suggested, "reduce sleep time," the action suggestion unit 52 suggests the action for day 3 of route B, which is "further reduce sleep time."

[0069] Thus, the action suggestion unit 52 proposes actions with a statistically high sleep quality target, but it can propose appropriate actions to the subject by updating the next suggested action as needed in response to the actions the subject actually takes after the suggestion. In other words, the action suggestion unit 52 can change the recommended actions in response to the actions the subject actually takes after the suggestion.

[0070] Furthermore, the order of actions—whether to increase steps before decreasing sleep or decrease steps before increasing sleep—to reach the target position from the current state of the subject should ideally be determined based on the density of points representing the user, using a visualization map like the one shown in Figure 14(a). For example, by proposing an order of actions that progresses from a high density of points representing the user to a low density, it is possible to first propose actions that are more common and less challenging. This reduces the likelihood of users giving up on actions aimed at improving their sleep because they find them difficult.

[0071] In the above configuration, the life log acquisition unit 42, life map creation unit 43, state estimation unit 44, deviation detection unit 45, risk prediction unit 46, feedback unit 47, visualization map creation unit 48, action recording unit 49, and execution determination unit 50 of server 1 are examples of the life log acquisition means, life map creation means, risk prediction means, feedback means, visualization map creation means, action recording means, and execution determination means of this disclosure, respectively. Furthermore, the alert output unit 51 and action suggestion unit 52 of the feedback unit 47 are examples of the alert output means and action suggestion means, respectively.

[0072] (Feedback processing) Next, we will explain the feedback processing by Server 1. Figure 15 is a flowchart showing an example of feedback processing by Server 1. This processing is achieved by the processor 12 shown in Figure 2(a) executing a pre-prepared program.

[0073] First, Server 1 retrieves the life logs of the subject and users of the same age and gender from Life Log DB 32 (Step S301). Next, Server 1 creates a life map based on the statistical information of the retrieved life logs, for example, with the average number of steps on the x-axis and the average sleep time on the y-axis (Step S302). Next, Server 1 estimates the subject's recent state based on the subject's most recent life logs (Step S303).

[0074] Server 1 predicts the risk of a decline in the subject's sleep quality based on the life map and the subject's recent condition (step S304). Next, Server 1 determines whether or not to issue an alert (step S305). If the risk is above a threshold, Server 1 determines to issue an alert (step S305; Yes) and sends an alert to the subject's terminal 2 informing them that there is a high risk of a decline in sleep quality (step S306).

[0075] On the other hand, if the risk is below the threshold, Server 1 determines that it will not output an alert (Step S305; No) and determines whether the subject's most recent subjective assessment of sleep quality is good or not (Step S307). If the subjective assessment of sleep quality is good (Step S307; Yes), Server 1 outputs advice to the subject terminal 2 that it is desirable to maintain the current state (Step S309) and terminates the feedback process. On the other hand, if the subjective assessment of sleep quality is poor (Step S307; No), Server 1 outputs advice to the subject terminal 2 to improve sleep quality, such as getting sunlight upon waking or doing moderate stretching (Step S308) and terminates the feedback process.

[0076] After outputting an alert in step S306, Server 1 creates a visualization map and, based on the subject's recent state, searches for a route to improve sleep quality on the visualization map (step S310). Next, Server 1 outputs advice suggesting recommended actions to the subject's terminal 2 based on the searched route (step S311). Next, Server 1 records the actions taken by the subject based on the subject's life log within a predetermined period after the suggested actions were proposed (step S312).

[0077] Next, Server 1 determines whether the target has been reached based on the actions performed by the subject (Step S313). Specifically, Server 1 determines, based on the subject's most recent life log, whether the number of steps and sleep time indicated by the subject's most recent state have reached the number of steps and sleep time indicated by the target location on the visualization map.

[0078] If it is determined that the goal has not been reached (Step S313; No), Server 1 compares the action performed by the subject with the action suggested to the subject and determines whether the subject performed the suggested action (Step S314). If the subject performed the suggested action (Step S314; Yes), Server 1 returns to the process in Step S311 and suggests the next recommended action based on the route explored earlier (Step S311).

[0079] On the other hand, if the subject has not taken the suggested action (step S314; No), Server 1 returns to the process in step S310 and, based on the subject's most recent state, re-searches for a route to improve sleep quality on the visualization map (step S310). Next, Server 1 proposes the following recommended action based on the re-searched route (step S311).

[0080] If it is determined that the goal has been reached during the process in step S313 (step S313; Yes), Server 1 completes the feedback process.

[0081] In the feedback process shown in Figure 15, if the risk of poor sleep quality is below a threshold but the subjective evaluation of sleep quality is poor, advice such as getting sunlight upon waking is output. However, this disclosure is not limited to this, and the process may proceed to step S310, where advice suggesting recommended actions is output based on the visualization map and the subject's most recent condition.

[0082] Furthermore, in the feedback process shown in Figure 15, in steps S310 and S311, regardless of the subject's subjective evaluation of sleep quality, if the risk is above a threshold, recommended actions are suggested and output as advice to the subject's terminal 2. However, this disclosure is not limited to this, and if the subject's subjective evaluation of sleep quality is good, the server 1 may output advice such as, "Your perceived sleep quality seems good, but there is a high risk of your sleep quality deteriorating, so please be careful," instead of recommending actions.

[0083] According to this risk prediction system 100, based on life logs, it can predict the risk of a decline in a subject's sleep quality. It can provide feedback to the subject, such as alerts and advice, not only when the subject's subjective assessment of sleep quality is low, but also when it is high, as long as the risk exceeds a threshold. In other words, it can make the subject aware of the risk of a decline in sleep quality and support their healthcare.

[0084] Furthermore, the risk prediction system 100 can plan what behavioral changes should be made to improve sleep quality based on the life log, and provide appropriate advice to the subject if the risk is above a threshold or if the subjective evaluation of sleep quality is low. Specifically, based on the life map and the subject's recent life log, the risk prediction system 100 can detect that the subject's condition deviates from the normal and provide advice on specific actions to recover from the deviation. The risk prediction system 100 can also determine whether the subject has taken the actions advised and appropriately update the subsequent advice based on the determination result. As a result, the risk prediction system 100 can always provide appropriate advice to subjects who have difficulty following the advice.

[0085] [First variation] The life map shown in Figure 6 was created based on life logs from February, but this disclosure is not limited to this, and the life map can be created based on life logs from any period as desired. Figure 16(a) is a life map created based on the life logs of a male in his 60s, who is the same age and gender as the target individual, for September. Comparing the September life map 70 shown in Figure 16(a) with the February life map 40 shown in Figure 6, it can be seen that the average number of steps is slightly lower for users in the September life map compared to February. Thus, as life logs change with temperature and climate, the threshold for detecting deviations and the recommended actions to be suggested also change, so it is desirable to create a life map based on life logs from an appropriate period.

[0086] Figure 16(b) is an example of converting the Life Map 70 into a three-dimensional graph with the density of points representing users as the Z-axis. Figure 16(c) is an example of converting the Life Map 70 into a graph where the density of points representing users is represented by contour lines.

[0087] [Second variation] Life maps and visualization maps are created based on life logs, with two or more parameters representing lifestyle trends as axes. Temperature and atmospheric pressure are thought to affect the autonomic nervous system and have a significant impact on sleep quality. Therefore, in addition to life logs, weather information such as daily temperature and atmospheric pressure may be stored in a designated database and used as parameters for the life map and visualization map axes. In this case, the life map creation unit 43 and the visualization map creation unit 48 create life maps and visualization maps that utilize weather changes based on life logs and weather information, and the action suggestion unit 52 proposes recommended actions that reflect the weather changes. As a result, the risk prediction system 100 can predict risks by considering weather conditions that cause changes in atmospheric pressure, such as typhoons, and provide appropriate feedback to the subject according to the prediction results.

[0088] Alternatively, instead of storing life logs and weather information in separate databases, weather information may be treated as part of the life log.

[0089] [Third variation] The risk of reduced sleep quality may be predicted using a machine learning model, which is a risk prediction model. For example, the risk prediction unit 46 may construct a risk prediction model that outputs an optimized numerical value indicating risk when a life map and data on the subject's recent state are input. Training data is used to construct (generate) the risk prediction model. Training data is data that associates the input data input to the learning of the risk prediction model with the corresponding ground truth data. The input data consists of various life maps and data on the subject's recent state, and the ground truth data is a numerical value indicating risk. The risk prediction unit 46 trains the risk prediction model to output a numerical value indicating risk based on the life map and data on the subject's recent state input as input data. Examples of machine learning methods include models using neural networks. According to this, the risk prediction unit 46 can use the numerical value indicating risk output by the risk prediction model as the subject's risk.

[0090] [Fourth variation] In the above embodiment, the subject uses subject terminal 2, but this disclosure is not limited thereto, and the subject may use subject terminal having the functionality of server 1. In this case, the subject terminal can predict the risk by performing the feedback processing that was performed by server 1, and can provide appropriate feedback to the subject according to the prediction result.

[0091] [Second Embodiment] Figure 17 is a block diagram showing an example of the functional configuration of the risk prediction device in this disclosure. The risk prediction device 90 comprises a life log acquisition means 91, a life map creation means 92, a state estimation means 93, a risk prediction means 94, and a feedback means 95.

[0092] Figure 18 is a flowchart illustrating an example of processing by the risk prediction device 90. The life log acquisition means 91 acquires life logs, including sleep quality (step S401). The life map creation means 92 creates a life map representing lifestyle trends based on the life logs (step S402). The state estimation means 93 estimates the subject's current state based on the subject's most recent life logs (step S403). The risk prediction means 94 predicts the risk of the subject's sleep quality deteriorating based on the life map and the subject's current state. The feedback means 95 provides feedback to the subject if the risk is above a threshold (step S405).

[0093] In addition, some or all of the above embodiments (including modifications, the same applies hereinafter) may also be described as follows, but are not limited to the following.

[0094] (Note 1) A means of acquiring life logs, including sleep quality, A life map creation method that creates a life map representing lifestyle trends based on the aforementioned life log, A state estimation means that estimates the subject's recent state based on the subject's recent life log, A risk prediction means that predicts the risk of a decline in the sleep quality of the subject based on the life map and the subject's recent condition, If the aforementioned risk exceeds a threshold, a feedback means for providing feedback to the subject, A risk prediction device equipped with the following features.

[0095] (Note 2) The life map includes deviation detection means that detects when the subject's recent state deviates from the subject's normal state and acquires deviation information related to the deviation. The life log acquisition means acquires the life logs of the subject at multiple points in the past, The life map creation means creates a life map representing the lifestyle trends of the subject based on the subject's past life logs. The risk prediction means is a risk prediction device as described in Appendix 1, which predicts the risk based on the deviation information.

[0096] (Note 3) The aforementioned subject is one of several users, The life map includes deviation detection means that detects when the subject's recent state deviates from the normal state of users of the same gender and age as the subject, and acquires deviation information related to the deviation. The life log acquisition means acquires life logs of multiple users at multiple points in the past, The life map creation means creates a life map that represents the lifestyle trends of users of the same gender and age group as the target user, based on the life logs of multiple users. The risk prediction means is a risk prediction device as described in Appendix 1, which predicts the risk based on the deviation information.

[0097] (Note 4) The feedback means is a risk prediction device according to Appendix 1 that outputs an alert to the terminal used by the subject when the risk is above a threshold.

[0098] (Note 5) A visualization map creation method that creates a visualization map that visualizes the relationship between lifestyle trends and subjective evaluations of sleep quality based on the aforementioned life log, The system includes an action suggestion means that, when the aforementioned risk exceeds a threshold, suggests actions to improve sleep quality based on the visualization map and the subject's recent condition, The feedback means is a risk prediction device as described in Appendix 4, which outputs advice indicating actions to improve sleep quality to a terminal used by the subject.

[0099] (Note 6) A behavioral recording means that records the actions taken by the subject based on the subject's life log during a predetermined period after suggesting actions to improve sleep quality, The system includes an execution determination means that compares the actions performed by the subject with the actions proposed to the subject, and determines whether or not the subject performed the proposed actions. The action suggestion means is a risk prediction device as described in Appendix 5, which suggests the next action to improve sleep quality based on the result of the judgment, the visualization map, and the subject's recent condition.

[0100] (Note 7) The aforementioned life log includes information on several parameters that represent lifestyle trends, The life map and the visualization map are based on two or more parameters, each with an axis. The aforementioned parameters include information about the weather for one day, The life map creation means creates a life map using the weather changes, The action suggestion means is a risk prediction device as described in Appendix 6 that suggests an action that reflects the changes in weather.

[0101] (Note 8) The risk prediction means is a risk prediction device according to Appendix 1, which predicts the risk of the subject's sleep quality decreasing by using a machine learning model that has been trained to output an optimized numerical value indicating the risk in response to input of the life map and data relating to the subject's recent condition.

[0102] (Note 9) A risk prediction method performed by a risk prediction device, By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. A risk prediction method that provides feedback to the subject if the aforementioned risk exceeds a threshold.

[0103] (Note 10) A program executed by a risk prediction device equipped with a computer, By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. A program that causes the computer to perform a process to provide feedback to the subject if the risk is above a threshold.

[0104] (Note 11) The aforementioned map creation means is a risk prediction device as described in Appendix 1, which creates the life map based on a life log recorded when the user felt they had good sleep quality.

[0105] (Note 12) The aforementioned life log includes information on several parameters that represent lifestyle trends, The life map and the visualization map each have two or more parameters as axes. The aforementioned action suggestion means is a risk prediction device as described in Appendix 6, which suggests which parameters to perform and in what order.

[0106] (Note 13) The aforementioned parameters are sleep duration and step count, as described in Appendix 12, for the risk prediction device.

[0107] (Note 14) The action suggestion means targets the position on the visualization map where subjective evaluations of high sleep quality are most frequent, and proposes which parameters to perform and in what order to reach the target position from the position showing the subject's most recent state. This is the risk prediction device described in Appendix 12.

[0108] (Note 15) The action suggestion means targets the position on the visualization map where the subjective evaluation of poor sleep quality is lowest, and proposes which parameters to perform and in what order to reach the target position from the position showing the subject's most recent state. This is the risk prediction device described in Appendix 12.

[0109] Furthermore, some or all of the configurations described in Appendices 2-8 and 11-15, which are dependent on Appendice 1 above, may also be dependent on Appendices 9 and 10 in the same way as Appendices 2-8 and 11-15. Moreover, not limited to Appendices 1, 9, and 10, some or all of the configurations described as appendices may also be dependent on various hardware, software, various recording means for recording software, or systems, without departing from the embodiments described above.

[0110] While the present disclosure has been described above with reference to embodiments, the present disclosure is not limited to the embodiments described above. Various modifications to the structure and details of the present disclosure may be understood by those skilled in the art within the scope of the present disclosure. That is, the present disclosure includes the entire disclosure, including the claims, and of course, various modifications and alterations that those skilled in the art may make in accordance with the technical idea. [Explanation of Symbols]

[0111] 1 server 2. Target user's device 11, 21 Interfaces 12, 22 processors 13, 23 memory 14, 24 Recording media 15, 25 Display section 16, 26 Input section 41 Personal Information Acquisition Department 42 Life Log Acquisition Unit 43 Life Map Creation Department 44 State Estimation Unit 45 Deviation detection unit 46 Risk Prediction Department 47 Feedback Department 48 Visualization Map Creation Department 49 Action Records Department 50 Execution Determination Unit 51 Alert Output Section 52 Action Proposal Department 100 Risk Prediction Systems

Claims

1. A means of acquiring life logs, including sleep quality, A life map creation method that creates a life map representing lifestyle trends based on the aforementioned life log, A state estimation means that estimates the subject's recent state based on the subject's recent life log, A risk prediction means that predicts the risk of a decline in the sleep quality of the subject based on the life map and the subject's recent condition, If the aforementioned risk exceeds a threshold, a feedback means for providing feedback to the subject, A risk prediction device equipped with the following features.

2. The life map includes deviation detection means that detects when the subject's recent state deviates from the subject's normal state and acquires deviation information related to the deviation. The life log acquisition means acquires the life logs of the subject at multiple points in the past, The life map creation means creates a life map representing the lifestyle trends of the subject based on the subject's past life logs. The risk prediction device according to claim 1, wherein the risk prediction means predicts the risk based on the deviation information.

3. The aforementioned subject is one of several users, The life map includes deviation detection means that detects when the subject's recent state deviates from the normal state of users of the same gender and age as the subject, and acquires deviation information related to the deviation. The life log acquisition means acquires life logs of multiple users at multiple points in the past, The life map creation means creates a life map that represents the lifestyle trends of users of the same gender and age group as the target user, based on the life logs of multiple users. The risk prediction device according to claim 1, wherein the risk prediction means predicts the risk based on the deviation information.

4. The risk prediction device according to claim 1, wherein the feedback means outputs an alert to the terminal used by the subject when the risk is above a threshold.

5. A visualization map creation method that creates a visualization map that visualizes the relationship between lifestyle trends and subjective evaluations of sleep quality based on the aforementioned life log, The system includes an action suggestion means that, when the aforementioned risk exceeds a threshold, suggests actions to improve sleep quality based on the visualization map and the subject's recent condition, The risk prediction device according to claim 4, wherein the feedback means outputs advice indicating actions to improve sleep quality to a terminal used by the subject.

6. A behavioral recording means that records the actions taken by the subject based on the subject's life log during a predetermined period after suggesting actions to improve sleep quality, The system includes an execution determination means that compares the actions performed by the subject with the actions proposed to the subject, and determines whether or not the subject performed the proposed actions. The risk prediction device according to claim 5, wherein the action suggestion means suggests the next action to improve sleep quality based on the result of the determination, the visualization map, and the subject's recent condition.

7. The aforementioned life log includes information on several parameters that represent lifestyle trends, The life map and the visualization map are based on two or more parameters, each with an axis. The aforementioned parameters include information about the weather for one day, The life map creation means creates a life map using the weather changes, The risk prediction device according to claim 6, wherein the action suggestion means suggests an action that reflects the change in weather.

8. The risk prediction device according to claim 1, wherein the risk prediction means predicts the risk of the subject's sleep quality decreasing by using a machine learning model that has been trained to output an optimized numerical value indicating the risk in response to input of the life map and data relating to the subject's recent condition.

9. A risk prediction method performed by a risk prediction device, By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. A risk prediction method that provides feedback to the subject if the aforementioned risk exceeds a threshold.

10. A program executed by a risk prediction device equipped with a computer, By collecting lifestyle logs, including sleep quality, Based on the aforementioned life log, a life map representing lifestyle trends is created. Based on the subject's most recent life log, the subject's most recent condition is estimated. Based on the aforementioned life map and the subject's recent condition, the risk of a decline in the subject's sleep quality is predicted. A program that causes the computer to perform a process to provide feedback to the subject if the aforementioned risk exceeds a threshold.

Images