JP7395423B2 - Lifestyle monitoring device, system, program, and method capable of providing information related to meals, sleep, and exercise - Google Patents

Description

The present invention relates to a technique for determining and providing information regarding a living body from data detected and measured from the living body.

In recent years, technologies have been developed that use sensors to detect various biological signals caused by various activities of humans and animals, and utilize the biological data obtained through signal processing in various situations. Here, as the sensor, for example, a wristwatch type pulse sensor, an earphone type pulse sensor, a headband type electroencephalogram sensor, etc. are used. Furthermore, biological signals detected by such sensors are processed and processed by, for example, a smartphone carried by the user, and are used in various applications.

Recently, especially due to the increasing number of elderly people and increasing health consciousness, attempts have been made to analyze and evaluate the lifestyle habits and health conditions of subjects from the obtained biological data.

For example, Patent Document 1 discloses a matrix display method applicable to health management that makes it easy to visually understand health conditions related to dementia and lifestyle-related diseases. Specifically, this display method displays the answers to questionnaire items related to lifestyle habits and symptoms, as well as data on sleep types and sleep disorders obtained from sleep tests, in a matrix on the vertical and horizontal axes, and displays individual answer results. The data are displayed separately so that the degree of good or bad for each item of sleep data can be grasped, and the degree of good or bad for each item of sleep data is also displayed separately so that the degree of good or bad for each item of sleep data can be grasped.

Furthermore, Patent Document 2 proposes a server system that allows users to easily know their dementia risk and contributes to the prevention of dementia. Specifically, in this server system, (1) the user's exercise information is acquired by communicating with the user's information communication terminal 110 via the network 130, and (2) the acquired exercise information is recognized as exercise information. A statistical test is performed on the data group associated with the dementia risk, the user's dementia risk is evaluated, and (3) a process of providing the dementia risk evaluation result to the information communication terminal 110 is performed.

Further, Patent Document 3 discloses an augmented reality system for user health analysis. The health analysis method implemented in the augmented reality system includes collecting data for an initial predictive model, continuing to collect additional data based on one or more data criteria, and further comprising: The method may include updating the initial predictive model and generating a revised predictive model based on the additional data collected. The augmented reality system also includes one or more sensors and a display device that outputs light with variable wavefront divergence that is configured to present virtual content to a user.

Japanese Patent Application Publication No. 2018-156558 Japanese Patent Application Publication No. 2018-191722 Special table 2019-517849 publication

As described above, various methods have been proposed for analyzing and evaluating a subject's lifestyle and health condition and presenting the results. However, the reality is that these conventional techniques have not yet been able to appropriately provide comprehensive information for leading a healthy life to subjects.

For example, it has been known for a long time that proper diet, exercise, and sleep every day are essential for maintaining good health. Therefore, it is thought that comprehensive information for leading a healthy life can be generated by comprehensively analyzing these implementation situations.

On the other hand, the technology described in Patent Document 1, for example, certainly allows for understanding the risks and causes of lifestyle-related diseases and dementia from the answers to questions about lifestyle habits, including the oral environment, and sleep data from sheet-type sleep sensors. Presented in an easy-to-understand format. However, since collection of data related to lifestyle habits other than sleep data relies on interviews, it is practically difficult to check these lifestyle habits, for example, every day.

Further, in the technique described in Patent Document 2, a statistical test is performed on the acquired exercise information using a data group in which exercise information and dementia risk are associated. In other words, the risk of dementia has not been determined based on lifestyle activities other than exercise, such as dietary habits, and there is currently no comprehensive information available to help people lead a life that can reduce the risk of dementia. has not yet been reached.

Furthermore, the technology described in Patent Document 3 does indeed provide augmented reality information about the user's health condition to a head-mounted display equipped with an audio microphone sensor that detects multiple types of biological signals, including chewing sounds and heart sounds. This enables the display to be performed with However, it has not yet come to the point of presenting comprehensive information for leading a healthy life. For example, it is now possible to analyze heartbeat, chewing, and even sleep patterns to determine whether a person has a medical condition that makes chewing difficult or whether they are eating or not. Regarding generation methods, it only lists the types of machine learning that can be used.

Therefore, the present invention provides a lifestyle monitoring device, system, program, and method that can generate and provide more comprehensive information regarding a user's lifestyle based on the amount of multiple lifestyle-related items including the user's meals. The purpose is to

According to the present invention, there is provided a lifestyle monitoring device for monitoring a user's lifestyle, comprising:
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from an output of an activity meter worn by the user;
Based on the determined meal-related amount and the determined exercise-related amount, information regarding the user's diet and exercise-related lifestyle includes whether or not the user is walking while chewing a predetermined luxury item. Provided is a lifestyle monitoring device having feedback means for generating or selecting and providing information related to the above .
According to the present invention, there is also a lifestyle monitoring device that monitors a user's lifestyle, comprising:
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
a mental state related amount determining means for determining a mental state related amount including an amount related to a smile of the user from the output of the myoelectric sensor;
Determine sleep-related amounts, including amounts related to sleeping and/or waking up of the user, and exercise-related amounts, including amounts related to exercise, including walking, of the user, from the output of the activity meter worn by the user. Activity meter related quantity determination means;
Based on the determined meal-related amounts, mental state-related amounts, sleep-related amounts, and exercise-related amounts, information related to the user's diet and sleep-related lifestyle, and information on the user's lifestyle related to exercise and mental status are provided. and feedback means for generating or selecting and providing such information,
The information related to the user's life related to meals and sleep is information related to judgment or advice regarding the relationship between the time related to meals and the time related to sleep, and is information related to the user's life related to exercise and mental state. A life monitoring device is provided in which the information is information regarding whether or not the user is determined to be smiling while walking.

As a preferred embodiment, the lifestyle monitoring device according to the present invention further includes a mental state related amount determining means for determining a mental state related amount including an amount related to the user's smile from the output of the myoelectric sensor. death,
The activity meter related amount determining means determines an exercise related amount including an amount related to walking of the user ,
The feedback means also provides information regarding the user 's lifestyle related to exercise and mental state based on the determined exercise -related amount and mental state-related amount ,
It is also preferable that the information related to the user's lifestyle related to exercise and mental state is information related to whether or not the user is determined to be smiling while walking .

Further, as another embodiment of the life monitoring device according to the present invention, the activity meter related amount determining means determines both the sleep related amount including the amount related to going to bed and/or waking up of the user , and the exercise related amount. death,
The present life monitoring device includes a mental state related amount determining means for determining a mental state related amount including an amount related to the smile of the user from the output of the myoelectric sensor ;
For each of the determined meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts, a deviation score is calculated based on the degree of deviation from the set reference value, and a warning score is obtained by integrating the deviation scores. and a warning score determining means for determining the
Preferably, the feedback means also provide information regarding the determined warning score.

Furthermore, in the above-described embodiment related to the warning score, the warning score determining means uses a modified sigmoid function whose variables are quadratic, including a lower limit reference value and an upper reference value as the reference value, to determine the deviation. It is also preferable to calculate a score.

In still another embodiment of the life monitoring device according to the present invention, the activity meter related amount determining means determines both the sleep related amount including the amount related to going to bed and/or waking up of the user , and the exercise related amount. decide,
The present life monitoring device includes a mental state related amount determining means for determining a mental state related amount including an amount related to the smile of the user from the output of the myoelectric sensor ;
Life pattern information that generates life pattern information that is information related to the lifestyle of the user from a plurality of meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts determined for a predetermined period. further comprising a generating means,
It is also preferable that the feedback means also provides information related to the generated lifestyle pattern information.

Furthermore, as a specific example of information provision in the life monitoring device according to the present invention, the activity meter related amount determining means determines a sleep related amount including an amount related to going to bed and/or waking up of the user ,
The feedback means generates or selects and provides information related to the user's diet and sleep-related lifestyle based on the determined meal-related amount and the determined sleep-related amount,
Information related to the user's life related to meals and sleep shall be information related to the result of comparing the time difference between the time when dinner ends and the time when the user plans to go to bed or the time when he or she went to bed with a predetermined reference value. is also preferable.

Further, as another specific example of information provision in the life monitoring device according to the present invention, the activity meter related amount determining means determines a sleep related amount including an amount related to going to bed and/or waking up of the user ,
The feedback means generates or selects and provides information related to the user's diet and sleep-related lifestyle based on the determined meal-related amount and the determined sleep-related amount,
Information related to the user's daily life related to meals and sleep is based on the result of comparing the time when the user went to bed within a predetermined time range from the time the lunch ended with a predetermined reference value. It is also preferable that it is information.

Furthermore, the lifestyle monitoring device according to the present invention is a device that can be worn on the head of the user, and includes the myoelectric sensor and the activity meter,
The myoelectric sensor includes at least one left electrode that contacts at least one skin position between the user's left ear pinna and cheek area, and at least one left electrode that contacts at least one skin position between the right ear pinna area and cheek area of the user. the left electrode and the right electrode, the myoelectric signals generated due to at least one of mastication, clenching or clenching, and raising the corners of the mouth; detectable;
It is also preferable that the activity meter includes an acceleration sensor and a pulse sensor.

Furthermore, in the lifestyle monitoring device according to the present invention, it is also preferable that the information related to the lifestyle provided is advice information and/or praise information for the user.

According to the present invention, there is also a life monitoring system for monitoring the life of a user, comprising:
a myoelectric sensor located on the head of the user;
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of the myoelectric sensor;
an activity meter worn by the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from the output of the activity meter;
Based on the determined meal-related amount and the determined exercise- related amount , information regarding the user's diet and exercise-related lifestyle is obtained when the user walks while chewing a predetermined luxury item. A lifestyle monitoring system is provided that includes a feedback means for generating or selecting and providing information regarding whether or not the patient is present .

According to the present invention, there is further provided a program that causes a computer installed in a life monitoring device that monitors a user's life to function,
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from an output of an activity meter worn by the user;
Based on the determined meal-related amount and the determined exercise- related amount , information regarding the user's diet and exercise-related lifestyle is obtained when the user walks while chewing a predetermined luxury item. A life monitoring program is provided that causes a computer to function as a feedback means for generating or selecting and providing information regarding whether or not a person is present .

According to the present invention, there is also a lifestyle monitoring method implemented by a computer installed in a lifestyle monitoring device that monitors a user's lifestyle, comprising:
From the output of the myoelectric sensor located on the user's head, the user's meal-related amount, including the amount related to chewing, is determined, and from the output of the activity meter worn by the user , the user's meal-related amount is determined. determining motion-related quantities including walking -related quantities;
Based on the determined meal-related amount and the determined exercise- related amount , information regarding the user's diet and exercise-related lifestyle is obtained when the user walks while chewing a predetermined luxury item. A lifestyle monitoring method is provided which includes the step of generating or selecting and providing information regarding whether or not the child is present .

According to the lifestyle monitoring device, system, program, and method of the present invention, it is possible to generate and provide more comprehensive information regarding the user's lifestyle based on the amount of multiple lifestyle-related items including the user's meals. can.

1 is a schematic diagram showing an embodiment of a lifestyle monitoring system including a lifestyle monitoring device according to the present invention. It is a graph showing an example of calculating a deviation score according to the present invention. 2 is a table showing an example of monitor data for each day of one week in each monitor data item according to the present invention. 4 is a radar chart displaying the results of calculating the deviation score and warning score of user 1 and user 2 in the example shown in FIG. 3. FIG. This is a graph displaying lifestyle pattern information for one week generated for a certain user (wearer). 3 is a problem behavior correspondence table held by an embodiment of the feedback means according to the present invention. 3 is a problem behavior correspondence table held by an embodiment of the feedback means according to the present invention. 3 is a problem behavior correspondence table held by an embodiment of the feedback means according to the present invention. 3 is a praise opportunity correspondence table held by an embodiment of the feedback means according to the present invention. FIG. 3 is a schematic diagram showing another embodiment of the lifestyle monitoring device according to the present invention.

Embodiments of the present invention will be described in detail below with reference to the drawings.

[Lifestyle monitoring system]
FIG. 1 is a schematic diagram showing an embodiment of a lifestyle monitoring system including a lifestyle monitoring device according to the present invention.

The lifestyle monitoring system of this embodiment shown in FIG.
(a) "Head attachment 2" as a biological signal acquisition device,
(a) "Wristband 3" equipped with an activity meter 31,
(C) "Mobile terminal 1" as a life monitoring device according to the present invention, which monitors the life of a user (wearer) who wears "head attachment 2" on the head and wears "wristband 3" on the arm. ”. Incidentally, "mobile terminal 1" may be a terminal owned by the user (wearer), or may be a terminal owned by a third party (for example, a manager or staff of the facility to which the user belongs). It is possible.

Here, the "head attachment 2", the "wristband 3", and the "mobile terminal 1" are based on, for example, Bluetooth (registered trademark), Wi-Fi (registered trademark), or other short-range wireless communication standards. They are equipped with signal interfaces (214, 32, 101) that are compliant wireless communication means, and are communicatively connected to each other. As a modification, the "head attachment 2" and the "wristband 3" may be communicatively connected to the "mobile terminal 1" by wire. First, the structures and functions of the "head attachment 2" and the "wristband 3" will be explained, and then the "mobile terminal 1" which is a lifestyle monitoring device according to the present invention will be explained in detail.

<Head attachment>
First, the head attachment 2 described in (a) above is a wearable device that is attached to the head of a living body (for example, a human user) and is capable of acquiring biological signals. In this embodiment, the biological signal acquired by the head mounted device 2 is a myoelectric signal, which is an electrical signal generated due to a movement of an intrafacial region or a movement related to facial expression.

Here, in this embodiment, we particularly focus on "chewing" and "smiling" (lifting of the corners of the mouth) as the movements of facial parts to be detected or movements related to facial expressions, and the muscles related to "chewing" and "smiling" are It is characterized by efficient detection of electric signals. Incidentally, after this, these acquired myoelectric signals are used to determine the meal-related amount and the mental state-related amount of the user (wearer), respectively. In addition, in this embodiment, as will be described later, myoelectric signals related to "clenching" are also detected and used.

Similarly, in FIG. 1, the head attachment 2 specifically includes:
(a) When the head attachment 2 is attached to the head, the left ear hook part and the right ear hook part contact the bases of the left and right auricles and function as GND (ground) electrodes, and the rear of these parts. A frame 22 having elasticity and conductivity that can sandwich the head, and has a frame connecting part that goes around and connects the parts to the rear of the head.
(b) The left reference electrode part 23a attached to the front end of the left ear hook part, the left detection electrode part 22a attached to the end of the arm extending forward from this left reference electrode part 23a, and the right A right reference electrode part 23b attached to the front end of the ear hook part, a right detection electrode part 22b attached to the end of an arm extending forward from the right reference electrode part 23b,
(c) A left bone conduction speaker 24a and a right bone conduction speaker 24b, which are provided to overlap with the left reference electrode part 23a and the right reference electrode part 23b, respectively, and vibrate the skin position in contact with these electrode parts. and,
(d) a GND band 25 whose both ends are connected to the front part of the left ear hook part and the front part of the right ear hook part, respectively, has elasticity and conductivity, and also functions as a GND electrode;
(e) Signal processing electrically connected to the frame 22 (the left and right ear hooks thereof), the GND band 25 of (d) above, the electrode group of (b) above, and the left and right bone conduction speakers of (c) above. 21.

Here, as shown in FIG. 1, the left detection electrode section 22a and the right detection electrode section 22b are arranged to abut (contact) the skin position of the cheekbones in front of the left and right ear holes of the user's head and in front of the left and right sideburns, respectively. As a result, it is possible to detect good myoelectric signals related to "chewing" and "smiling" with reduced artifacts.

In the head attachment 2, a left reference electrode section 23a and a right reference electrode section 23b are further arranged so as to contact the skin position in front of the left and right ear holes of the user's head and behind the left and right sideburns, respectively. In this way, the head mount 2 has a comfortable fit that does not interfere with vision or hearing as there are no parts that come into contact with the eyes, nose, or mouth, and one channel on the left side of the head and one channel on the right side of the head. This allows for a useful myoelectric sensor system with a total of 2 channels.

As a modification, the head mounted device 2 embodies a 1-channel myoelectric sensor system using the left detection electrode section 22a and the right detection electrode section 22b as a detection (reference) electrode and a reference (detection) electrode, respectively. Alternatively, it is also possible to implement a 1-channel myoelectric sensor system using the left reference electrode section 23a and the right reference electrode section 23b as a detection (reference) electrode and a reference (detection) electrode, respectively. be. By detecting myoelectric signals across the left and right sides of the head using such a 1-channel electrode section, it is possible to capture the entire left and right muscle activity and obtain a more stable and large myoelectric signal.

Furthermore, as shown in FIG. 1, the signal processing section 21 (e) above has a built-in battery and a computer (processor/memory), and the computer uses the power supplied from the battery to and performs audio output processing for the user (wearer). More specifically, the signal acquisition section 21 includes a signal conversion section 211, a pre-filter processing section 212, a signal control section 213, a communication interface 214, an audio output section 215, and an audio amplification section 216. The computer functions to perform myoelectric signal processing and audio output processing.

Here, among these, the signal conversion section 211 is
(Left a) A left negative (reference) electrode electrically connected to the left reference electrode section 23a,
(Left b) AC component of the potential difference between the left detection electrode section 22a and the electrically connected left positive (detection) electrode,
(Left c) Amplify by differential amplification of the GND potential at the left GND electrode electrically connected to the frame 22 and GND band 25, and digitize this analog myoelectric signal at a constant sampling frequency. This is a circuit that generates left myoelectric signal (digital) data. This makes it possible to detect skin potential in the range of, for example, plus or minus 0.1 to several hundred μV.

Further, the signal conversion unit 211 further includes:
(Right a) A right negative (reference) electrode electrically connected to the right reference electrode section 23b;
(Right b) The alternating current component of the potential difference between the left detection electrode part 22b and the electrically connected right positive (detection) electrode,
(Right c) Amplify by differential amplification of the GND potential at the right GND electrode electrically connected to the frame 22 and GND band 25, and digitize this analog myoelectric signal at a constant sampling frequency. Generate right myoelectric signal (digital) data. Incidentally, the signal conversion unit 211 has a DRL (Driven Right Leg) circuit that reduces common mode noise caused by a commercial power source, etc., and the GND electrode may be used as the DRL electrode.

The signal conversion unit 211 then performs window division processing on each of the generated left and right myoelectric signal (digital) data, and converts the time-series myoelectric signal data into a form that can be sequentially analyzed in real time. This makes it possible to realize a user interface that provides real-time feedback to the user (for example, audio presentation using a bone conduction speaker). Furthermore, it becomes easier to use the myoelectric sensor output data in various application programs.

Specifically, in this window division process, a window analysis section is provided in advance, and the myoelectric signal data corresponding to the analysis section is sequentially extracted while shifting this analysis section. For example, when the sampling frequency for digitization is 512Hz, the window analysis interval is set to 128 samples, and the time-series myoelectric signal data is divided into every 0.125 seconds (every 64 samples), and each divided interval is Data extraction may be performed.

Similarly, in FIG. 1, the pre-filter processing section 212 includes a notch filter processing section and a low-pass filter (LPF) processing section. Among these, the notch filter processing section performs band elimination filter processing for reducing periodic noise related to the commercial power source on each of the left and right myoelectric signal data. Further, the LPF processing section performs LPF processing to remove high frequency noise on each of the left and right myoelectric signal data that has been subjected to band elimination filter processing. Here, the LPF processing section performs the LPF processing by performing high pass filter (HPF) processing on the myoelectric signal data using, for example, a DC blocker, and subtracting the result from the original myoelectric signal data. It may be implemented.

Here, the signal processing section 21 transmits the left and right myoelectric signal data preprocessed by the pre-filter processing section 212 to the mobile terminal 1 via the signal control section 213 and the communication interface 214. Note that the audio output section 216 and the audio amplification section 216 in the signal processing section 21 will be explained later when the feedback section 15 (FIG. 1) of the mobile terminal 1 is described.

<Wristband>
Similarly, according to FIG. 1, the wristband 3 has a built-in battery, an activity meter 31, and a computer (processor/memory) that functions as a signal processing/control unit 311, and uses the power supplied from the battery. , the activity meter 31 and the computer (signal processing/control unit 311) process the acquired activity amount signal data.

Here, in this embodiment, the activity meter 31 is a known device that is equipped with an acceleration sensor and a pulse sensor and is capable of measuring the amount of activity. For example, the activity meter 31 may be an activity meter manufactured by Fitbit. Although it is possible to adopt an activity meter that does not include a pulse sensor, it is also preferable to use data related to pulse rate when determining sleep-related amounts and exercise-related amounts, which will be described later. The activity meter 31 of this embodiment also includes a pulse sensor. The pulse sensor of the activity meter 31 uses a transmissive or reflective optical sensor to measure changes in blood flow (changes in blood vessel volume) that occur with the heartbeat in the capillaries near the wrist. It has a mechanism that captures the pulse.

Furthermore, the activity meter 31 can be set to sample acceleration at, for example, 100 Hz and measure pulse rate at, for example, 5 minute intervals. Furthermore, it is also preferable to set the pulse rate to be continuously measured when the acceleration sensor detects (acceleration signal indicating) movement such as walking. Further, the measured number of steps and pulse rate may be displayed on the display of the wristband 3 as appropriate. Furthermore, if the activity meter 31 detects a condition such as a low battery voltage, a dead battery, or an incorrectly attached battery, a display or voice output may be displayed to prompt correction of the operating status of the activity meter 31. preferable.

Next, the signal processing/control unit 311 calculates, for example, the following activity meter-related quantities (a) to (h) from the acceleration signal and/or pulse signal that are the output signals from the activity meter 31 using a known method. At least one, preferably all, of:
(a) Body position (lying, standing or sitting): Determined based on acceleration and pulse rate (b) Number of steps: Calculated based on acceleration (c) Walking time information (walking start time, walking end time): Determined based on acceleration (d) Walking pace (number of steps per unit time): determined based on acceleration (e) Exercise intensity METs (Metabolic Equivalents)
(f) Activity amount Ex (Exercise): Calculated based on acceleration, preset and input weight value, and pulse rate (g) Sleep time information (bedtime, wake-up time): Determined based on acceleration and pulse rate (h) Resting heart rate: Calculated and determined based on pulse rate and acceleration

Here, the exercise intensity METs in (e) above is defined as the oxygen intake amount due to the exercise to be measured (energy consumption due to the exercise) when the oxygen intake amount at rest is 3.5 ml/kg/min. For example, according to the well-known Karvonen method, the following formula (1) METs = ("Heart rate" - "Resting heart rate") / ("Maximum heart rate" - "Resting heart rate") number”)×100
This is the amount calculated by . Here, the maximum heart rate is estimated by "maximum heart rate" = 220 - "age". In general, it is said that an aerobic exercise effect can be obtained if the exercise intensity METs calculated by the Karvonen method calculated using the above formula (1) is 50 to 70. Furthermore, the activity amount Ex in (f) above can also be calculated from the exercise intensity METs as Ex=METs×"exercise duration (h)".

Note that the resting heart rate in (h) above is the pulse rate measured while the user is resting in a sitting position. After the (wearer) wakes up, the pulse rate when an acceleration of less than a predetermined value is measured continuously for a predetermined period of time (for example, 3 minutes) shall be considered as the resting pulse rate for the day. I can do it. Furthermore, in addition to the above-mentioned activity meter-related quantities, the signal processing/control unit 311 generates (i) target heart rate as an activity meter-related quantity: "target heart rate" = METs x ("maximum heart rate" - "Resting heart rate") + "Resting heart rate"
may be determined.

<Mobile terminal (life monitoring device)>
Similarly, according to FIG. 1, the mobile terminal 1 as a lifestyle monitoring device according to the present invention (c) above has the following features:
(A) Meals for which the "meal-related amount" including the amount related to chewing by the user (wearer) is determined from the output of the myoelectric sensor (of the head mounted device 2) located on the head of the user (wearer) A related quantity determination unit 111;
(B) From the output of the activity meter 31 (of the wristband 3) worn by the user (wearer),
(b1) “Sleep-related amount” including the amount related to sleeping and/or waking up of the user (wearer); (b2) “Exercise-related amount” including the amount related to the number of steps and/or activity intensity of the user (wearer) ”
an activity meter related quantity determining unit 12 that determines one or both of the above;
(C) Based on the determined "meal-related amount" and the determined "sleep-related amount" and/or "exercise-related amount" (c1) related to the user's (wearer's) diet and sleep-related lifestyle. and (c2) "meal/exercise related life information" which is information related to the user's (wearer's) life related to diet and exercise.
and a feedback unit 15 that generates or selects and provides one or both of them.

In this way, the mobile terminal 1 utilizes both the output of the myoelectric sensor of the head mounted device 2 and the output of the activity meter 31 of the wristband 3 to monitor multiple meals including the user's (wearer's) meals. Based on amounts related to life-related items, that is, "meal-related amounts," "sleep-related amounts," and/or "exercise-related amounts," more comprehensive information regarding the user's (wearer's) life, i.e., "meal-related amounts," "sleep-related amounts," and/or "exercise-related amounts"・Sleep-related daily life information" and/or "meal/exercise-related daily life information" can be generated and provided.

Here, "meal/sleep-related lifestyle information" and "meal/exercise-related lifestyle information" (information related to the lifestyle) are feedback information, and as will be explained in detail later, advice information and information for the user (wearer). It is also preferable that the information is/or praise information. In this case, for example, the user (wearer) can obtain useful information regarding his/her own diet, sleep, diet, and exercise, and can utilize this information in his/her daily life.

Naturally, the activity meter that provides activity meter-related amount data to the mobile terminal 1 is not limited to that mounted on the wristband 3. For example, as will be explained later with reference to FIG. 10, it may be located on the user's head. Similarly, the myoelectric sensor that provides myoelectric signal data to the mobile terminal 1 is not limited to that provided in the head mount 2. For example, the electric sensor is from the left electrode, the right anchle near the cheeks, at least one left electrode, and the right anchle, which contacts at least one skin position from the left of the nurtivity of the user (wearer) to near the cheeks. and at least one right electrode that contacts at least one skin location between the left and right electrodes, the left electrode and the right electrode cause the bite to occur due to at least one of mastication, clenching (clenching), and raising the corners of the mouth. Various methods can be used as long as they can detect myoelectric signals.

However, both the head attachment device 2 and the wristband 3 of this embodiment can be easily and easily worn and have a good wearing feeling, so they can be worn without placing a large burden on the user (wearer). , it is possible to reliably monitor myoelectric signals and life-related quantities. Also, in this embodiment, there is no need to use a camera (image sensor) that generates images of the user (wearer), so the psychological burden on the user (wearer), such as worries about privacy invasion, is reduced. It is also possible to do so.

Note that the head mount 2 and the wristband 3 are wearable devices that have a good wearing feel as described above, so the user (wearer) wears these devices all day long and wears the head mount 2. The wristband 3 can also generate and output, for example, myoelectric signal data and activity meter-related amount data for 24 hours each day. Alternatively, when sleeping, the user (wearer) may remove the head mount 2, which may get in the way, and wear only the wristband 3 for generating activity meter-related amount data related to sleep.

Furthermore, as another preferred embodiment, the mobile terminal 1 includes:
(D) A mental state related amount determination unit 112 that determines a "mental state related amount" including the amount related to the user's smile from the output of the myoelectric sensor (of the head mounted device 2)
The activity meter related amount determination unit 12 determines both the “sleep related amount” and the “exercise related amount”, and the feedback unit 15 determines the determined “meal related amount” and “sleep related amount”. ”, “exercise-related amount”, and “mental state-related amount”, based on at least one of the user’s (wearer’s) diet, sleep, exercise, and mental state. It is also preferable to provide such information such as "lifestyle information (related to meals, sleep, exercise, and mental state)." This makes it possible to provide even more comprehensive information regarding the user's (wearer's) life.

In this regard, it has long been empirically said that proper eating, exercise, sleeping, and even laughing often bring about good health. Based on the output of myoelectric sensors and activity meters, it provides ways to utilize this knowledge in daily life.

For example, in one embodiment, the mobile terminal 1 shown in FIG. The system generates and presents "lifestyle information" such as lifestyle advice information to prevent and reduce lifestyle-related diseases and dementia, and helps the user (wearer) live a healthy life both physically and mentally. We can also support you.

Here, we will explain about "dementia," which can become a major problem in a super-aging society, and which can be prevented or alleviated by the mobile terminal 1.

Dementia is broadly classified into Alzheimer's type, Lewy body type, vascular type, and frontotemporal type. First, Alzheimer's dementia is thought to be caused by the accumulation of amyloid β in the brain, and it is known that the degree of amyloid β accumulation changes depending on how you sleep. In fact, there are reports showing that people who sleep about 7 hours have a lower incidence of dementia than those who sleep less than 7 hours or more than 7 hours. . Therefore, the mobile terminal 1 contributes to reducing the risk of developing Alzheimer's type dementia by generating and providing lifestyle advice information (as lifestyle information) regarding actual sleeping hours to the user (wearer). is also possible.

Furthermore, Lewy body dementia is caused by the occurrence of Lewy bodies, and it has been reported that one of the characteristic symptoms is awakening during sleep. Therefore, by providing the user (wearer) with lifestyle advice information (as lifestyle information) generated based on sleep-related amounts related to the presence or absence of mid-way awakening and its degree, the mobile terminal 1 can improve the Lewy body type. It will also be possible to contribute to the early detection and early treatment of dementia.

Furthermore, the onset of vascular dementia can be reduced by preventing lifestyle-related diseases such as arteriosclerosis. Therefore, the mobile terminal 1 provides the user (wearer) with lifestyle advice information (as lifestyle information) related to lifestyle habits such as diet, sleep, exercise, and mental state (laughter) (as will be explained in detail later). By producing and providing vascular dementia, it is also possible to contribute to reducing the risk of developing vascular dementia. Incidentally, it is known that tau protein and the protein TDP-43 are involved in the onset of frontotemporal dementia, but the cause of its onset remains unknown.

[Device configuration, life monitoring program]
In the functional block diagram of FIG. 1, a mobile terminal 1 as an embodiment of the life monitoring device according to the present invention includes a communication interface 101, a myoelectric information storage section 102, an activity amount information storage section 103, and a touch panel display ( TP/DP) 104, a microphone 105, a speaker 106, and a processor/memory.

Here, this processor memory stores an embodiment of the lifestyle monitoring program according to the present invention, and also has a computer function, and executes the lifestyle monitoring process by executing this lifestyle monitoring program. do. Therefore, the life monitoring device according to the present invention may be a mobile terminal such as a smartphone as in this embodiment, but it may also be a cloud server, a non-cloud server, a personal computer, etc., equipped with the life monitoring program according to the present invention. (PC), or a notebook or tablet computer.

Further, the processor memory includes a myoelectric sensor related amount determining section 11 including a meal related amount determining section 111 and a mental state related amount determining section 112, and an activity amount including a sleep related amount determining section 121 and an exercise related amount determining section 122. It has a total-related amount determination section 12 , a warning score determination section 13 , a lifestyle pattern information generation section 14 , a feedback section 15 , an input/output control section 16 , and a communication control section 17 . Note that these functional components can be regarded as functions of a lifestyle monitoring program stored in the processor memory. Furthermore, the process flow shown by connecting the functional components of the mobile terminal 1 (life monitoring device) with arrows in the functional block diagram of FIG. 1 can be understood as an embodiment of the lifestyle monitoring method according to the present invention. .

Similarly, in the functional block diagram of FIG. 1, the myoelectric information storage unit 102 receives muscle information that has been subjected to pre-filtering processing from (the signal control unit 213 of) the head mounted device 2 via the communication interface 101 and the communication control unit 17. Receive, store and manage electrical signal data.

In addition, the activity information storage unit 103 stores the determined activity meter related amount data (for example, the above (a)) from (the signal processing/control unit 311 of) the wristband 3 via the communication interface 101 and the communication control unit 17. ~ (i) data) is received, stored and managed.

Similarly, in the functional block diagram of FIG. 1, the myoelectric sensor-related amount determining unit 11 (meal-related amount determining unit 111 and mental state-related amount determining unit 112) appropriately selects information from the user (wearer) from the myoelectric information storage unit 102. The myoelectric signal data of is extracted, and the data is analyzed to determine (a) a meal-related amount including an amount related to mastication, and (b) a mental state-related amount including an amount related to smiling for the user.

Specifically, the myoelectric sensor related quantity determining unit 11, for example, as described in Japanese Patent Application Laid-Open No. 2019-107067, which is a patent application in which the inventor of the present application is the inventor,
(a) Generate acceleration component data of myoelectric signal data subjected to pre-filter processing,
(b) Calculate the representative value SD _W (specifically, the value determined from the standard deviation) related to the degree of data bias in the window analysis section in the acceleration component data,
(c) Based on the magnitude of the representative value SD _W , it is possible to determine the occurrence of myoelectric signals related to chewing or smiling (raised corners of the mouth). When the occurrence of mastication is determined here, when the time series data of the representative value SD _W shows a predetermined hysteresis N times (N is a positive integer), it may be determined that mastication has been performed N times.

Alternatively, the myoelectric sensor related quantity determination unit 11 may, for example, as described in Japanese Patent Application Laid-Open No. 2019-115410, which is a patent application in which the inventor of the present application is the inventor.
(d) Perform resonator filter processing on the time series data of the representative value SD _W in (b) above,
(e) Based on the periodicity of the representative value SD _W subjected to the resonator filter processing, it is also possible to determine whether a myoelectric signal is generated due to repeated mastication movements. Here, the peak position when the representative value SD _W fluctuates beyond a predetermined range is calculated, and if the time interval between adjacent peak positions is within a predetermined time range, it is determined that a myoelectric signal related to mastication has occurred. It is also preferable to make a judgment.

As a further preferable modification, the myoelectric sensor-related quantity determination unit 11 may, for example, as described in Japanese Patent Application No. 2019-098323, which is a patent application in which the inventor of the present application is the inventor.
(f) Sequentially import the time series data of the representative value SD _W in (b) above, and based on the imported data values, set the lower reference value corresponding to the minimum value of the data value and the maximum value of the data value. (a) The lower reference value and the subsequent upper reference value are determined or updated, and a data value smaller than the upper reference value that satisfies a predetermined condition is imported. or (a) when the upper standard value and the subsequent lower standard value are determined or updated, and a data value that is larger than the lower standard value that satisfies the predetermined condition is imported. It is also preferable to assume that a myoelectric signal related to chewing (or smiling) is generated, and to count the wave number of the myoelectric signal.

It is also preferable that the myoelectric sensor-related amount determination unit 11 determines the type of the generated myoelectric signal, that is, whether the generated myoelectric signal is related to mastication or smiling (raised corners of the mouth). For example, as described in JP 2018-139630, JP 2019-107067, and JP 2019-115410 related to patent applications in which the present inventor is the inventor,
(g) Calculate the feature amount including the representative value SD _W and the average power frequency in the time interval related to the determination that the myoelectric signal has occurred, and calculate the feature amount including the representative value SD W and the average power frequency, and use the relevant feature amount of the input signal that corresponds to the reference state for the feature amount. The degree of separation, which is the degree of separation from the unit space set by the amount, is calculated, and the type of the generated myoelectric signal can be determined based on the degree of separation.

More specifically, from the degree of separation from the unit space related to the reference state where no myoelectric signal is generated, the standard that combines the state where the myoelectric signal is generated and the state where no myoelectric signal is generated is determined. Based on the amount obtained by subtracting the degree of separation from the unit space related to the state and the degree of separation from the unit space related to the reference state in which the myoelectric signal was generated, the type of the generated myoelectric signal, that is, the myoelectric signal related to mastication is determined. It may also be possible to determine whether the signal is a smile or a myoelectric signal related to a smile.

In addition, as the above unit space and separation degree,
(a) Unit space in MT (Mahalanobis Taguchi) method and value calculated from Mahalanobis distance;
(a) Unit space in the MTA (Mahalanobis-Taguchi Adjoint) method and a value calculated from the Mahalanobis distance;
(c) It is possible to adopt a value calculated from the unit space and characteristic value in the T method, or (d) a value calculated from the unit space and RT distance in the RT (Recognition Taguchi) method. .

Further, by using the myoelectric signal data from the head mounted device 2 constituting the 2-channel myoelectric sensor system, the myoelectric sensor-related amount determining unit 11 can obtain, for example, a patent patent whose inventor is the present inventor. As stated in the specification of Japanese Patent Application No. 2019-207369,
(a) Calculate the left side representative value SD _W related to the degree of myoelectric signal generation for the myoelectric signal data from left channel 1 at each predetermined time interval,
(b) Calculate the right side representative value SD _W related to the degree of myoelectric signal generation for the myoelectric signal data from the right side 1ch at each predetermined time interval,
(c) Based on the difference between the left side representative value SD _W and the right side representative value SD _W , determine the side (left side, right side, or both sides) on which a strong myoelectric signal of a predetermined level or higher is generated;
(d) recording the time course of myoelectric signal data on the determined side (left side, right side, or both sides) every time a predetermined movement occurs in the user (wearer);
(e) It is also preferable to determine the body habit of the user (wearer) related to left-right bias based on the time course of myoelectric signal data recorded for a plurality of predetermined movements.

By performing such processing, the myoelectric sensor-related quantity determination unit 11 can determine, for example, the presence or absence of "biased mastication" in which the chewing motion tends to be biased to the left or right side of the mouth, It is also possible to determine whether or not a ``one-sided smile'' is likely to be performed.

In any case, by analyzing the myoelectric signal data as described above, the myoelectric sensor related quantity determination unit 11 determines whether "(biased) mastication" has been performed or not, and whether "(unilateral) smiling" has occurred. It is possible to determine whether or not the event occurred, as well as information regarding their actions, number of occurrences, and counts. In the present embodiment, the myoelectric sensor-related quantity determining unit 11 uses the above-described analysis method to determine whether or not "bite binding" has occurred (whether or not a myoelectric signal related to biting binding has occurred). ), and information regarding the number of occurrences/counts thereof can also be determined.

In addition, the myoelectric sensor related amount determining unit 11 calculates the myoelectric integral value by adding together the product of the myoelectric amplitude value exceeding a predetermined threshold and the sampling time in the time interval determined as chewing or smiling, The electromyogram integral value may be used as the amount of chewing or the amount of smile generation. Here, the predetermined threshold value is set to a value that can remove regularly mixed noise. Furthermore, it is also possible to calculate the similar integral value from the above-mentioned representative value SD _W to determine the amount of chewing and the amount of smile occurrence. In addition, by dividing the calculated amount of chewing and smile occurrence by the time interval determined to be chewing or smiling, the size and degree of mastication movement and the size and degree of smile in the relevant time interval can be determined. You can also do it.

Next, the meal-related amount determination section 111 of the myoelectric sensor-related amount determination section 11 determines, for example, the following meal-related amount from the above-described analysis results.
(a) Meal start time, which is the time at which chewing was determined for the first time after a predetermined time has elapsed from the end time of the last meal; (b) Chewing is determined for a predetermined period of time (for example, 10 minutes) from the time at which chewing was determined for the last time. (c) Meal time, which is the time difference between the meal start time and the next meal end time (d) One continuous mastication action per mouth , the mastication time of one mouthful, which is the difference between the time when the first chewing was determined and the time when the last chewing was determined, and the average chewing time per one mouthful (e) for a predetermined period (for example, one day). ), the mastication time in the relevant predetermined period as the aggregate value of the mastication time for each bite

(f) The number of times of mastication, which is the number of times of mastication in one mouth, where one continuous chewing action is considered as one mouth; and the average number of times of mastication per one mouth; Total number of times of mastication in the specified period as a total value of the number of times of chewing (h) Interval time which is the average value of the time interval between one bite and the next one during meal time (i) Predetermined period (for example, one day) Mastication pace (mastication tempo) calculated by dividing the number of times of mastication by the mastication time
(j) Breakfast intake time, which is the chewing time during breakfast.

By the way, the chewing times in (d) and (e) above and the number of mastications in (f) and (g) above are generally correlated, so by calculating the other amount from the determined amount of either one, Good too. For example, if chewing is performed at a cycle of 0.7 seconds, 10 minutes of chewing time corresponds to 857 times of mastication (≒1/0.7×60×10).

Further, the mental state related amount determining unit 111 of the myoelectric sensor related amount determining unit 11 determines, for example, the following mental state related amount from the above-described analysis results.
(a) Number of smiles, which is the number of times it was determined to be a smile in a predetermined period (for example, one day); (b) Smile time, which is the time interval that was determined to be a smile in a predetermined period (for example, one day)

Similarly, in the functional block diagram of FIG. 1, the sleep-related amount determining section 121 of the activity meter-related amount determining section 12 receives activity meter-related amount data of the user (wearer) from the activity amount information storage section 103 as appropriate; The data is analyzed to determine sleep-related amounts including amounts related to going to bed and/or waking up for the user. The sleep-related amounts that are specifically determined are, for example, as follows.
(a) Bedtime and wake-up time (c) After-dinner sleep onset time, which is the time difference between the end of dinner time and the next bedtime (b) Sleep time, which is the time difference between bedtime and the next wake-up time (c) Sleep Awakening time during sleep (d), which is the waking time (time in non-sleep state) during judgment; Variation in bedtime (for example, the absolute value of the difference between bedtime and critical bedtime) in a predetermined period (for example, one week) standard deviation), and variation in wake-up time (for example, standard deviation of the absolute value of the difference between the wake-up time and the scheduled wake-up time)

Similarly, in the functional block diagram of FIG. 1, the exercise-related amount determination section 122 of the activity meter-related amount determination section 12 receives the user's (wearer's) activity meter-related amount data from the activity amount information storage section 103 as appropriate; The data is analyzed to determine exercise-related quantities, including quantities related to the number of steps and/or activity intensity of the user. The motion-related quantities that are specifically determined are, for example, as follows.
(a) Number of steps in a predetermined period (for example, one day) (b) Walking start time and end time for each walking movement in a predetermined period (for example, one day)

(c) Walking time, which is the time difference between the walking start time and the walking end time for each walking movement; (d) The total walking time for each walking movement in a given period (for example, one day) as the aggregated value of the walking time for each walking movement. Walking time (e) Walking pace (number of steps per unit time) during a predetermined period (for example, one day)
(f) Exercise intensity METs in a predetermined time interval (e.g. daily training time)
(g) Activity amount Ex (Exercise) during a predetermined period (for example, one day)
(h) Exercise time in a given period (e.g. 1 day)

The various related amount determination processes in the myoelectric sensor related amount determining section 11 and the activity meter related amount determining section 12 have been described above. In this embodiment, in addition to the above-mentioned related quantities, these functional components also provide user-reported information (numerical values) regarding the following items (related to dementia countermeasures), for example, on a touch panel display by the user. 104 (operation of a displayed slider icon, etc.).
(a) Work activities: (none) 0 ←｜→ 1 (well implemented)
(b) Conversation with people: (none) 0 ←｜→ 1 (well implemented)
(c) Meal: (sluggish) 0 ←｜→ 1 (good eating)
(d) Sleep: (insomnia) 0 ←｜→ 1 (good sleep)
(e) Exercise: (none) 0 ←｜→ 1 (I worked as hard as I could)
(f) Laughter: (none) 0 ←｜→ 1 (I laughed so hard)
(g) Forgetfulness: (severe) 0 ←｜→ 1 (none)
(h) Defecation: 0 (none) / 0.33 (hard stool) / 0.67 (soft stool) / 1 (easy stool)
(i) Weight: (text input or slider icon operation)
(j) Systolic blood pressure: (text input or slider icon operation)
(k) Diastolic blood pressure: (text input or slider icon operation)
(l) Diary: (text input or selection of selection item icon)

<Warning score determination>
Similarly, in the functional block diagram of FIG. 1, the warning score determination unit 13 calculates the meal-related amount, sleep-related amount, exercise-related amount, and mental state-related amount (hereinafter also referred to as monitor data) determined as described above. For each, a "deviation score" related to the degree of deviation from the set "reference value" is calculated, and a "warning score" is determined by combining the "deviation scores".

Here, the "deviation score" detects behavior (problem behavior) that may be a problem for the user (wearer), for example, from the ideal lifestyle of "chewing well, walking well, sleeping well, laughing well", and This is an indicator that allows you to get an overview of the situation. Furthermore, in this embodiment, the "reference value" used to calculate the "deviation score" is determined without assuming that the monitor data necessarily follows a normal distribution. For standard values for conventional medical examinations, etc., this defines a range of ±2σ (σ is the standard deviation) of the measured value of a healthy (deemed) person, for example, on the assumption that the diagnostic results follow a normal distribution. However, in this embodiment, the monitor data (meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts) do not necessarily follow a normal distribution.

Specifically, the "reference value" can be statistically calculated, for example, from monitor data determined for a large number of users (wearers). For example, out of a large number of users (wearers), the monitor data of those who are determined to be healthy through a general health checkup is tested to determine the degree of agreement with multiple known probability distribution models. One of these values may be adopted and, for example, the upper limit value and lower limit value that define a range of approximately 95% of the distribution may be used as the "reference value." At this time, it is also preferable to classify the monitor data by attribute information of the user (wearer), such as age and gender, and to set a "reference value" for each such attribute type.

In addition, if it is difficult to test the probability distribution model as described above, for example, divide the possible range of monitor data into N equal parts (N is an integer of 2 or more), and calculate the monitor data that corresponds to each divided interval. The frequency is calculated to determine the frequency distribution, and the upper limit defines the frequency range between 2.5% of the lower limit of the overall frequency and 2.5% of the upper limit (5% frequency when both are combined) The value and the lower limit value may be used as a "reference value."

Furthermore, as a modification, the "reference value" may be set based on the findings of an epidemiological survey. Here, we set strict "standard values" (so that the normal range is narrow) for monitoring items for which the risk is clearly determined to be high through epidemiological surveys, and on the other hand, we determine that the risk is recognized but the degree is low. The "reference value" may be set loosely and widely for the monitored item. Furthermore, it is also possible for the setter to set the "standard value" based on experience, without relying on distribution models, epidemiological surveys, etc.

Next, the warning score determination unit 13 calculates and determines a deviation score for each monitor data item using the set reference value.

For example, if the monitor data value is x (≧0), the lower limit reference value is c1 (≧0), the upper limit reference value is c2 (≧0), and c is a set constant, the deviation score r is calculated by the following formula: It can be calculated using
(2) r＝log ₁₀ (xt)－c
(3) xt=(x-c1)×(x-c2)
However, when xt is less than 1 (xt<1) in equation (3), xt substituted into equation (2) is 1 (xt=1).

Here, a specific example of the deviation score r when c=0 will be described later using FIGS. 2(A) and 2(B). Incidentally, even in the case of c>0, if the calculated r takes a negative value (r<0), it is also preferable to consider r=0. In any case, by setting the c value using the above equation (2), it is possible to prevent the deviation score r from rising rapidly due to a minute deviation in the monitor data. Furthermore, the base of the logarithm in the above equation (2) can be set to any value while paying attention to the range of possible values of r.

As another preferred modification, the warning score determination unit 13 uses the set reference value to determine whether there are two variables, including the lower limit reference value cl and the upper reference value c2 as reference values, for each monitor data item. It is also preferable to calculate the deviation score r using the following modified sigmoid function. Specifically, in this case, the warning score determination unit 13 uses the following formula (4) r=1/(1+exp{η×(x−c1)×(x−c2)}), where the gain is η.
The deviation score r can be calculated by In this way, by calculating the discrepancy score using the transformed sigmoid function, the discrepancy score r can be kept within the range of 0 to 1, regardless of the scale of the monitor data, and as a result, This simplifies calculations for outliers.

Here, a specific example of the deviation score r when η=0.005 will be described later using FIGS. 2(C) and (D). In any case, by setting η to a sufficiently small value near 0, the lower limit can be set so that the deviation score does not change suddenly at the lower limit reference value cl or the upper reference value c2 and immediately take the value 1. It is also possible to make the slope of the graph more gentle near the reference value cl and the upper reference value c2.

As yet another modification, the warning score determining unit 13 may perform the following formula (5) r=tanh{η×(x−c1)×(x−c2)}+c
The deviation score r may be calculated by Here, tanh(x) (hyperbolic tangent) is tanh(x) = (exp(2x) - 1) / (exp(2x) + 1), so x is a value that can take the entire range of real numbers. It takes a value in the range of -1 to 1 even if x = 0, and takes a value of 0 when x = 0. Therefore, for example, by setting c=1, the deviation score r in the above equation (5) will have a value of 1 at x=c1 or x=c2 (η×(x-c1)×(x-c2)=0). It can be set so that the possible values fall within the range of 0 to 2.

Furthermore, if the risk ratio exceeding the upper and lower limits is known through an epidemiological survey, it is also preferable to obtain a deviation score by multiplying the r value calculated as described above by this risk ratio.

Additionally, if the monitor data matches a predetermined probability distribution model for each monitor data item, check which position in the probability distribution the monitor data value corresponds to, and determine the probability value p at this corresponding position in the probability distribution. Then, the following formula (6) r=1−p
It is also possible to calculate the deviation score r. By the way, if the monitor data does not match any probability distribution model, divide the possible range of the monitor data into N equal parts (N is an integer greater than or equal to 2) and calculate the frequency of the monitor data that falls into each divided interval. The frequency distribution may be determined in advance, the relative frequency (empirical probability value) p in the interval to which the monitor data corresponds is obtained, and the deviation score r may be calculated using the above equation (6).

FIG. 2 is a graph showing an example of calculating a deviation score according to the present invention.

First, Figures 2 (A) and (B) show monitor data (items) such as "sleeping time," "smiling time," "exercise time," "chewing time," "dispersion of wake-up time," "breakfast intake time," and "dinner time." A graph showing the time dependence of the discrepancy score calculated using the above formula (2) (when c=0) for each of the post-sleep onset time and the waking time during sleep is shown. Next, FIGS. 2(C) and (D) show the monitored data of "sleeping time", "smiling time", "exercise time", "chewing time", "wake-up time variation", "breakfast intake time", "time to fall asleep after dinner" A graph showing the time dependence of the deviation score calculated using the above equation (4) (when η=0.005) for each of the "wake time during sleep" is shown.

Here, in both graphs, the vertical axis is the deviation score r (dimensionless), and the horizontal axis is the time (minutes) at which the item corresponding to the monitor data item was performed. Further, the lower limit reference value c1 and the upper limit reference value c2 as reference values are set as follows.
<<Sleep time>>
Lower limit standard value c1 (minutes) and upper limit standard value c2 (minutes) of sleep time:
c1=6.5×60, c2=8.5×60
<<Smile time>>
Lower limit standard value c1 (minutes) and upper limit standard value c2 (minutes) of total smiling time in a day:
c1=0.5×60, c2=3×60

<<Exercise time>>
Lower limit reference value c1 (minutes) and upper limit reference value c2 (minutes) of exercise time with exercise intensity of 3 METs or more per day:
c1=0.375×60, c2=4×60
<<Chewing time>>
Lower limit reference value c1 (minutes) and upper limit reference value c2 (minutes) of mastication time at the time of mastication judgment:
c1=30, c2=90
<<Wake-up time variation>>
In the case of one day, the lower limit reference value c1 (minutes) and the upper limit reference value c2 (minutes) of the absolute value of the difference (minutes) between the scheduled wake-up time and the actual wake-up time, and in the case of a predetermined period (for example, one week), Lower limit standard value c1 (minutes) and upper limit standard value c2 (minutes) for the standard deviation (minutes) of the absolute value (minutes) of the difference on each day:
c1=-60, c2=4×60
Note that the lower limit reference value c1 is set to a negative value so that the deviation score r=0 when the dispersion is 0 minutes.

<<Breakfast intake time>>
Lower limit reference value c1 (minutes) and upper limit reference value c2 (minutes) of chewing time at breakfast:
c1=10, c2=90
<<Time of falling asleep after dinner>>
Lower limit standard value c1 (minutes) and upper limit standard value c2 (minutes) of the difference (minutes) between bedtime and last meal end time:
c1=3×60, c2=8×60
<<Wake time during sleep>>
Lower limit reference value c1 (minutes) and upper limit reference value c2 (minutes) of wakefulness time during sleep, which is the wakefulness time (time in non-sleep state) during sleep judgment:
c1=-60, c2=30
Note that the lower limit reference value c1 is set to a negative value so that the deviation score r=0 when the waking time during sleep is 0 minutes.

Under the setting of the reference values (upper and lower reference limits) as explained above, in the graphs of Figures 2 (A) to (D), the deviation score r = 0 within the reference value range, and if the reference value range is slightly If it deviates from , the deviation score r will rise from zero. In addition, in the case where the deviation score r is significantly outside of this reference value range, the deviation score r (which is the case where the above formula (2) is adopted) is x60 (minutes)), and takes the value 1 in FIGS. 2(C) and (D) (this is the case where the above equation (4) is adopted).

In this way, no matter which of the above formulas (2) and (4) is adopted, for each monitor data item, the deviation score is such that the deviation from the preferred reference value range is reliably reflected in the value. Therefore, r can be suitably determined. In particular, when formula (4) is adopted, the maximum value of the deviation score r can be fixed to 1, so it is impossible for the deviation score r to become an extremely large value, and abnormal monitor data This also makes it possible to index values appropriately.

Incidentally, it is also possible to use the "number of times of mastication" instead of the above-mentioned "chewing time". In this case, the lower limit reference value c1 explained in "Chewing time" above is 2571 (≒30 x 60 x 1/0.7) times in terms of the number of mastications, and the upper limit reference value c2 is 7714 (≒90) in terms of the number of mastications. ×60×1/0.7) times. In addition, if the monitor data value is not the time (minutes) but the number of times of mastication per day, the lower limit reference value c1 and the upper limit reference value c2 can be set as follows:
c1=800×3, c2=1500×3
Can be set to .

With such settings, the deviation score r can be kept within a value of about 7 even if it deviates significantly from the reference value range. Also, although this is an extreme example, even if you continue chewing at a frequency of 0.7 seconds for 24 hours, the number of times you chew will be 123429 (≒ 1/0.7 x 60 x 60 x 24), and the deviation score will be around 10. be.

Furthermore, in the present embodiment, the warning score determination unit 13 determines a "warning score" that is a composite of the "deviation scores" determined as described above. For example, if the deviation score of monitor data item k in a certain day is rk, then the warning score tot' for that day is calculated by the following formula (6) tot' = (Σ _k rk) / (number of monitor data items)
It may be calculated by Here, Σ _k rk _is the summation of deviation scores rk for monitor data item k. Therefore, it is understood that the above formula (6) is a formula for calculating the average value of the deviation scores rk.

Alternatively, the warning score tot for a predetermined period (for example, one week or one month) is calculated using the following formula (7) tot = (Σ _k ave_rk + Σ _j rj_SD) / (total number of monitor data items including the number of monitor data items related to standard deviation) )
It can be calculated by Here, rj_SD is the deviation score for the predetermined period in the monitor data item j related to the standard deviation of the predetermined period, such as "wake-up time variation", and ave_rk is the deviation score for the monitor data item j related to the standard deviation above. It is the average value of the deviation score rk (for each day) in the predetermined period in the monitor data item k other than rk. Therefore, it is understood that the above formula (7) is a formula for calculating the average value of the deviation scores rk and kj_SD over the predetermined period.

An example in which the warning score was calculated will be described below with reference to FIGS. 3 and 4. Here, in this example, after calculating the deviation score r for a certain week (Sunday to Saturday) for user (wearer) 1 and user (wearer) 2, the deviation score r is calculated using the above formula (7). This is an example of calculating the warning score tot for one week.

FIG. 3 is a table showing an example of monitor data for each day of one week in each monitor data item.

Here, regarding the monitor data item "wake-up time variation" in the table shown in FIG. 3, the monitor data value is the absolute value of the difference between the wake-up time and the scheduled wake-up time on each day from Sunday to Saturday. Furthermore, in the rightmost column of the table, the average value of each monitor data item for the week (Sunday to Saturday) is recorded.

FIG. 4 is a radar chart displaying the results of calculating the deviation score and warning score for user 1 and user 2 in the embodiment shown in FIG.

Here, in FIG. 4(A), for each of users 1 and 2, using the table of FIG. 3, the deviation score r of each monitor data item is calculated by the above formula (2), and then the above formula (7) The results of calculating the warning score tot using . Here, in this radar chart, the outermost position has a score value of 0, and the innermost position has a score value of 5.

On the other hand, in FIG. 4(B), using the table in FIG. 3 for each of users 1 and 2, the deviation score r of each monitor data item is calculated by the above formula (4), and then the above formula (7) is The results of calculating the warning score tot using this method are shown. Here, in this radar chart, the outermost position has a score value of 0, and the innermost position has a score value of 1.

In both the radar charts of FIGS. 4A and 4B described above, the area of the polygon representing the score of user 2 is larger than the area of the polygon of user 1. As a result, it is understood that user 2 is leading a more favorable (healthier) lifestyle than user 1 from the viewpoint of the set monitor data items. Additionally, as a result, the warning score tot for user 2 is also smaller.

In this way, by displaying the calculated warning score tot and deviation score r in the form of a radar chart, it becomes possible to easily grasp trends and problems in the user's (wearer's) life.

Of course, it is also possible to create daily (daily) or even monthly (monthly) radar charts for the warning score tot and deviation score r. It is also possible to display time-series changes in the warning score tot and deviation score r using a line graph or bar graph. This makes it possible to easily discover items with risks. Further, from the viewpoint of grasping trends in general lifestyle behavior, only the time-series changes in the warning score tot may be displayed. Furthermore, it is also preferable to visualize data on other indicators such as body weight, blood pressure, resting heart rate, hearing ability, etc. using a line graph or a bar graph.

<Life pattern information determination>
Returning to the functional block diagram of FIG. 1, the lifestyle pattern information generation unit 14 generates meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts (monitor data) determined for a predetermined period (for example, one week). ), in this embodiment, "life pattern information" which is information related to the lifestyle of the user (wearer) is generated from three of the following: meal-related amounts, sleep-related amounts, and exercise-related amounts.

FIG. 5 is a graph displaying lifestyle pattern information for one week generated for a certain user (wearer).

The graph in Figure 5 shows the monitor data items for each day of the week (each day of the week): "Woke up (time)," "Breakfast start (time)," "Breakfast end (time)," "Walking start (time)," "Walking End (time),” “Start lunch (time),” “End lunch (time),” “Start nap (time),” “End nap (time),” “Start dinner (time),” “End dinner (time),” and “Go to bed. (Time)" (unit: hour) is shown. Here, the time values of these monitor data items are the average values (black circles) of the relevant values over the most recent weeks, and the graph also shows the error range of ±2σ (σ is the standard deviation) of the relevant values. Bars are also shown.

In addition, in the graph in Figure 5, a line connecting the time values (black circles) of each day (each day of the week) is attached for each monitor data item, making it easy to understand how each monitor data value changes over time. It is now possible to do so. That is, the user's (wearer's) lifestyle pattern can be easily viewed from the graph, and trends and problems in the lifestyle pattern can be grasped at a glance.

Note that although the graph of FIG. 5 does not list monitor data items related to mental state related quantities, such as "smiling time," it is of course possible to display this in a graph as well. For example, just above the scale line for each day (each day of the week), the total smiling time (minutes) for that day may be displayed, or for each day (each day of the week), it may be determined that a predetermined number of smiles or more were observed. The time zone may be displayed using a bar of a predetermined color. Furthermore, it is also preferable to display the average value of the time value of each monitor data item for one week in different colors, for example.

<Determine feedback information>
Returning to the functional block diagram of FIG. 1, the feedback unit 15 provides feedback to the user ( Generates, selects, and outputs lifestyle information (meals, sleep, exercise, mental state) related to at least one of the user's (wearer's) diet, sleep, exercise, and mental state. ) and the user's (wearer's) diet and exercise-related lifestyle information, which can be generated, selected, and provided. It has become.

Further, the feedback unit 15 may also output information related to the warning score determined by the warning score determination unit 13, and further output information related to the lifestyle pattern information generated by the lifestyle pattern information generation unit 14. It is also good.

Regarding the feedback information (lifestyle information, warning score information, lifestyle pattern information) output here, for example,
(a) Monitor data of each monitor data item (meal-related amount, sleep-related amount, exercise-related amount, and mental state-related amount), warning score, and analysis results of lifestyle pattern information,
(b) It is equipped with an output selection table that associates in advance the "display/audio output contents" that should be provided and notified to the user (wearer), and using this output selection table, " It is also preferable to select "display/audio output content" and output it as feedback information.

Further, the feedback information is displayed from the touch panel display 104 or the speaker 106 via the input/output control unit 16 in accordance with the user's (wearer's) instructions using the touch panel display 104 or the microphone 105, or in the form of a push. It may also be output as information or audio information (including an alarm sound to notify an emergency).

Furthermore, it is also preferable that the feedback information is transmitted to the SNS (Social Networking Service) server 4, external server 5, etc. via communication via the communication control unit 17 and the communication interface 101 and used. In this case, feedback information may be provided, for example, to the user's (wearer's) cohabitants, close relatives, relatives, family doctor, administrator, friends, or even a conversational robot in the form of SNS, email, or conversational audio. It is also possible to be notified and provided.

Furthermore, the feedback information is transmitted to the communication interface 214 of the head mounted device 2 worn by the user through communication via the communication control unit 17 and the communication interface 101, and the signal control unit 213 and the audio output unit 215 output the audio signal. , and further converted and amplified by the audio amplification unit 216 into an audio signal that can be input to the bone conduction speakers 24a and 24b, and output as audio information (including an alarm sound to notify an emergency). good.

The feedback information determination process in the feedback unit 15 will be described in detail below.

<Feedback information regarding meals>
The feedback unit 15 grasps the current state of the user's (wearer's) diet from the monitor data (meal-related amount, etc.) of each monitor data item, warning score, and lifestyle pattern information, and provides feedback regarding advice, praise, etc. Generate and select information (advice information, praise information, etc.). In this embodiment, for example, we recommend that you take your time to chew your food thoroughly, and also encourage people to develop the habit of eating breakfast in order to adjust their daily rhythm, foster physical and mental health, and reduce the risk of dementia. We take the position of recommending that it be installed.

Specifically, the feedback unit 15 can monitor the chewing time for one breakfast (meal after waking up) and determine whether the chewing time is 10 minutes or more. Furthermore, taking into consideration the possibility that the patient is ingesting low-calorie foods such as gum, the upper limit of the chewing time may be set to, for example, 60 minutes, and it may be determined whether the chewing time is 60 minutes or less. .

Also, whether the duration of one meal, including breakfast, as mentioned above, deviates from the standard range (is it too long or too short), in other words, whether eating has become a problem behavior? It is also preferable to judge using the deviation score r. Furthermore, it may be determined whether the number of times of chewing (chewing time) in one meal is insufficient or not using the deviation score r.

Furthermore, it is also preferable that the feedback unit 15 monitors mastication movements from the viewpoint of preventing lifestyle-related diseases and dementia. For example, on the website of the Ministry of Health, Labor and Welfare, there is a description of the impact of oral function on health, and specifically, for elderly people, deterioration of oral function leads to insufficient energy intake through meals, resulting in decreased muscle strength and immunity. It has been pointed out that there are cases where people become bedridden due to weakening of their strength and developing pneumonia, etc. Furthermore, it is important to maintain oral function in order to maintain "opportunities to interact with others". It is stated that it is essential to do so.

Furthermore, it is generally known that there is a correlation between eating too quickly and obesity, and it is said that it is a good idea to correct eating too quickly in order to prevent lifestyle-related diseases and vascular dementia. In addition, there are studies that show that chewing your meals slowly and chewing your meals slowly increases postprandial meal-induced thermogenesis, increases cerebral blood flow, and improves short-term memory. Studies have also reported that the initial secretion ability of insulin is increased by strengthening masticatory movements.

Specifically, the feedback unit 15 determines, for example, whether the number of times of chewing per mouthful during breakfast is 30 or more, and whether the total number of times of chewing during breakfast is 1000, and determines the number of times. If it is determined that the condition is met, praise information (feedback information) is determined; on the other hand, if it is determined that the condition is not met, it is determined that the condition is satisfied, that the child chews his/her food thoroughly, or that he/she chews gum, etc. after eating. It is also preferable to determine advice information to prompt the user to do so. Note that such processing is not limited to breakfast, but can also be applied to lunch and dinner. Further, the feedback information as described above may be determined based on the mastication time instead of the number of times of mastication.

Furthermore, if chewing is not determined for a predetermined period of time after the breakfast time period, the feedback unit 15 may determine advice information to the effect that breakfast is to be eaten. Further, if chewing is interrupted for a predetermined period of time or more during a meal, or if the mastication pace fluctuates by a predetermined amount or more, feedback information (warning information) related to a warning can be determined in consideration of the possibility of physical illness. Furthermore, if the chewing pace fluctuates more than a predetermined amount during a meal, considering the possibility of a decline in masticatory ability, advice information such as "chew firmly with your back teeth 30 times" is determined, Furthermore, in order to check whether such chewing is possible, the device may be set to shift to a test mode.

In addition, the feedback unit 15 provides feedback information about the result of comparing the time difference between the dinner end time and the scheduled bedtime or bedtime with a predetermined reference value as the user's (wearer's) eating/sleeping lifestyle information. It is also preferable to For example, in order to recommend the habit of not eating for 3 hours before going to bed in order to get a good night's sleep, for example, when chewing is detected 3 hours before the scheduled bedtime (estimated from lifestyle pattern information), or when chewing is detected at bedtime, If the time difference between the time and the dinner end time is 3 hours or less, advice information (feedback information) such as "avoid eating before going to bed" may be determined.

Furthermore, it is also preferable that the feedback unit 15 determines advice information to urge the user to eat dinner if chewing is not determined within 4 hours before the scheduled bedtime (estimated from the lifestyle pattern information). Further, for example, if it is 3 to 4 hours before the scheduled bedtime, advice information may be determined to recommend a light meal (with less digestive burden). Furthermore, if the average sleeping time determined from the lifestyle pattern information is less than six and a half hours, it is also possible to determine advice information that urges the user to eat dinner earlier.

Further, the feedback unit 15 may determine the advice information regarding the time to eat dinner and/or breakfast so that the fasting time from the dinner end time to the next breakfast start time is 8 hours or more. Furthermore, if mastication is determined at a time other than breakfast, lunch, and dinner, it is also possible to determine advice information to the effect that the user will not consume too many calories.

In addition, the feedback unit 15 compares the sleeping time when sleeping (napping) is detected within a predetermined time range from the lunch end time with a predetermined reference value as the user's (wearer's) eating/sleeping lifestyle information. It is also preferable to use information regarding the results as feedback information.

For example, if the bedtime (nap time) is less than 30 minutes or more than 1 hour, the feedback unit 15 may determine advice information that prompts an appropriate nap time. In some cases, when the nap time exceeds one hour, an audio alarm may be output as feedback information related to the warning.

So far, we have described the process of determining feedback information related to meals mainly from information related to mastication. In this way, chewing is a movement involved in the basic activity of eating, and in a sense, it can also be seen as an indicator of the will to live. In addition, a smile (raised corner of the mouth) is also an expression of the activity of showing a smile and trying to be involved in society.Furthermore, depending on the situation, a smile (raised corner of the mouth) is also an expression of an activity in which one is trying to be involved in society.Furthermore, depending on the situation, a smile (raised corner of the mouth) is also an expression of an activity that is performed when one is feeling frustrated or working hard, and is stressed out. It can also be seen as a type of activity aimed at confronting the situation.

Furthermore, chewing, smiling (raised corners of the mouth), and clenching are also known to have the effect of reducing stress, and are considered to be an expression of a person's desire to reduce stress and lead a better life. In other words, these can be understood as vitality indicators (motivation for life indicators) that indicate the level of motivation for life. Furthermore, such lifestyle motivation index is not only an indicator of the desire to live a better life, but also an index for evaluating the degree of decline in oral function that can cause dementia and aspiration. It is also possible to capture it.

Here, for example, the total amount of myoelectric signals captured near the left and right cheekbones (total muscle The amount of activity) can also be used as an indicator of motivation for living. Furthermore, in addition to the amount of muscle activity in the head, information such as the amount of walking (number of steps) and "whether or not you wake up within a specified time every morning" may also be incorporated as elements for determining the lifestyle motivation index. . In this case, for example, the life motivation index may be calculated per day or for a longer period.

Alternatively, using the warning score tot, use the following formula (8) (life motivation index) = 1 - tot / (maximum possible value of the warning score tot)
It is also possible to obtain a lifestyle motivation index. In this case, the life motivation index is the amount determined by all the set meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts (all monitor data), and is a value from 0 to 1. A value of 1 indicates the highest motivation for life, and a value of 0 indicates the lowest motivation for life.

In any case, it is also preferable that the feedback unit 15 outputs information related to the lifestyle motivation index determined as described above as feedback information.

<Feedback information related to mental state (smile)>
In this embodiment, when measuring the mental state, attention is paid to smiles (raised corners of the mouth). In other words, in general, when you smile a lot, it means that you are having pleasant communication with others and are in a good mental state. It is estimated that the amount will decrease. It is also known that the frequency of laughter generally decreases with age, and that there is a relationship between cognitive function and the frequency of laughter.

For example, using the discrepancy score r, the feedback unit 15 uses content that makes you feel happy (for example, comedy content or dialogue content) may be output as feedback information. It is also possible to determine feedback information that notifies cohabitants, managers, etc. that smiling time is insufficient. Basically, although the smiling time is monitored, it is also preferable to treat the problem of no smiling (raised corners of the mouth) as a problem and to determine feedback information corresponding to this.

Note that the time obtained by subtracting the chewing time from the time when it was determined that the myoelectric signal was generated may be monitored as the approximate smiling time. Moreover, the smiling time for each day can also be determined by taking a moving average over one week (7 days), for example. Furthermore, if the smiling time of the user (wearer) has been continuously monitored for several years, for example, it is also preferable to visualize the smiling time using a time-series (for example, weekly, monthly, or yearly) moving average graph.

Moreover, the feedback unit 15 may display or notify the user by voice each time a smile (raised corner of the mouth) is determined as feedback information. This also allows the user (wearer) to confirm that the device is operating normally. Furthermore, the user (wearer) may respond positively to the device by smiling (raising the corners of the mouth). This can also increase the number of times the user (wearer) smiles.

<Feedback information related to exercise>
Walking is an exercise that can be done easily and is considered to be good for health. Regarding walking, it is recommended to walk at least 5,000 steps a day, and it is also recommended to do interval walking (walking in which you alternate between slow walking and brisk walking, for example, 5 times for 3 minutes each, for a total of 30 minutes). has been done. In addition, the Ministry of Health, Labor and Welfare recommends that people under 65 years of age engage in exercise with an activity level of 23Ex per week, and has also disclosed a trial calculation for doing this by walking only. The feedback unit 15 may output feedback information to that effect when the monitored activity amount Ex is insufficient based on the above criteria.

Furthermore, in the present embodiment, the feedback unit 15 determines, for example, that the activity time with an exercise intensity of 3 METs or more (or the activity time equivalent to 5000 steps or more) is the exercise time, and uses the deviation score r to determine whether the exercise time is insufficient. If the exerciser is doing it or is doing it excessively (the deviation score r is above a predetermined value), advice information such as urging an appropriate amount of exercise time may be output as feedback information. Further, it is also preferable to output advice information that recommends exercising in the morning, 30 minutes after breakfast. Furthermore, when the feedback unit 15 detects a heart rate close to the maximum heart rate of the user (wearer), it is also possible to output advice/warning information to stop exercise and urge rest as feedback information. .

For example, the feedback unit 15 may notify the remaining number of steps to be taken or the walking time (assuming 10 minutes is spent for 1000 steps) in a certain day. In addition, if the target number of steps has not been reached even at a predetermined time in the evening or after dinner, the free time is estimated from the lifestyle pattern information, and when the free time becomes available, the message "Would you like to take a short walk?" is sent. It is also preferable to output advice information. Note that the walking time (to be performed) for each day can also be determined by taking a moving average over one week (7 days), for example. Furthermore, when the number of steps or walking time reaches a goal, it is also possible to introduce a system in which praise information to that effect is posted to, for example, the SNS server 4 and can be praised by other people.

Furthermore, based on the exercise-related amount (monitor data) related to a predetermined lifestyle behavior by the user (wearer), the feedback unit 15 outputs information for a predetermined period from the meal end time as lifestyle information related to the lifestyle behavior after the meal. It is also preferable to determine, as feedback information, information regarding whether or not the lifestyle movement has been performed within the range.

For example, if you set tooth brushing (tooth brushing movement) as the above-mentioned lifestyle action and can detect the brushing movement from the acceleration change pattern by the acceleration sensor of the wristband 3, then 30 minutes have passed since the end of the meal. However, if the brushing movement is not determined, an audio alarm is output as feedback information related to the warning, and further, if the brushing movement is still not determined for a predetermined period of time (for example, 5 minutes), advice information to urge teeth brushing is output. You can. In addition to this information, it is also possible to present educational information to the effect that oral care is effective in preventing dementia.

Although tooth brushing (tooth brushing exercise) is set here as the relevant lifestyle action after a meal, other lifestyle actions such as defecation behavior or muscle training such as push-ups can also be set as the relevant lifestyle action. . Furthermore, the brushing motion of the teeth can also be detected by an acceleration sensor of the head mount 7, which will be explained in detail later with reference to FIG. That is, the head attachment 7 (FIG. 10) can also output the above-mentioned advice information to encourage tooth brushing.

Additionally, the feedback unit 15 provides information on whether the user (wearer) is walking while chewing a predetermined luxury item including gum, as the user's (wearer's) diet/exercise-related lifestyle information. It is also preferable to do so. For example, while the user (wearer) is walking, advice information such as ``Would you like to chew gum?'' may be output, or when the user (wearer) is determined to be chewing while walking, praise may be given. It is also preferable to output information. Furthermore, the feedback unit 15 provides information related to whether or not the user (wearer) is smiling while walking, as the user's (wearer) exercise/mental state related lifestyle information. is also preferable. For example, when walking and smiling are determined at the same time, feedback information may be determined as praise information.

<Feedback information related to sleep>
In this embodiment, feedback information is determined with the goal of getting good quality sleep. In general, in order to get good quality sleep, for example, in the morning, you should expose yourself to sunlight to reset your body clock, take a 30-minute nap after lunch, and do moderate aerobic exercise during the day. It is recommended that you eat dinner at least 3 hours before going to bed and get 6 to 7 hours of sleep to prevent amyloid-β accumulation. Therefore, for example, it is also preferable to set the lower and upper limit reference values of sleep time to 5.5 hours and 8.5 hours, respectively.

As such sleep-related feedback information, the feedback unit 15 determines whether or not sleep-related problem behavior is occurring, for example.
(a) Whether the amount of sleep is insufficient or too long;
(b) Whether or not bedtime is not determined even after a predetermined period of time has elapsed from the scheduled bedtime (staying up late);
(c) Whether or not mid-way awakening occurs more than a predetermined amount;
(d) Determine whether or not there is a dispersion in wake-up times that is greater than a predetermined value using, for example, the deviation score r, and if it is determined that such problematic behavior is occurring, advice information to correct this behavior. may be determined.

Here, mid-awakening in (c) above may correspond to REM sleep behavior disorder, which is a characteristic symptom of Lewy body dementia, and it is important to determine feedback information as a countermeasure for this dementia. becomes. For example, it is also preferable to set the lower limit reference value of this mid-way awakening to zero, and to detect even the slightest mid-way awakening as a problem behavior. Incidentally, the pulse rate is taken as an index of the depth of sleep, and when the pulse rate rises above a predetermined value, it can be determined that midway awakening has occurred.

Moreover, since the above-mentioned (d) variation in wake-up time may also be a symptom of dementia, it is important to determine feedback information as a countermeasure against dementia. Furthermore, this variation in wake-up time (the smallness of it) can also be taken as an indicator of motivation for living, and in that respect, it is important monitoring data. For example, the lower limit reference value of this wake-up time variation can be set to 0 to 10 minutes, and the upper limit reference value can be set to 60 minutes. It is also preferable that the wake-up time variation is determined for each day of the week, weekdays, and holidays.

Here, regarding the (upper and lower limits) reference values of sleep time, for example, extract the sleep time on days when the sleep was 1 (good sleep) in the input (report) on the touch panel display 104 by the user, and calculate the center of these values. The upper and lower limit reference values may be determined by calculating representative values such as the maximum value, the mode value, and the average value. In addition, we modeled not only the sleep time on days when we had a good night's sleep, but also the relationship between the sleep hours on days when we had a good night's sleep and all the monitor data from the previous day, using multivariate analysis such as multiple regression analysis and machine learning. It is also preferable to predict a suitable sleep time for a good night's sleep that corresponds to the day's activities, and to set upper and lower reference limits based on this. Furthermore, it is also possible to determine the scheduled bedtime time from the predicted suitable sleep time, further determine the dinner start time, and output advice information that encourages daily activities in accordance with these times.

The feedback unit 15 also indicates that within two hours after dinner, the pulse rate of the user (wearer) has continued for a predetermined period of time (estimated from past measurement results) and has approached the pulse rate during normal sleep. In this case, an audio alarm (or a vibration alarm, an electrical stimulation alarm, or an alarm notification to a conversational robot or a cohabitant) may be output as feedback information related to the warning so that the user does not fall asleep. This makes it possible to avoid falling asleep immediately after dinner, reducing the burden on the digestive organs, and making it easier to achieve a good night's sleep.

Further, the feedback unit 15 may output advice information urging the user (wearer) to go to bed eight hours before the scheduled wake-up time estimated from the lifestyle pattern information. At that time, it is also possible to output command information such as setting the touch panel display 104 to a blue light reduction mode, dimming the lighting in the room where the user (wearer) stays, or turning off audio equipment, etc. .

Further, when the feedback unit 15 determines that the pulse rate tends to increase during the preset scheduled wake-up time (zone), the feedback unit 15 outputs a wake-up alarm, such as an audio alarm, a vibration alarm, or a light alarm. Furthermore, a command to start up lighting equipment, air conditioning equipment, and audio equipment in a room, and a command to let in outside light to an IoT (Internet of Things) window may be output. This allows the user (wearer) to wake up when his or her pulse rate increases, that is, when his or her sleep becomes light, allowing him or her to have a comfortable morning.

In addition, the user (wearer) is configured to be able to input information to the device that he or she was able to wake up comfortably by shaking the arm on which wristband 3 (its acceleration sensor) is attached, for example, three times. You can. Further, it may be possible to notify that it was a bad awakening by shaking the device strongly, for example, 10 times. Furthermore, it is also preferable that the feedback unit 15 sets/updates the scheduled wake-up time to be applied to later notifications from a log recorded in association with the wake-up time and the corresponding comfort level of wake-up.

Furthermore, if the pulse rate is higher than a predetermined value during the time before and after waking up, the feedback unit 15 determines that the autonomic nervous system is activated and the circulatory system is under load, and sends feedback information related to a warning. The system then outputs advice that recommends entering a room at an appropriate temperature, stopping physical activity (lying down if possible), and breathing in and out for 4 seconds and exhale for 6 seconds to regulate breathing. Good too. Additionally, if you are exercising at a predetermined intensity of METs or more during the morning hours, it is possible to output advice/warning information urging you to take a break. Furthermore, during the day, if the pulse rate (which can also be considered as alertness level) is lower than a predetermined level, advice information such as recommending chewing gum or walking is determined, and furthermore, if the pulse rate is within the normal range. It is also preferable to output praise information that praises the user when the user recovers.

<Feedback information related to problem behavior>
It is also preferable that the feedback unit 15 utilizes a problem behavior response table or a praise opportunity response table as a preferred feedback implementation mode.

6, 7, and 8 are problem behavior correspondence tables held by one embodiment of the feedback unit 15. Moreover, FIG. 9 is a praise behavior correspondence table held by one embodiment of the feedback unit 15.

In the tables in Figures 6 to 8, various problem behaviors are listed, and for each of these problem behaviors, there are "notification conditions" for outputting warning information (feedback information), and "notification conditions" for canceling the output of warning information. "Recovery conditions", "escalation" which is a staged presentation mode when no correction is made even after warning information is presented to the user (wearer), and "notification time" are determined. In addition, Figure 9 lists various opportunities for praise that can be praised, and for each of these praise opportunities, the content of praise (reward) when praise is given on a daily basis, and the content of praise (reward) when praise is given on a weekly basis. The content of the praise (remuneration) is determined.

By using such a problem behavior correspondence table and a praise opportunity correspondence table, the feedback unit 15 can perform suitable feedback processing according to the user's (wearer's) action content and situation. By the way, the various types of feedback information (advice information, warning information, etc.) that have been explained so far can be suitably implemented by using such problem behavior response tables and praise opportunity response tables. It will become.

Here, by using the "medals" as rewards in the praise opportunity corresponding table in Fig. 9, users can compete for medal acquisition rankings among their friends or organizations they belong to, or even play an item collection type or trading card type game. It is also preferable to have the Here, users (wearers) can pay the collected "medals" to, for example, view their favorite content, draw fortunes, obtain stamps for use on SNS, and costumes and accessories for their avatars. It may be set to do so.

<Other feedback information>
The feedback unit 15 may further provide training information as feedback information. For example, users (wearers) can receive functional maintenance training content that can be carried out in a game-like manner to prevent the decline in oral function, which can cause aspiration pneumonia, and the decline in cognitive function, which is a symptom of dementia. It is also preferable to provide the

More specifically, a game was developed in which the myoelectric signals of chewing and smiling (raised corners of the mouth) were monitored, the participants were given chewing training and smiling training, and when they were able to perform good chewing and smiling, they were praised. mification content can be provided. Here, while looking at the rhythm game screen, you can smile at the right time (raise the corner of the mouth), maintain the biggest smile (raise the corner of the mouth) for a few seconds, and even win a lottery or gacha every 1000 times when chewing gum. It is also possible to train the muscles that work during these movements by promoting the movements of smiling and chewing. Further, game content may be provided in which a plurality of users (wearers) share the number of chewing times with each other and perform rankings for a predetermined period on a game screen.

Furthermore, the feedback unit 15 may output advice information that urges the user to keep a diary, which is also useful for maintaining cognitive function. For example, if it is detected from the user's input (report) or lifestyle pattern information that no diary has been recorded by the scheduled bedtime, it is preferable to prompt the user (wearer) to keep a diary for today.

Further, the feedback unit 15 grasps and stores the time when the user does not wear or use the head attachment 2 or the wristband 3 as the unwearing time, and stores, for example, the statistical values (average value, standard deviation, , median value, mode value, etc.), it is also preferable to exclude the non-wearing time from the period covered by statistical processing to calculate appropriate statistical values.

Furthermore, the feedback unit 15 collects and digitizes monitor data (for example, sleep time, number of steps (exercise time), meal time (number of times chewing), smiling time, resting pulse rate, For example, a significant difference test is performed between the most recent January's monitor data and the previous January's monitor data. It is also preferable to output praise information praising this fact when a significant difference is recognized that the monitor data of a person) is approaching a more preferable data value.

[Other embodiments of life monitoring device]
FIG. 10 is a schematic diagram showing another embodiment of the lifestyle monitoring device according to the present invention.

According to FIG. 10, the lifestyle monitoring device of this embodiment is a head-mounted device 7, and this head-mounted device 7 specifically includes a frame 702, a left detection electrode section 702a, a right detection electrode section 702b, and a left detection electrode section 702b. The signal processing section 71 includes an electrode section 703a, a right reference electrode section 703b, a left bone conduction speaker 704a, a right bone conduction speaker 704b, a GND band 705, and a signal processing section 71, and further includes an activity meter 711, myoelectric information Storage unit 712, activity level information storage unit 713, microphone 715, speaker 716, communication interface 717, signal processing/control unit 721, signal conversion unit 722, pre-filter processing unit 723, audio output unit 724, audio amplification unit 725, meal A myoelectric sensor related amount determining section 73 including a related amount determining section 731 and a mental state related amount determining section 732, an activity meter related amount determining section 74 including a sleep related amount determining section 741 and an exercise related amount determining section 742, and a warning score. It has a determining section 75, a lifestyle pattern information generating section 76, a feedback section 77, an input/output control section 78, and a communication control section 79.

The components and functional components of the head mount 7 listed above are the components and functions of the same name in the daily life monitoring system (including the mobile terminal 1, the head mount 2, and the wristband 3) shown in FIG. It has the same structure, configuration, and contents as the constituent parts, and performs the same functions and actions. That is, the head attachment 7 is a single device that performs the same operations and processes as the lifestyle monitoring system shown in FIG. 1, and serves as a lifestyle monitoring device that achieves similar effects.

However, unlike the pulse sensor of the activity meter 31 (FIG. 1), the pulse sensor of the activity meter 711 in the head mounted device 7 is different from the pulse sensor of the activity meter 31 (FIG. 1). The capacitance change) is measured by a transmissive or reflective optical sensor, and this is used to detect the pulse. Therefore, in the head-mounted device 7, the optical probe 706 of the pulse sensor of the activity meter 711 is placed in contact with the earlobe position, or is provided in the form of a clip that pinches the earlobe. be.

Furthermore, unlike the mobile terminal 1, the head mounted device 7 is not provided with a touch panel display, but when displaying warning scores r, lifestyle pattern information, and even feedback information, these information can be displayed through the communication interface. 717 to the mobile terminal 8 owned by the user (wearer), and displayed on the touch panel display of the mobile terminal 8. Further, when the user (wearer) is wearing a wristband, it is also possible to display this information on the display of the wristband. Furthermore, this information (warning score, lifestyle pattern information, and even feedback information) may be output from the speaker 716 as preset sound (for example, a predetermined melody sound, beep tone sound, etc.) or synthesized sound. preferable.

Incidentally, both the head mount 7 described above and the head mount 2 shown in FIG. 1 are wearable devices that clip the head from behind, but are of course not limited to this. For example, it may be a headphone type, hair band type, or headgear type wearable device.

Further, as another embodiment, it is preferable that the head mount 7, the head mount 2, and the mobile terminal 1 (FIG. 1) described above can activate a hearing test mode. Hearing loss can impede communication and increase the risk of developing dementia, so hearing tests are important.

Specifically, in the hearing test mode in this embodiment, for example, each test sound of 125Hz, 250Hz, 500Hz, 1000Hz, 2000Hz, 4000Hz, and 8000Hz is sequentially transmitted through the left bone conduction speaker (704a, 24a) or the right bone conduction speaker (704b, 24b), and the user (wearer) inputs the result of hearing or not into the mobile terminal 8 (touch panel display) or the touch panel display 104 (FIG. 1) for testing. .

For example, if such a hearing test is performed when the device is used for the first time and every other week thereafter, and it is detected that the hearing ability has decreased below a predetermined standard, the feedback unit (77, 15) may, for example, Advice information that prompts diagnosis at a medical institution may be determined and output. In addition, the output (vibration amplitude) level of the left bone conduction speaker (704a, 24a) and the right bone conduction speaker (704b, 24b) is increased (adjusted) according to (the degree of decline in) the hearing ability of the user (wearer). It is also preferable.

Additionally, the hearing acuity of the left and right ears is compared for each frequency, and if the hearing acuity in the left and right ears is more than a certain level, the user (wearer) is notified that there is a suspicion of uneven mastication. You can. Here, the next time the meal is determined, the feedback unit (77, 15) outputs advice information such as "Let's chew 15 times with the right side. Next time, let's chew 15 times with the left side." is also preferable. Incidentally, the occurrence of uneven mastication, which is a problem here, can also be detected as a meal-related amount (monitor data) related to mastication when a 2-channel myoelectric sensor system is adopted.

Furthermore, in the head-mounted device 7 and the daily life monitoring system shown in FIG. As a rejection in case of a case or as an affirmation in a case of continuation/resumption, e.g.
(a) (When using head mount 7 or head mount 2) Smiling continuously for 3 seconds or more is an affirmative response, while chewing with the molars 3 or more times is a rejection. It is also preferable to make it a response, or,
(b) (When using the wristband 3) It is also preferable that waving the arm three times serves as an affirmative response, and waving the arm 10 or more times as a rejection response.

As explained in detail above, according to the present invention, amounts related to a plurality of life-related items including meals of the user (wearer), that is, "meal-related amounts," "sleep-related amounts," and/or "exercise-related amounts." It is possible to generate and provide more comprehensive information related to the user's (wearer's) life, ie, "meal/sleep related life information" and/or "meal/exercise related life information" based on the "quantity".

Furthermore, for example, as one embodiment, the lifestyle monitoring device or system according to the present invention determines a user's (wearer's) "meal-related amount," "sleep-related amount," "exercise-related amount," and "mental state-related amount." The device generates and presents "lifestyle information", such as lifestyle advice information, to prevent and reduce lifestyle-related diseases and dementia, for example, so that the user (wearer) can lead a life that is healthy both physically and mentally. It is also possible to support

Regarding the various embodiments of the present invention described above, various changes, modifications, and omissions within the scope of the technical idea and viewpoint of the present invention can be easily made by those skilled in the art. The above description is merely an example and is not intended to be limiting in any way. The invention is limited only by the claims and their equivalents.

1 Mobile terminal (life monitoring device)
101, 214, 32, 717 Communication interface 102, 712 Myoelectric information storage unit 103, 713 Activity amount information storage unit 104 Touch panel display (TP/DP)
105, 715 Microphone 106, 716 Speaker 11, 73 Myoelectric sensor related amount determination section 111, 731 Meal related amount determination section 112, 732 Mental state related amount determination section 12, 74 Activity meter related amount determination section 121, 741 Sleep related amount determination section Amount determination unit 122, 742 Exercise-related amount determination unit 13, 75 Warning score determination unit 14, 76 Life pattern information generation unit 15, 77 Feedback unit 16, 78 Input/output control unit 17, 79 Communication control unit 2 Head attachment 21 , 71 signal processing section 211, 722 signal conversion section 212, 723 pre-filter processing section 213 signal control section 215, 724 audio output section 216, 725 audio amplification section 22, 702 frame 22a, 702a left detection electrode section 22b, 702b right detection Electrode section 23a, 703a Left reference electrode section 23b, 703b Right reference electrode section 24a, 704a Left bone conduction speaker 24b, 704b Right bone conduction speaker 25, 705 GND band 3 Wrist band 31, 711 Activity meter 311, 721 Signal processing/ Control unit 4 SNS server 5 External server 7 Head mounted device (life monitoring device)
706 Probe 8 Mobile terminal

Claims (13)

A lifestyle monitoring device that monitors a user's lifestyle,
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from an output of an activity meter worn by the user;
Based on the determined meal-related amount and the determined exercise-related amount, information regarding the user's diet and exercise-related lifestyle includes whether or not the user is walking while chewing a predetermined luxury item. 1. A lifestyle monitoring device comprising: feedback means for generating or selecting and providing information related to the above. A lifestyle monitoring device that monitors a user's lifestyle,
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
a mental state related amount determining means for determining a mental state related amount including an amount related to a smile of the user from the output of the myoelectric sensor;
Determine sleep-related amounts, including amounts related to sleeping and/or waking up of the user, and exercise-related amounts, including amounts related to exercise, including walking, of the user, from the output of the activity meter worn by the user. Activity meter related quantity determination means;
Based on the determined meal-related amounts, mental state-related amounts, sleep-related amounts, and exercise-related amounts, information related to the user's diet and sleep-related lifestyle, and information on the user's lifestyle related to exercise and mental status are provided. and feedback means for generating or selecting and providing such information,
The information related to the user's life related to meals and sleep is information related to judgment or advice regarding the relationship between the time related to meals and the time related to sleep, and is information related to the user's life related to exercise and mental state. A life monitoring device characterized in that the information is information regarding whether or not the user is determined to be smiling while walking. further comprising a mental state related amount determining means for determining a mental state related amount including an amount related to a smile of the user from the output of the myoelectric sensor,
The activity meter related amount determining means determines an exercise related amount including an amount related to walking of the user,
The feedback means also provides information regarding the user's lifestyle related to exercise and mental state based on the determined exercise-related amount and mental state-related amount,
The lifestyle according to claim 1 , wherein the information related to the user's lifestyle related to exercise and mental state is information related to whether or not the user is determined to be smiling while walking. Monitor device. The activity meter related amount determining means determines both a sleep related amount including an amount related to going to bed and/or waking up of the user, and the exercise related amount,
The lifestyle monitoring device includes:
a mental state related amount determining means for determining a mental state related amount including an amount related to a smile of the user from the output of the myoelectric sensor;
For each of the determined meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts, a deviation score is calculated based on the degree of deviation from the set reference value, and a warning score is obtained by integrating the deviation scores. and a warning score determining means for determining the
4. The lifestyle monitoring device according to claim 1, wherein the feedback means also provides information regarding the determined warning score. The warning score determining means calculates the deviation score using a modified sigmoid function whose variable is quadratic and includes a lower limit reference value and an upper limit reference value as the reference value. 4. The lifestyle monitoring device according to item 4 . The activity meter related amount determining means determines both a sleep related amount including an amount related to going to bed and/or waking up of the user, and the exercise related amount,
The lifestyle monitoring device includes:
a mental state related amount determining means for determining a mental state related amount including an amount related to a smile of the user from the output of the myoelectric sensor;
Life pattern information that generates life pattern information that is information related to the lifestyle of the user from a plurality of meal-related amounts, sleep-related amounts, exercise-related amounts, and mental state-related amounts determined for a predetermined period. further comprising a generating means,
The lifestyle monitoring device according to any one of claims 1 to 5 , wherein the feedback means also provides information related to the generated lifestyle pattern information. The activity meter related amount determining means determines a sleep related amount including an amount related to going to bed and/or waking up of the user,
The feedback means generates or selects and provides information related to the user's diet and sleep-related lifestyle based on the determined meal-related amount and the determined sleep-related amount,
Information related to the user's life related to meals and sleep shall be information related to the result of comparing the time difference between the time when dinner ends and the time when the user plans to go to bed or the time when he or she went to bed with a predetermined reference value. The lifestyle monitoring device according to any one of claims 1 to 6 , characterized in that: The activity meter related amount determining means determines a sleep related amount including an amount related to going to bed and/or waking up of the user,
The feedback means generates or selects and provides information related to the user's diet and sleep-related lifestyle based on the determined meal-related amount and the determined sleep-related amount,
Information related to the user's daily life related to meals and sleep is based on the result of comparing the time when the user went to bed within a predetermined time range from the time the lunch ended with a predetermined reference value. The lifestyle monitoring device according to any one of claims 1 to 7 , characterized in that the information is information. The lifestyle monitoring device is a device that can be worn on the head of the user, and includes the myoelectric sensor and the activity meter,
The myoelectric sensor includes at least one left electrode that contacts at least one skin position between the user's left ear pinna and cheek area, and at least one left electrode that contacts at least one skin position between the right ear pinna area and cheek area of the user. the left electrode and the right electrode, the myoelectric signals generated due to at least one of mastication, clenching or clenching, and raising the corners of the mouth; detectable;
The life monitoring device according to any one of claims 1 to 8 , wherein the activity meter includes an acceleration sensor and a pulse sensor. The lifestyle monitoring device according to any one of claims 1 to 9 , wherein the information related to the lifestyle is advice information and/or praise information for the user. A life monitoring system that monitors a user's life,
a myoelectric sensor located on the head of the user;
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of the myoelectric sensor;
an activity meter worn by the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from the output of the activity meter;
Based on the determined meal-related amount and the determined exercise-related amount, information regarding the user's diet and exercise-related lifestyle includes whether or not the user is walking while chewing a predetermined luxury item. 1. A lifestyle monitoring system comprising: feedback means for generating or selecting and providing information related to A program that operates a computer installed in a life monitoring device that monitors the life of a user,
Meal-related amount determining means for determining a meal-related amount including an amount related to mastication of the user from the output of a myoelectric sensor located on the head of the user;
an activity meter-related amount determining means for determining an exercise-related amount including an amount related to walking of the user from an output of an activity meter worn by the user;
Based on the determined meal-related amount and the determined exercise-related amount, information regarding the user's diet and exercise-related lifestyle includes whether or not the user is walking while chewing a predetermined luxury item. A life monitoring program characterized by causing a computer to function as a feedback means for generating or selecting and providing information related to the following. A lifestyle monitoring method implemented by a computer installed in a lifestyle monitoring device that monitors a user's lifestyle, the method comprising:
From the output of the myoelectric sensor located on the user's head, the amount related to the user's meal, including the amount related to chewing, is determined, and from the output of the activity meter worn by the user, the amount of food related to the user's gait is determined. determining a motion-related quantity including a quantity relating to;
Based on the determined meal-related amount and the determined exercise-related amount, information regarding the user's diet and exercise-related lifestyle includes whether or not the user is walking while chewing a predetermined luxury item. 1. A lifestyle monitoring method comprising the step of generating or selecting and providing information related to the following.

Images