KR102494874B1 - Method and apparatus for predicting dementia based on Activity of daily living - Google Patents

Abstract

A method for predicting dementia based on daily living ability information is provided. Dementia prediction method based on daily living ability information according to an embodiment of the present invention includes the steps of obtaining daily living ability information related to dementia prediction by monitoring indoor behavior of a user through an IoT device installed indoors, using the daily living ability information determining a lifestyle pattern for the user's behavior, and determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the lifestyle through the daily life ability information. can do.

Description

Method and apparatus for predicting dementia based on daily living ability information {Method and apparatus for predicting dementia based on Activity of daily living}

The present invention relates to a method and apparatus for predicting dementia based on daily living ability information. More specifically, it relates to a dementia prediction method and device based on daily life ability information that can predict and diagnose the occurrence of dementia early by monitoring daily life in a user's actual living space based on smart home technology.

Dementia is emerging as a social problem in the current population decline and aging society. In particular, since dementia cannot be cured at the current level of medical care, the only solution is to predict dementia early and slow it down. However, in the case of the elderly living alone or those who are vulnerable to medical care, there is no way to predict dementia early because there is a lack of social conditions to closely observe the risk of dementia. In particular, since aging elderly people often spend most of their time at home, there is a limit to monitoring the elderly through outside personnel.

Activity of Daily Living (ADL) refers to basic activities of daily living necessary for self-care, and it is treated as an important factor in predicting dementia. If daily living ability information is monitored in real time and usefully used, it may be possible to predict dementia in the elderly at an early stage.

Therefore, there is a demand for an efficient method for predicting dementia using daily life ability information for the elderly who spend most of their time at home or in a similar life pattern.

Registered Patent Publication No. 10-2102931 "Dementia Prediction System" (registered on April 14, 2020)

The technical problem to be solved by the present invention is to provide a dementia prediction method based on daily living ability information that can early predict and diagnose the occurrence of dementia by monitoring the daily life of the user's actual living space based on smart home technology. is to do

Another technical problem to be solved by the present invention is to provide a dementia prediction method based on daily life ability information that can monitor the user's behavior in detail as the user's daily life is monitored according to the daily life collection items of the user's actual living space. is to do

Another technical problem to be solved by the present invention is to provide a dementia prediction method based on daily living ability information that can accurately determine and predict the user's risk of dementia by more closely observing the user's wandering along the moving path.

The technical problems of the present invention are not limited to the technical problems mentioned above, and other technical problems not mentioned will be clearly understood by those skilled in the art from the description below.

In order to solve the above technical problem, a method for predicting dementia based on daily living ability information according to an embodiment of the present invention includes the steps of obtaining daily living ability information related to dementia prediction by monitoring a user's indoor behavior through an IoT device installed indoors. Determining a lifestyle pattern for the user's behavior using the daily living ability information, and detecting that the user's behavior deviates from the living pattern through the daily living capability information obtained after determining the daily living pattern. In this case, determining that the user is at risk of dementia may be included.

In one embodiment, the method may include detecting an operation in daily life of the user according to a pre-classified daily life collection item using data obtained from an IoT device installed indoors.

In an embodiment, the method may further include analyzing the user's indoor loitering according to the user's movement information using a wearable device attached to the user.

In one embodiment, the analyzing of the user's loitering according to the user's movement information may include determining whether the user is repeatedly moving without a purpose for the visit place based on information about the user's heart rate and walking speed. The method may further include detecting whether or not the user is loitering.

In one embodiment, the step of determining the living pattern may include receiving data about the characteristics of the spatial structure in which the user lives, and determining the living pattern of the user according to the characteristics of the spatial structure. there is.

In an embodiment, the determining of the life pattern may further include receiving data on the user's physical characteristics and determining the user's life pattern according to the user's physical characteristics.

In an embodiment, the step of determining the life pattern may further include receiving data about the degree of dementia symptoms of the user and determining the lifestyle of the user according to the degree of dementia symptoms.

In an embodiment, the detecting of the user's motion in daily life may include collecting data according to a first rule defining the arrangement of sensors for detecting the motion of the user of the normal group.

In one embodiment, the step of collecting data according to the first rule is to collect data through a vibration sensor, a door sensor, a motion sensor, a temperature and humidity sensor, a smart plug, and a lidar sensor in the room where the user of the normal group is located. steps may be included.

In an embodiment, the detecting of the user's motion in daily life may include collecting data according to a second rule defining the arrangement of sensors for detecting the motion of the user in the dementia group.

In one embodiment, the step of collecting data according to the second rule is to collect data through vibration sensors, door sensors, motion sensors, temperature and humidity sensors, smart plugs, and lidar sensors in the room where the user of the dementia group is located. steps may be included.

In one embodiment, the step of determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the lifestyle pattern, the value of the daily life ability information obtained after the lifestyle pattern is determined in a time-sequential manner. Expressing as a chart expressed in time series, averaging the chart expressed in time series with an average value for each predetermined interval and converting it into a staircase chart, assigning a symbol character corresponding to the average value in the staircase chart to It may include symbolizing the staircase chart, and determining whether the behavior of the user deviates from the life pattern by using the symbolized symbol characters.

In one embodiment, the step of determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the lifestyle pattern, the activity time of the daily living ability information obtained after the lifestyle pattern is determined, and the lifestyle calculating a similarity of activity times when there is an overlapping activity time among activity times of a pattern, and classifying whether the user is normal or at risk of dementia by using the similarity of activity times; The similarity of activity time is calculated as {Infimum(E0, En) - supremum(Sn, S0)}/Dn * 100, and E0 is the observation point and end time of daily life ability information obtained after determining the life pattern. where En is the average end time of the daily living ability information obtained after the life pattern is determined, Sn is the average start time of the daily living ability information obtained after the life pattern is determined, and S0 is, This is the start time of the observation point of daily living ability information obtained after determining the life pattern.

In one embodiment, the step of determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the lifestyle pattern, and the user's sense of balance and Determining walking ability, indoor moving distance, and classifying whether the user is normal or at risk of dementia using the sense of balance, walking ability, and indoor moving distance.

In one embodiment, the step of classifying whether the user is normal or at risk of dementia using the sense of balance and walking ability is Single-leg stance (SLS), Multidirectional reach test, Timed up and go (TUG) , measuring the sense of balance and walking ability using any one of walking velocity, cadence, and gait stability ratio.

To solve the technical problem, an apparatus for predicting dementia based on daily living ability information according to another embodiment of the present invention includes a processor, a network interface, a memory that is executed by the processor and loads a computer program, and the computer program Including a storage for storing, wherein the computer program monitors the user's indoor behavior through an IoT device installed indoors to obtain daily life ability information related to dementia prediction, and the daily life ability information to obtain the user's daily life ability information. It may include an instruction for determining a lifestyle pattern for behavior, and an instruction for determining that the user is in a dementia risk state when the user's behavior is detected as deviating from the lifestyle pattern through the daily life ability information.

1 is a schematic diagram of a dementia prediction system based on daily life ability information according to an embodiment of the present invention.
2 is a flowchart of a method for predicting dementia based on daily living ability information according to an embodiment of the present invention.
3A and 3B are exemplary diagrams illustrating a smart home system installed in a user's residence to monitor a user's behavior.
4 is a diagram for explaining pre-classified daily life collection items collected from user's behavior.
FIG. 5 is a diagram for explaining step S200 of FIG. 2 in detail.
6 is a diagram illustrating an example of data according to a user's personal characteristics.
FIG. 7 is a diagram for explaining step S300 of FIG. 2 in detail.
8 is a diagram for explaining a method of analyzing a user's loitering according to a user's moving path.
9 and 10 are diagrams for explaining specific examples of calculating the similarity of activity times.
11 is a diagram for explaining a user's visiting frequency and staying time measured using a lidar sensor.
12 is a schematic diagram of a dementia prediction system based on daily life ability information according to another embodiment of the present invention.
13 is a hardware configuration diagram of an apparatus for predicting dementia based on daily life ability information according to another embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. Advantages and features of the present invention, and methods for achieving them, will become clear with reference to the embodiments described below in detail in conjunction with the accompanying drawings. However, the present invention is not limited to the embodiments disclosed below and may be implemented in various different forms, only the present embodiments make the disclosure of the present invention complete, and the common knowledge in the art to which the present invention belongs It is provided to fully inform the holder of the scope of the invention, and the present invention is only defined by the scope of the claims. Like reference numbers designate like elements throughout the specification.

Unless otherwise defined, all terms (including technical and scientific terms) used in this specification may be used in a meaning commonly understood by those of ordinary skill in the art to which the present invention belongs. In addition, terms defined in commonly used dictionaries are not interpreted ideally or excessively unless explicitly specifically defined. Terminology used herein is for describing the embodiments and is not intended to limit the present invention. In this specification, singular forms also include plural forms unless specifically stated otherwise in a phrase.

Hereinafter, some embodiments of the present invention will be described with reference to the drawings.

1 is a schematic diagram of a dementia prediction system based on daily life ability information according to an embodiment of the present invention.

Referring to FIG. 1 , a dementia prediction system based on daily living ability information according to an embodiment of the present invention may include a dementia prediction server 100, a wearable device 200, and an IoT device 300.

The dementia prediction server 100 may monitor the daily life of the user's actual living space using data received from the wearable device 200 and the IoT device 300 of the user. The dementia prediction server 100 may monitor a user's daily life and determine a life pattern according to the user's behavior. The dementia prediction server 100 determines that the user is at risk of dementia when the user's behavior deviates from the life pattern or when the user's behavior behaves similar to that of a patient with dementia through daily living ability information obtained after determining the lifestyle pattern. can judge The dementia prediction server 100 may store and manage information received from the wearable device 200 and the IoT device 300 in the DB 400 .

The wearable device 200 is a device that can be attached to a user and can detect and transmit biometric information and location information of the user to the dementia prediction server 100 . For example, the wearable device 200 may detect biosignals of the user's heart rate and walking speed. Also, the wearable device 200 may detect the user's current location and generate location information. The wearable device 200 may interwork with the user's terminal 210 to transmit/receive information with the user's terminal 210 . That is, the wearable device 200 may transmit/receive information directly to/from the dementia prediction server 100 or transmit/receive information to/from the dementia prediction server 100 via a user terminal.

The IoT device 300 may be composed of a device that generates information about the operation of an electronic product and a detection signal of a sensor. The IoT device 300 may detect a user's behavior in real time. The IoT device 300 may be installed in a house where a user resides and built into a smart home system. In addition, the IoT device 300 may be installed in various places such as public institutions, nursing facilities, and medical facilities. In one embodiment, the IoT device 300 may identify the user through the wearable device 200 of the user. For example, the IoT device 300 may identify which user is in the room through the wearable device 200 of the user and detect the identified user's behavior even if a user other than the user enters the building.

An electronic product is a device operated by a user's manipulation, and a sensor may be composed of various sensors capable of detecting a user's motion. The IoT device 300 is set to a group of similar types of devices or devices in an adjacent location, and each device group is connected to a gateway 310 that transmits and receives information to and from the dementia prediction server 100, and transmits the detected information to predict dementia. It can be transmitted to the server 100.

Dementia prediction system based on daily living ability information according to an embodiment of the present invention can predict and diagnose the occurrence of dementia at an early stage by monitoring a user's daily life based on smart home technology.

Hereinafter, a dementia prediction method based on daily life ability information according to another embodiment of the present invention will be described with reference to FIGS. 2 to 9 . This embodiment may be performed by a computing device. For example, the computing device may be the dementia prediction server 100 described with reference to FIG. 1 . In describing the present embodiment, description of a performer of some operations may be omitted. At this time, the performing entity is the computing device.

2 is a flowchart of a method for predicting dementia based on daily living ability information according to an embodiment of the present invention.

In step S100 of FIG. 2 , daily life ability information related to prediction of dementia may be acquired by monitoring the user's behavior. The user's behavior can be detected by IoT devices and wearable devices installed indoors. The dementia prediction server may receive information about the operation of the IoT device and the bio-signal and location information of the user measured by the wearable device.

The dementia prediction server may integrate and manage the information about the user's behavior into information called daily life ability information. That is, the information received from the IoT device and the wearable device may consist of various types of information such as sensor detection signals, operation signals of electronic products, or location information. These data values themselves are used as information to predict dementia hard to be Accordingly, the dementia prediction server may pre-process or classify the information so as to be easily used as information for determining the user's behavior, and generate daily living ability information, which is information about the user's behavior.

The dementia prediction server may monitor the user's behavior through the generated daily living ability information. For example, the dementia prediction server may generate daily life ability information about a user's movement location by using information about the operation of an electronic product or information about where an object has been moved. The dementia prediction server may generate daily life ability information about actions of the user, such as sitting, walking, lying down, washing, etc., by using the signal detected by the sensor. In addition, the dementia prediction server may generate daily living ability information about the user's health status using the user's bio-signal.

Hereinafter, a detailed method of acquiring daily life ability information by monitoring a user's indoor behavior will be described with reference to FIGS. 3 and 4 .

3A and 3B are exemplary diagrams illustrating a smart home system installed in a user's residence to monitor a user's behavior, and FIG. 4 is a diagram for explaining pre-classified daily life collection items collected from the user's behavior.

Referring to FIG. 3A , the IoT device may be installed indoors of a user's residence and implemented as a smart home system. IoT devices can be composed of indoor electronic products or sensors installed in various locations indoors. The IoT device may transmit information about the operation of the electronic product to the dementia prediction server or transmit a signal generated by detecting a user's action to the dementia prediction server. As shown in the interior plan 10 of FIG. 3A and the interior plan 10-2 of FIG. 3B, power meters are attached to electronic products, and motion sensors, audio sensors, and contact sensors are installed throughout the room. can When the electronic product is operating, the power meter may measure power consumption of the electronic product and transmit information on the power consumption to the dementia prediction server. The motion sensor and the contact sensor may transmit a detected signal to the dementia prediction server when a user's motion is detected, such as when the user moves from the living room to the bathroom or from the bathroom to the living room. The audio sensor may transmit information about the user's voice and sound generated by the user's movement to the dementia prediction server. In this case, the wearable device may transmit the user's location information and the user's biological signal to the dementia prediction server.

3A is a design diagram 10-1 of a room showing sensors arranged according to a first rule in a room where normal users are located. In one embodiment, sensors may be disposed according to a first rule defining the arrangement of sensors for detecting motions of users of the normal group, and data may be collected from the placed sensors.

Data may be collected through a vibration sensor, a door sensor, a motion sensor, a temperature and humidity sensor, a smart plug, and a LIDAR sensor in a room where a user of the normal group is located. Vibration sensors may be placed in medicine cabinets, refrigerators, rice cookers, sinks, and toilets, door sensors may be placed in front doors, microwave ovens, and closets, and motion sensors may be placed in shoe cabinets, TVs, drawers, and desks. Temperature and humidity sensors are placed in the bathroom and gas range, smart plugs are placed in the TV and bed, and lidar sensors can be placed on the sofa on the wall side.

Arrangement of the sensor according to the first rule described with reference to FIG. 3A may indicate a location where ADL collection items of users of the normal group can be most effectively collected. Specifically, the arrangement of the sensors according to the first rule may be an optimal arrangement for observing the user of the normal group, which was acquired through the experimental results while going through the experiment process of comparing the motions of the user of the dementia patient and the user of the normal group. there is. Accordingly, the arrangement of the sensors according to the first rule may be an arrangement method set to collect the maximum amount of ADL collection items of users in the normal group using the minimum number of sensors.

3B is a design diagram 10-2 of a room showing sensors arranged according to the second rule in the room where a user of the dementia group is located. In one embodiment, sensors may be disposed according to a second rule defining the arrangement of sensors for detecting motions of a user of the dementia group, and data may be collected from the disposed sensors.

Data may be collected through a vibration sensor, a door sensor, a motion sensor, a temperature and humidity sensor, a smart plug, and a LIDAR sensor in a room where a user of the dementia group is located.

Vibration sensors may be placed on refrigerators, sinks, rice cookers, medicine cabinets, toilets, vacuum cleaners, and trash cans, door sensors may be placed on front doors and microwave ovens, and motion sensors may be placed on balcony windows and TVs. Temperature and humidity sensors are placed in toilets and gas stoves, smart plugs are placed in washing machines, electric pads and TVs, and lidar sensors are placed on sofas on the wall side to collect data on the user's location and movement in the room.

Arrangement of the sensor according to the second rule described with reference to FIG. 3B may indicate a location where the ADL collection items of the user of the dementia group can be most effectively collected. Specifically, the arrangement of the sensors according to the second rule may be an optimal arrangement for observing the user of the dementia group learned through the experiment results while going through the experiment process of comparing the motions of the user of the dementia patient and the normal group. there is. Accordingly, the arrangement of the sensors according to the second rule may be an arrangement method set to maximize the collection of ADL collection items of a user in the dementia group using the minimum number of sensors.

Dementia prediction method based on daily life ability information according to an embodiment of the present invention monitors the daily life of a dementia patient and predicts dementia early using data collected through the arrangement of sensors described in FIGS. 3A and 3B. Therefore, in the field of dementia, where early prediction is more important than treatment, there is an advantage in being able to predict dementia patients at an early stage with minimal cost and minimal time.

In addition, since the dementia prediction method based on daily living ability information using the arrangement of sensors according to the first rule and the second rule has excellent performance in predicting dementia early compared to the results predicted by the arrangement of other sensors, In the future, the method of predicting dementia according to the sensor positions of the first rule and the second rule may be regarded as a technology that is likely to be studied in depth in the field of dementia technology.

Referring to FIG. 4 , the dementia prediction server may classify the daily life ability information according to the daily life collection items 30 set in advance. In another embodiment, the dementia prediction server can cluster information received from IoT devices and wearable devices into predetermined items by machine learning, and according to the clustered items, daily life collection items 30 can be set. Activity of Daily Living (ADL) includes using the phone, preparing food, cooking, using household appliances, doing housework, locking the door, grooming, managing belongings, taking medication, talking about recent events, managing money, It can be classified into items such as toilet use, indoor behavior and walking, transportation use and loitering.

Among daily life ability information, 'keeping appointments and gatherings', 'leisure and hobbies', 'going out a short distance', 'shopping', and 'use of public transportation' can be located based on GPS, but 'user' Since there are limitations in collecting daily life ability information such as 'confirmation of whether the person has acted properly regarding recent actions', 'talking about recent events', 'managing belongings', and 'managing money', there are limitations It can be collected through guardian interviews or observer identification.

In particular, by using the data collected in FIGS. 3A and 3B, the user's life patterns of cooking, cleaning, door locking, using home appliances, housework, grooming, taking medicine, and moving indoors can be determined

For example, a cooking operation may use a sensor disposed on a microwave oven, a refrigerator, a rice cooker, a sink faucet, or a gas stove, and a cleaning operation may use a sensor disposed on a sink faucet (washing dishes) or a washing machine. The following operation may use sensors disposed on a front door, an electric pad, and a TV.

The operation of using a home appliance may use a sensor disposed in a washing machine, a microwave oven, or a TV, and an operation of doing housework may use a sensor disposed in a washing machine, a vacuum cleaner, or a sink in the case of washing dishes. For grooming, a sensor placed in the bathroom can be used for a shower, a sensor placed in a closet for dressing, a sensor placed in a medicine cabinet for taking medicine, and a main living space for indoor movement. It can be determined through the movement line in (living room).

As such, with reference to FIGS. 3 and 4 , a detailed method of acquiring daily life ability information by monitoring indoor behavior of a user has been described, and step S200 will be described again with reference to FIG. 2 .

In step S200, a lifestyle pattern for the user's behavior may be determined using the daily living ability information. The dementia prediction server may determine the user's life pattern according to the daily life collection items. The dementia prediction server can determine the user's life pattern for phone use, food preparation, cooking, and home appliance use according to the items collected in daily life. For example, the dementia prediction server analyzes voice volume, word usage pattern, voice tone, phone usage time, data usage, and direction of movement when determining the life pattern of phone use to determine the user's usual life pattern of phone use. can decide

Hereinafter, a detailed method of determining a user's life pattern will be described with reference to FIGS. 5 and 6 .

Referring to FIG. 5 , in step S201, data according to a user's personal characteristics may be obtained. Specifically, data on the characteristics of the spatial structure in which the user lives, data on the physical characteristics of the user, and data on the degree of dementia symptoms of the user may be obtained. Data according to the personal characteristics may be input from a medical institution, a public institution, or a user's terminal. In step S203, a lifestyle pattern for a user's behavior may be determined in consideration of the input data.

Referring to FIG. 6 , the dementia prediction server may receive data about characteristics of a spatial structure in which a user lives, and determine a life pattern of the user according to the characteristics of the spatial structure. Data on the characteristics of spatial structure can be divided into house structure, sleeping pattern, meal pattern, etc. Type 1 is 'a structure using a dining table without using a bed in a detached house', and Type 2 is 'a bed and dining table in an apartment' , and Type 3 may be data of a 'structure that does not use beds and dining tables in rental APT'.

The dementia prediction server may receive data on the user's physical characteristics and determine the user's life pattern according to the user's physical characteristics. Data on physical characteristics can be divided into gender, age, family status, etc. Type 1 is 'male, 70's living alone', Type 2 is 'woman, 70's, married couple', Type 3 is 'male, 80's, It may be the data of living alone.

The dementia prediction server may receive data about the degree of the dementia symptom of the user and determine the life pattern of the user according to the degree of the dementia symptom. Data on the degree of dementia symptoms can be divided into normal, MCI, and AD, and Type 1 can be 'normal', Type 2 can be 'MCI', and Type 3 can be 'AD' data.

In step S203, a life pattern reflecting the individual characteristics of the user can be determined by considering the data on the characteristics of the structure of the space where the user lives, the data on the physical characteristics of the user, and the data on the degree of dementia symptoms of the user. there is.

In step S300 of FIG. 2 , when it is detected that the user's behavior deviates from the life pattern through daily life ability information, it can be analyzed that the user is in a dementia risk state. For example, assuming that the life pattern of phone use is less than 1 minute and the repetition of the same word within the same sentence is less than 3 times, the user's phone usage time is over 1 hour and within the same sentence When repetition of the same word exceeds 10 times or occurrence of an epigram sentence is detected, the dementia prediction server may determine that the user is in a dementia risk state. In addition, as will be described below in FIG. 7 , the dementia prediction server may determine that the user is in a dementia risk state when the user's moving path continues to repeat the same path.

In one embodiment, the difference in ADL between a normal user and a dementia user can be derived through a first-step analysis (difference between normal and dementia by ADL activity), a second-step analysis (Dementia Index of ADL), and a third-step (dementia prediction model) analysis. there is.

In the first stage analysis (the difference between normal and dementia by ADL activity), in the case of cooking, the total number of times of cooking by period, the temperature and humidity cycle of the gas stove, the cooking time, and the cooking movement can be analyzed. In the case of cleaning, it can be analyzed by checking the total number of cleanings per period, opening/closing of faucets, and movement lines. In the case of door locking, it can be analyzed through the opening/closing of the front door, the electric pad, and the ON/OFF of the TV. In the case of the use of home appliances, it can be analyzed through the number of times of use, use time, use cycle, and movement of each product. In the case of doing housework, it can be analyzed through the frequency of cleaning/washing, time, cycle, and movement. In the case of grooming, it can be analyzed through the shower-toilet temperature and humidity cycle, the number of closet doors and time zone. In the case of taking medicine, it can be analyzed through the number of times of use of the medicine cabinet and the time of use. In the case of indoor movement, it can be analyzed through movement from living space to kitchen to room, movement distance, movement time, and location.

In the second stage analysis (Dementia Index of ADL), (a) average ADL activity score is calculated, (b) abnormal behavior score is calculated for each ADL activity (number of times to take medicine, time zone, number of cleaning, number of showers, number of cooking, etc.), (C) An ADL abnormal behavior score can be calculated with a combination of two or more sensors. For example, if the front door is closed & the lighting is low, but the TV or electric pad is ON, if the sink faucet is open but the location continues to stay in a location other than the kitchen, if you wake up at night and wander around, a score will be calculated as an abnormal behavior. can Then, (d) indoor movement and abnormal patterns (pattern analysis different from normal movement: staying in a different place for a very long time, wandering frequently, etc.) are analyzed, and long-term ADL changes can be analyzed and verified.

Step 3 (dementia prediction model) collects ADL data, calculates average (A) and ADL dementia score (B, C, D, E) for each ADL, and classifies based on dementia analysis rules and machine learning-based classification (machine Application of learning and deep learning analysis algorithms and development of analysis models through learning), through which dementia can be predicted early.

So far, the overall process of the dementia prediction method based on daily life ability information according to an embodiment of the present invention has been described. Hereinafter, a detailed method of predicting dementia by monitoring a user's loitering among the daily life collection items of FIG. 4 will be described with reference to FIGS. 7 and 8 .

FIG. 7 is a diagram for explaining step S300 of FIG. 2 in detail, and FIG. 8 is a diagram for explaining a method of analyzing a user's loitering according to the user's moving path.

Referring to FIG. 7 , motions in the user's daily life according to the pre-classified daily life collection items may be detected in step S301. At this time, whether the user is loitering according to the movement may be monitored. In step S302 , the user's loitering may be analyzed using the information on the movement path collected by the wearable device.

In step S303, a method of determining the user's loitering may be different depending on whether the user is located indoors or outdoors. If it is determined in step S303 that the user is located indoors, steps S304 to S306 may be performed, and if it is determined that the user is located outdoors, steps S307 to S309 may be performed.

When it is determined in step S303 that the user is located indoors, information about the use of the building may be checked in step S304. The use information of the building may include information about what purpose the building is used for, such as a school, an institution, a company, an entertainment establishment, a restaurant, and the like. In step S305, if it is determined that the building is a place out of the user's life pattern by analyzing whether the building information according to the use information of the building is related to the user's life pattern, step S306 may be performed. At this time, if it is determined that the user stays longer than the threshold time, it may be determined that the user is loitering in step S310. If the user leaves the building without staying longer than the threshold time, return to step S302 and the user's behavior may be monitored. If the building in step S305 is not a place out of the user's life pattern, step S307 may be performed, and the operation of step S307 will be described later.

If it is determined in step S303 that the user is not located indoors, it may be determined that the user is staying outdoors. Then, in step S307, it may be determined whether the user's location is a place the user frequently visits. If the frequently visited place is an exceptional place designated in advance in step S308, the user's bio-signal may be analyzed. In step S309, when it is determined that the user's cognitive ability is reduced based on the bio-signal, it may be determined that the user is wandering in step S310. In step S309, when it is determined that the user's cognitive ability is not deteriorated on the bio-signal, it is determined that the user is not loitering, and the user's behavior can be monitored by returning to step S302.

Referring to FIG. 8 , the dementia prediction server may receive GPS data 70 of a user's movement path from the wearable device. The dementia prediction server may determine whether the user is located in a building through this path.

In this case, the dementia prediction server may generate loitering detection results 80 classified into cases in which the user's physical condition is normal, loitering, and loitering exceptions. The dementia prediction server may receive biosignals such as the user's heart rate and walking speed. The dementia prediction server may determine whether the user's physical state is in a normal pattern and generate data 80 as a result of the loitering detection. In one embodiment, the dementia prediction server may determine whether the user is loitering by determining whether the user is repeatedly moving without a purpose to the visit place based on information about the user's heart rate, walking speed, and visit destination. .

Dementia prediction method based on daily life ability information according to an embodiment of the present invention analyzes the user's loitering by analyzing the user's moving path and collects information on the use of buildings to analyze the user's loitering. It has the advantage of being able to predict precisely.

In addition, when it is determined in step S300 whether or not it is detected as out of the life pattern, it may be determined using a symbol character.

Specifically, a time-sequentially expressed chart may be used to use symbolized symbol characters. The time-sequentially expressed chart is a time-sequentially expressed chart of the values of the daily life ability information obtained after the life pattern is determined. For example, a chart expressed in time series is a chart created in the process of performing SAX (Symbolic Aggregate approXimation' or 'Symbolic ApproXimation') analysis method. In the chart expressed in time series, the time series data is divided into time windows by the SAX analysis method, and after the dimension is reduced, the complex time series shape can be converted into simple data.

First, a chart expressed in time series can be converted into a stair-type chart by averaging with an average value for each predetermined interval. For example, if daily living ability information in the form of a time series is input, this daily living ability information is converted into a simplified piecewise aggregate approximation (PAA) in a stepped data form. PAA is a method of splitting a time series and connecting them to each average value.

Thereafter, the step-type chart may be symbolized by assigning a symbol letter corresponding to an average value in the step-type data chart. For example, a simplified PPA in the form of step-like data is expressed in the form of a symbol again in text form (aabbbccb), which is called SAX. This is a method of mapping symbolic symbols.

Using these coded symbol characters, it can be determined whether the user's behavior deviates from the life pattern. For example, it is determined whether the symbolized symbol character matches the user's life pattern, and if not matched, the user's behavior may be determined to deviate from the user's life pattern.

In this case, it may be determined that the user's behavior deviates from the lifestyle pattern based on whether the user's lifestyle pattern is similar to the lifestyle pattern at risk of dementia by calculating the distance shown in the time series pattern similarity graph expressed in time series.

Specifically, in order to find similar behavior patterns, after performing distance calculation (Euclidean, DTW, etc.) A similar pattern of behavior can be identified. That is, it is possible to classify ADL behavioral patterns between normal and dementia elderly through a distance calculation method for behavioral patterns similar to those of dementia patients.

In addition, when it is determined in step S300 whether it is detected as being out of the lifestyle pattern, it may be determined using the similarity between the daily living ability information and the activity time of the lifestyle pattern. This embodiment will be described with reference to FIGS. 9 and 10 . 9 and 10 are diagrams for explaining specific examples of calculating the similarity of activity times.

In this step, if there is an overlapping time between the activity time of the daily living ability information obtained after determining the life pattern and the activity time of the life pattern, the similarity of the activity time may be calculated. For example, the similarity of activity time may be used when calculating the similarity with respect to time of detailed ADL activities (cooking, using electric appliances, etc.). The similarity of activity times may be used when ADL activity times to be compared overlap. Whether the user is normal or at risk of dementia may be classified using the similarity of activity time. The similarity of activity time can be calculated by Equation 1.

Here, E0 is the observation point and end time of the daily living ability information obtained after the life pattern is determined, and En is the average end time of the daily living ability information obtained after the life pattern is determined. Sn is the average starting time of the daily living ability information obtained after determining the life pattern, S0 is the average required time at the point of observation of the daily living ability information obtained after the living pattern is determined, and Dn is the average time required for the daily living ability information It's time.

As a first case, as shown in FIG. 9, when Sn (average start time) of the life pattern is after S0 (start time of observation) and En (average end time) is after E0 (end time of observation), similarity of activity time can be calculated as E0 - Sn / Dn * 100.

As a second case, as shown in FIG. 10, when Sn (average start time) of the life pattern is before S0 (start time of observation) and En (average end time) is before E0 (end time of observation), similarity of activity time can be calculated as En - S0 / Dn * 100.

In addition, when it is determined in step S300 whether it is detected as out of the lifestyle pattern, it may be determined using a lidar sensor. This embodiment will be described with reference to FIG. 11 . 11 is a diagram for explaining a user's visiting frequency and staying time measured using a lidar sensor.

In this step, the user's visiting frequency and staying time may be measured using a LiDAR sensor installed indoors. Afterwards, the user's sense of balance and walking ability may be determined using the measured visit frequency and stay time.

Specifically, the frequency of visit and the time of stay can be checked by using a lidar sensor installed indoors, such as indoor movement frequency, movement area, movement distance, and main living space in the room for each time period. In this step, whether indoor loitering, sleep activity at night, and sleep disorder may be analyzed through the indoor movement frequency, movement area, movement distance, and main living space for each time period.

For example, the moving distance that a patient with dementia moves indoors can be measured using a LiDAR sensor installed indoors. In the case of dementia patients, the daytime movement distance is shorter than that of non-dementia patients, and the night time movement distance is longer than that of non-dementia patients. In addition, since the moving distance of a patient with dementia is not constant and irregular compared to the moving distance of an ordinary person who is not a patient with dementia, the moving distance of a patient with dementia can be used as an analysis result for identifying a patient with dementia. However, since the moving distance of dementia patients is sometimes not clearly distinguished from the moving distance of non-dementia patients, it is necessary to measure the moving distance very precisely in order to classify dementia patients and non-dementia patients by using the moving distance. There is a need.

Since the invention according to this embodiment detects the movement of a patient with dementia very precisely through the lidar sensor installed in the room of the actual living space where the user actually lives, the indoor movement frequency, movement area, movement distance, and major factors in the room by time period There is an advantage in that it is possible to easily analyze whether or not a person has dementia using only the living space. In addition, since this embodiment utilizes a lidar sensor without a camera function for 'privacy protection' of the user, it has the advantage of being able to monitor infringements while protecting the privacy of dementia patients. In addition, the present embodiment has the advantage of being able to monitor infringement without causing inconvenience to the user by using a lidar to minimize inconvenience in the user's daily life without using a sensor that the user wears inconveniently.

As shown in the indoor design 90 of FIG. 11, the visit frequency measured using LIDAR can be displayed as a large and small size of the circle size 91, and the staying time measured using LIDAR has different shades ( 92) can be distinguished. The frequency of visits to Zone 1 is 97 and the duration of stay is measured as 22 minutes, while the frequency of visits to Zone 2 is 200 and the duration of stay is measured as 277. In this way, visit frequencies and stay times of zones 3 to 6 can be measured, and the sum of visit frequencies and stay times for all zones can be calculated.

In this step, the user's sense of balance and walking ability can be determined by applying the measured frequency of visits and length of stay of the user to any one of timed up and go (TUG), walking velocity, cadence, and gait stability ratio.

Timed up-and-go (TUG) is a test that measures the time required for a subject to sit in a chair, stand up, turn around at a point 3 meters in front of them, and return to their original place. .

Walking velocity is a test that measures the walking time to 10 meters and 5 meters.

Cadence is a test in which steps are counted while walking 5 meters and measured as steps/s' (time taken).

Gait stability ratio (GSR) is divided by walking velocity in Cadence [steps/m], and a higher value means less dynamic movement of the subject.

Afterwards, the sense of balance and walking ability are determined in the same way as above, and using these results, the difference between the night and day activities of the normal and dementia elderly is analyzed to determine whether there is a sleep disorder (activity) during the night and indoors during the day. Loitering can be detected. In addition, whether the user is normal or at risk of dementia can be classified using the results of the sense of balance and walking ability. For example, the difference in ADL between a normal person and an elderly person with dementia can be classified using the results of balance and walking ability. 12 is a schematic diagram of a dementia prediction system based on daily life ability information according to another embodiment of the present invention.

Referring to FIG. 12, the dementia prediction system based on daily life ability information according to another embodiment of the present invention includes a dementia prediction server 100, a wearable device 200, an IoT device 300, a service DB 400, and early dementia A diagnosis web 500 and a sensor management web 600 may be included. Dementia prediction system based on daily life ability information according to another embodiment of the present invention is mostly the same as the configuration described in FIG. This is an embodiment in which a web and sensor management web 600 is added. Since the wearable device 200 and the IoT device 300 in this embodiment are the same as those in FIG. 1 , descriptions thereof will be omitted.

The dementia prediction server 100 may be implemented as Mobius Server, an IoT Open Platform. The dementia prediction server 100 can perform data storage access, requester, and responder, use MQTT, COAP, and HTTP protocols, and can be implemented as an MQTT proxy server using Node Js platform. Mobius Server can process data by connecting to Filesystem, Mobius DB, and HDFS. The dementia prediction server 100 may store and manage information received from the wearable device 200 and the IoT device 300 in the service DB 400. The dementia early diagnosis web 500 can manage the user's risk of dementia through the web using information stored in the service DB 400, and the sensor management web 600 can separately collect and manage only sensor information.

Dementia prediction system based on daily living ability information according to the present embodiment in step S300

Hereinafter, an exemplary computing device 500 capable of implementing the devices described in various embodiments of the present invention will be described with reference to FIG. 13 .

13 is an exemplary hardware configuration diagram illustrating a computing device 500 .

As shown in FIG. 13, the computing device 500 loads one or more processors 510, a bus 550, a communication interface 570, and a computer program 591 executed by the processor 510. It may include a memory 530 and a storage 590 for storing the computer program 591. However, only components related to the embodiment of the present invention are shown in FIG. 13 . Accordingly, those skilled in the art to which the present invention pertains can know that other general-purpose components may be further included in addition to the components shown in FIG. 13 .

The processor 510 controls the overall operation of each component of the computing device 500 . The processor 510 may include at least one of a Central Processing Unit (CPU), a Micro Processor Unit (MPU), a Micro Controller Unit (MCU), a Graphic Processing Unit (GPU), or any type of processor well known in the art. can be configured to include Also, the processor 510 may perform an operation for at least one application or program for executing a method/operation according to various embodiments of the present disclosure. Computing device 500 may include one or more processors.

Memory 530 stores various data, commands and/or information. Memory 530 may load one or more programs 591 from storage 590 to execute methods/operations according to various embodiments of the present invention. For example, when the computer program 591 is loaded into the memory 530, the logic (or module) shown in FIG. 4 may be implemented on the memory 530. An example of the memory 530 may be RAM, but is not limited thereto.

The bus 550 provides a communication function between components of the computing device 500 . The bus 550 may be implemented as various types of buses such as an address bus, a data bus, and a control bus.

The communication interface 570 supports wired and wireless Internet communication of the computing device 500 . The communication interface 570 may support various communication methods other than Internet communication. To this end, the communication interface 570 may include a communication module well known in the art.

Storage 590 may non-temporarily store one or more computer programs 591 . The storage 590 may be a non-volatile memory such as read only memory (ROM), erasable programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), flash memory, or the like, a hard disk, a removable disk, or a device well known in the art. It may be configured to include any known type of computer-readable recording medium.

Computer program 591 may include one or more instructions in which methods/operations according to various embodiments of the invention are implemented. When computer program 591 is loaded into memory 530, processor 510 may execute the one or more instructions to perform methods/acts according to various embodiments of the present invention.

In one embodiment, including a processor, a network interface, a memory that is executed by the processor to load a computer program, and a storage for storing the computer program, the computer program, via an IoT device installed indoors An instruction for obtaining daily living ability information related to dementia prediction by monitoring the user's indoor behavior, an instruction for determining a lifestyle pattern for the user's behavior using the daily living ability information, and an instruction for determining the daily living ability information of the user through the daily living ability information It may include an instruction for determining that the user is at risk of dementia when the behavior of the user is detected as deviating from the lifestyle pattern.

The methods according to the embodiments of the present invention described so far can be performed by executing a computer program implemented as a computer readable code. The computer program may be transmitted from the first computing device to the second computing device through a network such as the Internet, installed in the second computing device, and thus used in the second computing device. The first computing device and the second computing device include both a server device, a physical server belonging to a server pool for a cloud service, and a fixed computing device such as a desktop PC.

The computer program may be stored in a recording medium such as a DVD-ROM or a flash memory device.

Although the embodiments of the present invention have been described with reference to the accompanying drawings, those skilled in the art to which the present invention pertains can be implemented in other specific forms without changing the technical spirit or essential features of the present invention. can understand that Therefore, the embodiments described above should be understood as illustrative in all respects and not limiting.

Claims (32)

In a method performed by a computing device,
Obtaining daily life ability information related to dementia prediction by monitoring indoor behavior of a user through an IoT device installed indoors;
receiving data about the degree of dementia symptoms of the user;
determining a lifestyle pattern for the user's behavior by using the daily living ability information and whether the level of dementia symptoms of the user is included among normal, MCI, and AD; and
Determining that the user is in a dementia risk state when the user's behavior is detected as deviating from the life pattern through daily living ability information obtained after determining the life pattern,
Dementia prediction method based on daily living ability information. According to claim 1,
The step of obtaining the daily life ability information,
Including the step of detecting motions in the user's daily life according to pre-classified daily life collection items using data obtained from an IoT device installed indoors,
Dementia prediction method based on daily living ability information. According to claim 2,
The step of detecting the user's motion in daily life,
Monitoring the use of the electronic product by a user by detecting the power amount of the electronic product through a power meter installed in the room, detecting the user's movement through a motion sensor, and detecting the user's touch through a vibration sensor. doing,
Dementia prediction method based on daily living ability information. According to claim 2,
Further comprising the step of analyzing the user's indoor loitering according to the user's movement information,
Dementia prediction method based on daily living ability information. According to claim 4,
The step of analyzing the user's indoor loitering according to the user's movement information,
Further comprising detecting whether the user is wandering by determining whether the user is repeatedly moving without a purpose based on information about the user's heart rate and walking speed.
Dementia prediction method based on daily living ability information. According to claim 2,
The step of detecting the user's motion in daily life,
Collecting data according to a first rule defining the arrangement of sensors for detecting motions of users of the normal group,
Dementia prediction method based on daily living ability information. According to claim 6,
Collecting data according to the first rule,
Collecting data through vibration sensors, door sensors, motion sensors, temperature and humidity sensors, smart plugs and lidar sensors in the room where the normal group of users are located,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of disposing the vibration sensor in the medicine cabinet, refrigerator, rice cooker, sink, and toilet,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of arranging the door sensor in the front door, microwave oven, and closet,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of arranging the motion sensor in the shoe cabinet, TV, drawer and desk,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of disposing the temperature and humidity sensor in the bathroom and gas range,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of arranging the smart plug to the TV, washing machine, electric pad and fan,
Dementia prediction method based on daily living ability information. According to claim 7,
Collecting data according to the first rule,
Including the step of the lidar sensor being disposed adjacent to the sofa to collect data of the user's location and movement in the room,
Dementia prediction method based on daily living ability information. According to claim 2,
The step of detecting the user's motion in daily life,
Collecting data according to a second rule defining the arrangement of sensors for detecting motions of a user of a dementia group,
Dementia prediction method based on daily living ability information. According to claim 14,
Collecting data according to the second rule,
Collecting data through vibration sensors, door sensors, motion sensors, temperature and humidity sensors, smart plugs and lidar sensors in the room where the user of the dementia group is located,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of arranging the vibration sensor in a refrigerator, sink, rice cooker, medicine cabinet, toilet, vacuum cleaner, and trash can,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of disposing the door sensor on the front door and the microwave oven,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of arranging the motion sensor on the balcony window and the TV,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of disposing the temperature and humidity sensor in the bathroom and gas range,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of arranging the smart plug in a washing machine, an electric pad, a TV and an electric fan,
Dementia prediction method based on daily living ability information. According to claim 15,
Collecting data according to the second rule,
Including the step of the lidar sensor being disposed adjacent to the sofa to collect data of the user's location and movement in the room,
Dementia prediction method based on daily living ability information. According to claim 1,
The step of determining the life pattern is,
Receiving data on the characteristics of the spatial structure in which the user lives, and determining the user's life pattern according to the characteristics of the spatial structure,
Dementia prediction method based on daily living ability information. According to claim 1,
The step of determining the life pattern is,
Receiving data on the user's physical characteristics and determining the user's life pattern according to the user's physical characteristics,
Dementia prediction method based on daily living ability information. delete According to claim 1,
Determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the life pattern,
expressing the values of daily life ability information obtained after determining the life pattern as a time-sequential chart;
converting the time-sequentially expressed chart into a staircase chart by averaging an average value for each predetermined interval;
symbolizing the step chart by allocating a symbol character corresponding to the average value in the step chart; and
Including the step of determining whether the behavior of the user deviates from the life pattern using the coded symbol character,
Dementia prediction method based on daily living ability information. According to claim 1,
Determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the life pattern,
determining whether a life pattern of the user is similar to a life pattern at risk of dementia by calculating a distance shown in a time series pattern similarity graph expressed in time series;
calculating a degree of similarity between activity times when an overlapping time exists between the activity time of daily living ability information obtained after determining the life pattern and the activity time of the life pattern; and
Classifying whether the user is normal or at risk of dementia using the similarity of the activity time,
The similarity of the activity time,
Calculated as {Infimum(E0, En) - supremum(Sn, S0)}/Dn * 100,
The E0 is the end time of the observation point of the daily living ability information obtained after the life pattern is determined,
En is the average end time of daily living ability information obtained after determining the life pattern,
Sn is the average starting time of daily living ability information obtained after determining the life pattern,
S0 is the average required time at the point of observation of the daily living ability information obtained after determining the life pattern,
Dn is the average time required for daily life ability information,
Dementia prediction method based on daily living ability information. According to claim 1,
Determining that the user is at risk of dementia when it is detected that the user's behavior deviates from the life pattern,
Determining the user's sense of balance, walking ability, frequency of visiting a specific space, staying time, and indoor moving distance by using a LiDAR sensor installed indoors; and
Classifying whether the user is normal or at risk of dementia using the sense of balance and walking ability, the frequency of visits and staying time in a specific space, and the indoor moving distance
Dementia prediction method based on daily living ability information. The method of claim 27,
The step of classifying whether the user is normal or at risk of dementia using the sense of balance and walking ability,
Determining the sense of balance and walking ability using any one of Timed up and go (TUG), Walking velocity, Cadence, and Gait stability ratio,
Dementia prediction method based on daily living ability information. processor;
network interface;
a memory that is executed by the processor and loads a computer program; and
Including a storage for storing the computer program,
The computer program,
An instruction to obtain daily life ability information related to dementia prediction by monitoring a user's indoor behavior through an IoT device installed indoors;
instructions for receiving data about the degree of dementia symptoms of the user;
an instruction for determining a lifestyle pattern for the user's behavior by using the daily living ability information and whether the level of dementia symptoms of the user is included among normal, MCI, and AD; and
Including an instruction for determining that the user is in a dementia risk state when the user's behavior is detected as deviating from the life pattern through the daily living ability information,
Dementia prediction device based on daily living ability information. According to claim 29,
The instructions for obtaining the daily life ability information include:
Including an instruction for detecting an operation in the user's daily life according to a pre-classified daily life collection item using data obtained from an IoT device installed indoors,
Dementia prediction device based on daily living ability information. 31. The method of claim 30,
The instruction for detecting the user's motion in daily life,
The user's use of the electronic product is monitored by detecting the power amount of the electronic product through a power meter installed in the room, the user's motion is detected through a motion sensor, and the sound generated by the user's motion is detected using an audio sensor. And including instructions for detecting the user's touch by a touch sensor,
Dementia prediction device based on daily living ability information. 31. The method of claim 30,
The instruction for detecting the user's motion in daily life,
Including an instruction for analyzing the user's indoor loitering according to the user's movement information,
Dementia prediction device based on daily living ability information.

Images