KR20210010548A - Information processing device, information processing method and program - Google Patents

Abstract

It provides an information processing apparatus, an information processing method, and a program capable of easily and accurately predicting human or animal dementia. The information processing device 10 includes a data acquisition unit 40 that acquires at least one selected from a group of human or animal surrounding environment data, biometric data, behavior data, image or video data, and audio data. , A prediction unit 30b for predicting the onset of behavioral/psychological symptoms (BPSD) in dementia or a timing of onset of the onset based on the data acquired by the data acquisition unit 40. The prediction unit 30b can predict BPSD by using at least one of inference analysis, regression analysis, hotspot analysis, proximity analysis, and spatiotemporal analysis. And a learning unit 30c that evaluates the difference between the prediction of the BPSD and the BPSD by the prediction unit 30b, and learns a function or a classifier based on the evaluation result.

Description

Information processing device, information processing method and program

The present invention relates to an information processing apparatus, an information processing method, and a program related to prediction of dementia (認知症; dementia).

[0002] A device for supporting the diagnosis of dementia has been proposed (see Patent Document 1).

Japanese Patent Laid-Open Publication No. 2017-217052

[0004] An object of the present invention is to provide an information processing apparatus, an information processing method, and a program capable of easily and accurately predicting the onset of dementia in humans or animals.

[0005] 1. Information processing device

The information processing device of the present invention,

At least selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. A data acquisition unit that acquires one type;

And a prediction unit for predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or a timing of the onset of the symptoms based on the data acquired by the data acquisition unit.

[0006] In the present invention,

The prediction unit uses at least one of inference analysis, regression analysis, hotspot analysis, proximity analysis, and spatiotemporal analysis to predict behavior and psychological symptoms in the dementia of the human or animal. It is possible to do.

[0007] In the present invention,

And an index conversion unit for indexing the data acquired by the data acquisition unit,

The prediction unit may perform prediction using a function or a classifier for inputting an index value indexed by the indexing unit and outputting a predicted value for behavioral and psychological symptoms in the dementia of the human or animal.

[0008] In the present invention, the classifier may be a support vector machine (SVM), a neural network, or a linear regression model.

[0009] In the present invention, the prediction unit evaluates the difference (差) between the prediction about the behavioral and psychological symptoms in the human or animal dementia and the actual behavioral and psychological symptoms of the human or animal, and It is possible to include a learning unit that learns the function or classifier based on the evaluation result.

[0010] In the present invention, the learning unit evaluates the difference between the prediction about the behavioral and psychological symptoms in the human or animal dementia predicted by the prediction unit and the actual behavioral and psychological symptoms of the human or animal So,

The relationship between the evaluation of the difference between the prediction of the behavioral and psychological symptoms in the human or animal dementia by the prediction unit and the actual behavioral and psychological symptoms of the human or animal, and the information on the data acquired by the data acquisition unit It is possible to learn the above function or classifier by deep learning.

[0011] In the present invention, the function is configured to include a parameter and a coefficient related to a numerical value related to the data acquired by the data acquisition unit,

The learning unit may adjust the coefficients by deep learning.

[0012] In the present invention,

The data acquisition unit acquires the environmental data and the biometric data,

The environmental data includes temperature and humidity around the person or animal,

The biometric data may include the human or animal's pulse rate, respiratory rate, and body temperature.

[0013] In the present invention, the environmental data and the biometric data, and information on the behavioral and psychological symptoms in the human or animal dementia corresponding to the environmental data and the biometric data are associated and stored It is possible to include a data storage unit.

[0014] In the present invention, the environmental data and the biometric data, and the environmental data and the biometric data and the corresponding information on the behavioral and psychological symptoms in dementia of the human or animal are correlated and configured It is possible to include a stored storage unit.

[0015] In the present invention, it is possible to include a coping method derivation unit for generating or selecting a coping method based on the prediction contents of the behavioral and psychological symptoms in the dementia of the human or animal predicted by the prediction unit Do.

[0016] In the present invention, the actual behavior and psychological symptoms of the human or animal are determined according to a judgment algorithm by the behavior and psychological symptom determination unit,

The learning unit evaluates the difference between the determination result by the behavioral/psychological symptom determination unit and the determination result of the behavioral/psychological symptom of the person or animal actually judged by a person, and learning the determination algorithm based on the evaluation result. It is possible to do.

[0017] 2. Information processing method

The information processing method of the present invention,

The data acquisition unit is selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. The process of acquiring at least one type,

The prediction unit includes a step of predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or a timing of onset of the onset based on the data acquired by the data acquisition unit.

[0018] 3. Program

The program of the present invention,

Computer

The data acquisition unit is selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. Acquiring at least one species, and

It is for causing the prediction unit to execute a step of predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or the onset time of the onset based on the data acquired by the data acquisition unit.

By predicting the onset or timing of the onset of BPSD, by early coping, it is possible to prevent the onset of BPSD in advance, and to significantly reduce the occurrence of BPSD itself. Since BPSD can be predicted in advance, it is possible to reduce the burden on caregivers.

1 is a diagram showing the overall configuration of an information processing flow.
2 is a diagram showing a configuration example of an information processing device.
3 is a diagram showing a configuration example of an information processing system.
4 is a diagram illustrating a processing flow of information processing.
5 is a diagram showing an example of data of five sense data.
6 is a diagram showing an example of environmental data data.
7 is a diagram showing an example of data of biometric data.
8 is a diagram showing an example of data of care recording.
9 is a diagram schematically showing a neural network.
10 is a diagram showing the correlation between behavior and BPSD.
Fig. 11 is a diagram summarizing output contents of each step in data processing.
12 is a diagram showing a correspondence table between a BPSD and a coping method.
13 is a flow schematic diagram of BPSD determination and learning of a determination algorithm.

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

[0022] An embodiment will be described taking predictions about behavioral and psychological symptoms in human dementia. In the following, behavioral and psychological symptoms in human dementia are simply referred to as "BPSD".

1. Outline of information processing device

(1) Information processing device

As shown in Fig. 1, the information processing device 10 is based on a function F1 for obtaining data, a function F2 for converting the data into an index, and a BPSD (human dementia) based on a prediction algorithm. The function of predicting the onset possibility and content of behavioral and psychological symptoms, and the onset timing (F3), the function of outputting the predicted result of BPSD (F4), and the function of generating and outputting the predicted BPSD coping method ( F5), a function of determining whether BPSD has occurred (F6), a function of verifying after confirmation and execution of symptoms (F7), and a function of learning a prediction algorithm based on the verification result (F7). Include.

[0023] (2) BPSD (behavior psychological symptoms of dementia)

Behavioral symptoms are symptoms revealed by observation. As evaluation items of behavioral symptoms, for example, roaming, dangerous manipulation, concealment, false work (symptoms of repeating movements that seem meaningless when viewed from the surroundings), trouble, collection, screaming and excitement are mentioned. Psychological symptoms are symptoms revealed by an interview. As an evaluation item of psychological symptoms, resentment, delusion, etc. are mentioned, for example. Behavioral and psychological symptoms may or may not appear depending on the person.

2. Configuration of the information processing device

(1) Functional configuration of the information processing device

As shown in Fig. 2, the information processing apparatus 10 is connected to devices connected to various communication networks via a communication network (for example, the Internet). As a device, for example, various IoT sensors can be applied, and specifically, an environmental sensor 60a, a biometric sensor 60b, a camera 60c, a microphone 60d, a position detection sensor (BLE tag, etc.) 60e ), a door opening/closing sensor 60f, a urinary amount sensor 60g in the bladder, and the like.

The information processing apparatus 10 includes a data acquisition unit 40, a processing unit 30, and a storage unit 20, as shown in FIG. 3.

[0026] The data acquisition unit 40 acquires at least one selected from a group of environment data of a person, biometric data of a person, behavior data of a person, image or video data of a person, and audio data of a person. The data acquisition unit 40 acquires, for example, environmental data and biometric data, and the environmental data includes a person's ambient temperature and humidity, and the biometric data can include a person's pulse, respiratory rate, and body temperature. Do.

The processing unit 30 includes an indexing unit 30a, a prediction unit 30b, a learning unit 30c, a coping method derivation unit 30d, and a BPSD determination unit 30e.

The index conversion unit 30a is for converting the data acquired by the data acquisition unit 40 into an index.

The prediction unit 30b predicts the onset of BPSD or the onset timing of the BPSD based on the data acquired by the data acquisition unit 40. The prediction unit 30b may predict BPSD by using at least one of inference analysis, regression analysis, hot spot analysis, proximity analysis, and spatiotemporal analysis.

The prediction unit 30b predicts using at least one of a function (20a) or a classifier (20b) for inputting an indexed index value in the indexing unit 30a and outputting a predicted value for BPSD It is possible to do.

The classifier 20b may be a support vector machine (SVM), a neural network, or a linear regression model. The function 20a may be configured to include parameters and coefficients related to numerical values for data acquired by the data acquisition unit 40. The neural network may be, for example, a three-layer feed forward type neural network composed of an input layer, an intermediate layer, and an output layer, as shown in FIG. 9.

The learning unit 30c evaluates the difference between the prediction of BPSD of the prediction unit 30b and the actual human behavior and psychological symptoms, and based on the evaluation result, a function 20a or a classifier 20b ) To learn at least one of them. The learning unit 30c can supply the machine learning algorithm to learn the function 20a or the classifier 20b. Specifically, the learning unit 30c evaluates the difference between the prediction about the BPSD predicted by the prediction unit 30b and the actual human behavior and psychological symptoms, and the prediction about the BPSD of the prediction unit 30b It is possible to learn the function (20a) or the classifier (20b) by deep learning on the relationship between the evaluation content of the difference between actual human behavior and psychological symptoms and the information on the data acquired by the data acquisition unit 40. Do. The learning unit 30c may adjust the coefficients by deep learning.

The coping method derivation unit 30d derives a coping method based on the onset possibility and content and onset time of the predicted BPSD.

BPSD determination unit (30e) is to determine whether or not BPSD based on the input data.

In realizing each function of the processing unit 30, it is possible to extract a feature amount of the input data and perform each function processing.

In the storage unit 20, for example, a function 20a, a classifier 20b, a correlation table 20c, a correspondence table 20d, and a data storage unit 20e may be stored.

[0037] The correlation table 20c may be composed of a database configured by associating environmental data and biometric data with information on BPSD corresponding to environmental data and biometric data. The correspondence table 20d may be composed of a database configured by associating the predicted occurrence of BPSD and the occurrence timing and a coping method. The data storage unit 20e includes data acquired by the data acquisition unit 40, data obtained by processing data by the processing unit 30, data inputted by the input unit 52, and various judgments by the processing unit 30 in advance. Data referenced for this purpose are stored.

The data acquisition unit 40 may acquire processed data, or may acquire raw data (measurement data or input data) and process it in the processing unit 30. Examples of the data acquired by the data acquisition unit 40 include the following.

[0039] (a) Five senses data

The five sense data, for example, as shown in Fig. 5, detects sound source detection, sound type identification, olfactory perception, face authentication, motion detection, distance range function, three-dimensional detection, and the like. Five sense data can be measured with a camera or a microphone. In addition, it is not particularly limited as long as it is a device capable of obtaining these data.

[0040] (b) environmental data

As for environmental data, as shown in FIG. 6, for example, temperature, humidity, illuminance, water quantity, and atmospheric pressure are measured, or movement distance is detected. Environmental data can be measured by a temperature sensor, a humidity sensor, an illuminance sensor, an atmospheric pressure sensor, or the like. In addition, it is not particularly limited as long as it is a device capable of obtaining these data.

[0041] (c) biometric data

Biometric data, for example, as shown in Fig.7, detects heart rate, respiration, bed departure (leaving the bed), sleep (non-REM sleep, detection of REM sleep), awakening, excretion, amount of exercise, body temperature, and the like. The sensor for detecting the biometric data is not particularly limited as long as it is capable of detecting the biometric data, and may be constituted by one or a plurality of sensors. Taking sleep as an example, a Doppler sensor can determine the non-REM sleep state and the REM sleep state. Regarding the excretion, the timing of excretion can be detected by a combination of an open/close sensor of a toilet door and a person detecting sensor in the toilet, or by analyzing a tag (BLE tag) having a communication function or an image of a camera. With regard to excretion, the content of excretion can be detected in consideration of the timing of the night urination time and the time required for excretion. The amount of exercise is calculated using software that calculates calories consumed, calculated from the distance traveled by a tag (BLE tag) having a communication function, or, from the record of a caregiver, moving the body such as exercise, cleaning, washing, etc. By calculating time, you can calculate the amount of exercise and calories consumed. In addition, it is not particularly limited as long as it is a device capable of obtaining these data.

[0042] (d) Record of nursing care (介護)

The items of the care record are, for example, as shown in FIG. 8, the needs of the person receiving care or the caregiver, the concern of the person receiving the care or the caregiver, subjective symptoms of the person receiving care, and the caregiver observed What they did, assess the condition of the person receiving care, the person receiving the care or the caregiver’s impressions, the caregiver’s response, the caregiver's speech, and what the caregiver did.

In nursing care recording, in order to improve the prediction accuracy of BPSD, it is preferable to have the following recording.

1) BPSD occurred

2) The time (time zone) and place (the role of the place, the relationship of the location, etc.) when the BPSD occurred must be clearly recorded

3) Conditions such as environmental factors (temperature, humidity, illuminance, air pressure, noise, off-flavor, presence of people, contents of conversation, behavior or attitude of the subject) at the time of BPSD should be recorded

4) The behavior and attitude of the subject before and after the occurrence of BPSD must be recorded. In addition, the environmental factors must be recorded.

By combining these, the role of the place where the BPSD occurs, the distance at which the place exists, the time zone in the life rhythm, the environmental factors for the place and time, the action content before and after it, and the relationship with the person before and after it. From the viewpoint of relations and the like, it is possible to infer a situation in which the probability of occurrence of BPSD increases with high precision.

(E) data acquisition method

The data acquired by the acquisition unit may be acquired as follows, for example.

1) Periodic acquisition of environmental data (temperature, humidity, air pressure, illuminance)

2) Periodic acquisition of biometric data (heart rate, respiration rate, sleep status, bed departure status)

3) Acquire the location information of the target user using the BLE tag and calculate the moving distance

4) By interlocking with AI cameras, it obtains the location information of the target user and calculates the moving distance.

5) Linking the above acquired data with the user ID and periodically sending it to the information processing device

[0046] In the above 1) periodic acquisition of temperature and humidity, the temperature and humidity sensor may transmit acquired data in a 10-minute interval from real time by default. Each gateway can make it possible to acquire the data of a target sensor by linking it with a user ID. Analysis becomes possible by acquiring a plurality of sensor data from one gateway and simultaneously supporting a plurality of users.

[0047] In the periodic acquisition of 2) atmospheric pressure and illuminance above, by connecting an expansion module mounted with an atmospheric pressure sensor and an illuminance sensor to an IoT gateway, information of atmospheric pressure and illuminance in a period of 10 minutes from real time Can be updated. It is possible to respond to each analysis by being able to support multiple users' data from one gateway.

[0048] In the periodic acquisition of 3) heart rate, respiration rate, sleep state, and out of bed state of the above, the Doppler sensor unit is installed on the wall around the bed, and the heart rate and respiration rate when entering the bed (着床時) , It is possible to provide notification of sleep conditions and events when leaving the bed or going to bed. Heart rate, respiration rate, and sleep status are measured every 10 minutes from real time and notified to the IoT gateway for analysis. You may be notified immediately when an event occurs when you leave the bed/bedding.

[0049] (2) Physical configuration of the information processing device

The information processing device 10 is composed of an electronic calculator, for example, a large electronic calculator such as a supercomputer, an electronic calculator equipped with a GPU (Graphics Processing Unit), a plurality of personal computers, a quantum computer, etc. Consists of

The processing unit 30 can be realized by a computing device such as a CPU. The storage unit 20 can be stored in a known storage device such as a ROM, a hard disk, or an external storage device (CD, DVD, etc.), for example. The storage unit 20 may be configured as a separate entity independently from the computing device, or may be configured in the computing device. The information processing device 10 may be configured with one electronic calculator or may be configured with a plurality of electronic calculators.

[0051] A program for executing information processing on the information processing device 10 can be stored in a storage device (eg, ROM, hard disk) or the like included in the information processing device 10.

The data acquisition unit 40 may be configured as, for example, a receiving unit capable of receiving data from a communication network.

The input unit may be made of a known input device such as a keyboard, a mouse, and a touch panel. As the display portion 54, a known display such as a liquid crystal display and an organic EL display can be applied. The transmission unit 56 transmits information to a terminal connected through a communication network, and a known transmission device can be applied.

[0054] 3. Information processing method

(1) Outline of information processing method

An information processing method will be described with reference to FIG. 4.

The data acquisition unit 40 acquires, from an IoT device or the like, the five senses data, environmental data, biometric data, behavioral data, and nursing care records of a person receiving care (S1).

The data acquired by the index conversion unit 30a is converted into an index, and an evaluation index parameter is calculated (S2).

The prediction unit 30b analyzes the evaluation index parameter and predicts BPSD (S3).

The coping method derivation unit 30d generates a coping method based on the predicted occurrence of BPSD and the time of occurrence (S4).

The data acquisition unit 40 acquires behavioral data and the like, and evaluates whether or not BPSD actually occurs by the BPSD determination unit 30e (S5).

Based on the difference between the BPSD and the actual symptoms predicted by the learning unit 30c, based on deep learning, learning (update) of the function 20a or the identifier (neural network) is performed (S6).

11 shows measurement data and processing data obtained in each step. Each of the data in FIG. 11 can be associated with each other and stored in the data storage unit 20e of the storage unit 20.

[0062] (2) Example of indexing processing

(a) environmental indicators

As for the environmental index, the degree of discomfort calculated for each temperature + humidity, atmospheric pressure, and illuminance can be used as an index. First, for temperature + humidity, it is determined whether or not it is unpleasant based on whether or not it is in the comfortable range using a graph showing a known comfort range.

As a treatment for this, first, it is determined whether or not the humidity is in the range of 40 to 70%, and if it is outside the range, it is regarded as unpleasant.

In addition, when the humidity is between 40% to 70%, it is determined whether the temperature is appropriate using the following formula.

When temperature and humidity exist in the range of this formula, it is considered to be comfortable, and when outside this range, it is considered to be unpleasant. When it is judged to be unpleasant, the distance from the center point of the comfort range to the current line of temperature and humidity and the intersection of the comfort range is used as an index of discomfort.

Next, for an unpleasant index of atmospheric pressure, the following equation is applied with a standard atmospheric pressure of 1013 hPa as a standard value.

Here, when the atmospheric pressure is P, the discomfort index becomes P'.

Next, it is an unpleasant index of illuminance, which depends on the state of the user, and the illuminance as a reference varies. The reference illuminance is set to I ₀ , and the following calculation formula is used.

With the illuminance as I, I', which is an unpleasant index of illuminance, is calculated. In addition, the standard illuminance is calculated as 200 lux when active, 50 lux when stable, and 20 lux when sleeping.

[0067] (b) biomarker

As for the heart rate, the heart rate for one minute is P and the average heart rate is P ₀ , and the biomarker P'is obtained by the following calculation formula.

Regarding the respiration rate, the respiration rate for 1 minute is B, the average respiration rate is B ₀ , and the biomarker B'is obtained by the following calculation formula.

[0069] (3) prediction processing

(a) prediction logic

When the prediction logic is P, it is possible to calculate P by performing matrix calculation based on the following equation.

Here, p1 to p5 represent each parameter, and α to ε'represent the degree of items related to the occurrence of behavioral and psychological symptoms (BPSD) of dementia.

[0070] (b) prediction of the occurrence time of BPSD

The prediction of the occurrence time of BPSD can derive the relevance of the time zone based on several reference points. As the index, the following time points are illustrated as candidates for the reference point.

1) 0 o'clock (date change point)

2) Sunrise (sunset)

3) Wake up (起床)

4) meal

[0071] For the occurrence of BPSD, the elapsed time from each reference point can be calculated, and a correlation can be derived from the bias. The relationship can be derived by performing regression analysis on the timing at which BPSD occurs and the elapsed time from the reference point. At this time, for waking up, whether sleep or nap at night is separately classified, and meals may be classified by types such as breakfast, lunch, dinner, and snacks. When the BPSD occurrence time is set to point A, the elapsed time from 0 o'clock, waking up at night, breakfast, lunch, and dinner, the elapsed time from the nearest one, after the nap, the elapsed time from the nap, and the elapsed time from the snack, In addition, it may be used as an index.

[0072] When a new job occurs, there is a tendency to increase the likelihood of having a memory impairment in which you forget the previous thing. When they become helpless without knowing what to do, they tend to engage in homologous behaviors repeating the same actions to find something, and to wander around without purpose. Due to excessive involvement by instructional orders, there is a tendency to show Yi No-seong, who feels annoying and uncomfortable, and suddenly becomes angry with a loud voice. In such a flow, there is a specific time difference, and it is possible to make predictions in consideration of time series.

In predicting the BPSD, unlike image analysis such as photographs, a function 20a or a classifier 20b in consideration of an influence due to a time series may be used. The learning of the function 20a or the classifier 20b is close to language, speech, and video analysis among deep learning, so it is sufficient to learn using long short-term memory (LSTM).

[0074] (c) BPSD prediction from environmental factors

In the occurrence of BPSD, since the influence of the surrounding environment is large, it is possible to analyze environmental factors, especially temperature, humidity, air pressure, illuminance, odor, and noise when BPSD occurs. For temperature and humidity, the degree of comfort can be diagnosed and used as an index based on the comfort index graph for the relationship between temperature and humidity. With respect to atmospheric pressure, illuminance, odor, and noise, each of the indicators is analyzed for the relationship with the occurrence of BPSD as a single indicator, and regression analysis is performed respectively, so that the influence and degree of each indicator on the BPSD can be calculated.

[0075] (d) prediction from natural language interpretation

BPSD can also be analyzed by natural language analysis. After specifying the part of speech included in a question or inquiry in natural language processing, a hypothesis is created, and then the hypothesis is supported or evidence is sought. According to the statistical model method of the evidence weighted score, the corresponding method is assigned based on the onset state of BPSD.

[0076] (e) predictions from behavior

As shown in Fig. 10, the correlation of BPSD can be seen from the behavior of a dementia patient. Thereby, the onset of BPSD can be predicted from the behavior of a dementia patient.

[0077] (4) How to cope

Based on the symptoms of BPSD, it is possible to create a coping method based on the function 20a or the classifier 20b. Further, based on the correspondence table 20d between the symptoms of BPSD and the corresponding method, a method of coping with BPSD may be selected. The correspondence table 20d as shown in FIG. 12 may be stored in the storage unit 20.

In addition, in the coping method, when it is necessary to create a diet (menu) content of a meal, nutritional calories may be calculated based on the amount of exercise or calories consumed, thereby helping to prepare a diet.

[0079] (5) Verification

(a) evaluation of occurrence

Verification of the onset of BPSD can be performed using the BPSD problem behavior evaluation scale (TBS: Troublesome Behavior Scale). TBS describes the destructive behavior or burden of elderly dementia patients.

[0080] TBS defines 15 items describing destructive behaviors or burdensome behaviors of elderly dementia patients and their frequency, and the applicant has reliability and validity as a measure to evaluate behavior abnormalities of persons with dementia Was confirmed. Furthermore, by performing an evaluation (e.g., 5-step evaluation) based on the frequency observed by the caregiver over a predetermined period of time (e.g., past 1 month), the problem behavior relatively frequently observed in a person with dementia Can be verified. The frequency can be divided into evaluation categories such as "once a day or more", "several times a week", "several times a month", and "none".

By analyzing the voice condition (音聲病態), BPSD analysis can be performed. Using Sensibility Technology, the ratio of four emotions and ``excitement'' in utterances: ``anger,'' ``joy,'' ``sad,'' and ``pyeongsang''. By analyzing the degree of, the degree can be displayed. Verification of the onset of BPSD may be performed by brain image diagnosis.

[0082] As shown in FIG. 13, in the evaluation of whether or not BPSD has onset and, in the case of onset, the onset content and the onset timing, whether or not BPSD has developed in the dementia patient is determined, the data acquisition unit 40 ) Can be determined by the BPSD determination unit 30e on the basis of the data acquired or the data input by the input unit 52. This BPSD judgment algorithm can be learned by the learning unit. The method of learning the BPSD determination algorithm can be performed in the same manner as the learning of the prediction algorithm of the prediction unit 30d.

Specifically, the data acquisition unit 40 acquires the data (F11), the index conversion unit 30a converts the data into an index, and determines whether the BPSD determination unit 30e corresponds to BPSD, and (F13), outputs the determination result of whether BPSD is applicable. It is used to verify the situation as to whether the input actual BPSD corresponds to the BPSD determination unit 30e and the result determined by the BPSD determination unit 30e (F15), and the learning unit 30c determines whether the BPSD determination unit 30e corresponds to the BPSD. The judgment algorithm is learned (F16).

(b) Evaluation of coping methods

Through the evaluation of BPSD coping, the reliability of the derivation of the care method is evaluated according to the degree of evaluation obtained. By performing analytics on a large amount of data collected from the care field, collecting insights and converting them into inspiration, it is possible to derive an appropriate care method. In the evaluation of how to cope with BPSD, based on the data acquired by the data acquisition unit 40 or the data inputted by the input unit 52, the processing unit 30 is asked whether or not BPSD has developed in a dementia patient. It is possible to verify the right and wrong of the coping method. Based on this verification result, the learning unit 30d learns the derivation algorithm of the coping method derivation unit 30d, and the derivation algorithm can be updated. The method of learning the derivation algorithm of the coping method derivation unit 30d can be performed in the same manner as the learning of the prediction algorithm of the prediction unit 30d.

[0084] 3. Action and effect

By predicting the onset or timing of onset of BPSD, it is possible to prevent the onset of BPSD in advance by early coping, and to significantly reduce the occurrence of BPSD itself. Since BPSD can be predicted in advance, the burden on caregivers can be reduced.

[0085] The above embodiments can be variously changed within the scope of the gist of the present invention. In the above embodiments, the behavioral and psychological symptoms of human dementia have been described, but the present invention is not limited thereto, and can be widely applied to the behavioral and psychological symptoms of dementia in animals.

[0086] 4. Application example

(1) AI system

The information processing device according to the present embodiment can be applied to the following AI (Artificial Intelligence) system. That is, the AI system includes an information processing device formed by a data storage unit, an analysis unit, an emotion processing unit, and a planning unit according to a knowledge expression method, data collected by automatic identification, automatic response, and automatic notification from an IoT gateway, and a human interface ( It can be configured as a data mining unit that combines nursing care record data collected by human interface), natural language interpretation, and speech recognition, and further performs statistical analysis.

Based on the knowledge data that may be used for care, it is possible to continuously grasp the state of the elderly with dementia, predict the onset of BPSD peculiar from the state change, and suggest an appropriate response method stored in advance. It has the following effects.

(a) Focus not only on caregivers but also on people with dementia

(b) Multidimensionally and flexibly, according to the needs of the caregiver and dementia.

(c) combination in case of adaptation of drug therapy

(d) In information sharing, it is possible to analyze in correlation with data obtained from IoT information while taking care based on the following information (a) to (c).

1) Specific information on how to deal with BPSD

2) Information to ensure the physical safety and well-being of people with dementia

3) Information to cope with difficult ADL (daily life movement) information

[0088] The basic research of AI can be applied and utilized to respond to behavioral and psychological symptoms (BPSD) of dementia based on reasoning or learning.

[0089] (2) Function of AI system

The AI system can have the following functions.

(a) Expert system

The knowledge of experts is accumulated as a rule, and the following references are lexicographically made by a system that solves problems using a method of reasoning.

(b) speech recognition

Talk to your smartphone or tablet. It specifies who is talking, and the computer understands what is said and makes sentences.

(c) natural language processing

The meaning of the sentenced information is understood by a computer, and the information is classified and recorded in the F-SOAIP (Life Support Record Method) so that information can be searched.

(d) emotional treatment

Based on the knowledge of cognitive science and ergonomics, sensations such as warm or cold feeling are received from environmental sensors and realized on a computer.

(e) image recognition

Contents taken with a camera, etc., are understood by a computer, and the brightness or color tone of the living room is classified as pleasant or unpleasant.

(f) machine learning

It is a system that finds consistent rules (patterns) from data collected by IoT sensors or nursing care records. It is strongly related to the statistical field of mathematics, and is interpreted and organized using, for example, the following statistical techniques.

[0090] 1) FTA (Fault Tree Analysis) analysis

As an analysis method for finding the cause from the result, the occurrence route, the cause of occurrence, and the probability of occurrence are analyzed for an event that is not desired to occur. In order to analyze the frequency of occurrence of BPSD, it logically follows the potential risk (fault) of the cause ("fault" here refers to events such as environment or human error). , By adding the probability of each occurrence, the probability that a basic event can occur is calculated.

2) ETA (Event tree analysis) analysis

Also referred to as risk prediction analysis, the process until BPSD occurs is interpreted based on the probability of occurrence and the success or failure of the response to the countermeasure (care).

3) HAZOP (Hazard and Operability Study) analysis

As a method of analyzing from the rule of thumb, in the scene where behavioral and psychological symptoms are onset, whether the condition has arisen as a result of the effect of not taking appropriate care for the parameter representing the subject's behavior or personality characteristics. How to analyze whether or not.

The parameters referred to herein are analyzed including data such as body temperature, breathing, pulse, temperature, humidity, atmospheric pressure, sleep time, etc., obtained from the IoT gateway, and guide words such as whether or not the event occurs repeatedly.

4) Information search

It is a system that finds what is needed for dementia care from the accumulated data.

[0091] (g) reasoning

It is a system to derive an answer without contradiction by integrating certain laws from various patterns.

1) Hot Spot Analysis

This is an analysis method that considers the space (place) where the past behavioral/psychological symptoms occur as a space (place) with a high probability of occurrence of behavioral/psychological symptoms.

2) Regression Methods

In addition to the past BPSD, other variables related to BPSD, such as environment and human relations, are used as independent variables to predict future BPSD by regression analysis.

3) Near-Repeat Methods

Based on the spatiotemporal proximity of one BPSD and the next BPSD, the future BPSD is predicted.

4) Spatiotemporal Analysis

Behavioral and psychological symptoms are predicted from movement patterns according to time changes in BPSD occurrence and various factors affecting them.

5) Risk Terrain Analysis

The risk plane is created from the proximity to the spatial factors affecting BPSD, and future BPSD occurrence is predicted.

[0092] (h) data mining

It is a system that combines database technology and machine learning to find information that is thought to be helpful from a large amount of unorganized data.It predicts BPSD by data mining techniques such as class classification and clustering, and provides a response method. To derive.

(i) Human interface

A smartphone or tablet device that allows a caregiver to more easily operate a device such as a computer can be applied.

(j) planning

In deriving an appropriate care response for BPSD, a system for determining the order in which care should be performed can be applied.

(k) multi-agent

It is a system that proposes a care method by investigating the occurrence of BPSD again by F-SOAIP for information when caregivers who solve BPSD gather and solve complex problems at the care site.

By utilizing the above, it can be applied to the response of BPSD of dementia based on reasoning or learning.

[0094] (3) Care method

Dementia requires care methods tailored to different behavioral and psychological symptoms. This is because different care methods affect dementia patients. The information processing device is capable of accumulating various documents and past information including information on standard care methods and the most excellent personal assistance methods (對人援助方法). From all of these, it is possible to specify what is the care method that provides the best option to be employed by the caregiver in nursing a dementia patient. Under the guidance of a nursing care expert, the dementia-responsive IoT service may collect the knowledge necessary to acquire skills in the nursing field. This is defined as "the corpus of knowledge in dementia care". The creation of the corpus can be started by loading a number of relevant literature into a dementia-responsive AI.

In addition, in the creation of the corpus, professional staff may be involved in order to select information, or to exclude all information that is not related to the field of the problem or out of date information, information considered to be lagging behind. This is defined as “curation of dementia-responsive content”.

Pre-processed by curation, indexes or other metadata that enables more efficient cooperation with content can be built from the field as a life support recording method (F-SOAIP). A dementia-responsive AI trained in a question-and-answer pair can continue to learn by continuously communicating with a bot. Furthermore, when new information is released, the dementia-responsive AI is also updated, so that it can always adapt to changes in knowledge and linguistic interpretation in a given field, and prepare to specify new insights or patterns hidden in the information. It is a method of specifying a part of speech included in a question or inquiry by natural language processing, generating a hypothesis, and then finding support or evidence for that hypothesis.

According to a statistical model method of evidence-weighted scores, a response method can be assigned based on the onset state of behavioral and psychological symptoms. The dementia-response AI can evaluate its reliability based on the degree of evaluation obtained by deriving a care method through a successful example of responding to behavioral and psychological symptoms. In short, the dementia-responsive AI allows you to always derive an appropriate care method by running analytics on a large amount of data collected from the care field, collecting insights and converting them into inspiration.

[0098] The above embodiments can be variously changed within the scope of the gist of the present invention.

The present invention can be applied as a management system in the care of dementia patients or animals suffering from dementia.

10: information processing device
20: memory
20a: function
20b: classifier
20c: correlation table
20d: correspondence table
20e: data storage unit
30: processing unit
30a: indexing unit
30b: prediction unit
30c: Department of Learning
30d: coping method generator
30e: BPSD judgment unit
40: data acquisition unit
52: input
54: display
56: transmitter
70: communication network

Claims (14)

At least selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. A data acquisition unit that acquires one type;
An information processing apparatus comprising a prediction unit for predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or a timing of the onset of the symptoms based on the data obtained by the data acquisition unit. The method of claim 1,
The prediction unit uses at least one of inference analysis, regression analysis, hotspot analysis, proximity analysis, and spatiotemporal analysis to predict behavior and psychological symptoms in the dementia of the human or animal. An information processing device that performs. The method according to claim 1 or 2,
And an index conversion unit for indexing the data acquired by the data acquisition unit,
The prediction unit is an information processing device in which prediction is performed using a function or a classifier for inputting an index value indexed by the indexing unit and outputting a predicted value for behavioral and psychological symptoms in the dementia of the human or animal. The method according to claim 1 or 2,
The classifier is an information processing device of a support vector machine (SVM), a neural network, or a linear regression model. The method according to claim 1 or 2,
The prediction unit evaluates the difference between the prediction of behavioral/psychological symptoms in the human or animal dementia and the actual behavioral/psychological symptoms of the human or animal, and based on the evaluation result, the function or An information processing device including a learning unit that learns a classifier. The method of claim 5,
The learning unit evaluates the difference between the prediction about the behavioral/psychological symptoms in the dementia of the human or animal predicted by the prediction unit and the actual behavioral/psychological symptoms of the human or animal,
The relationship between the evaluation of the difference between the prediction of the behavioral and psychological symptoms in the human or animal dementia by the prediction unit and the actual behavioral and psychological symptoms of the human or animal, and the information on the data acquired by the data acquisition unit An information processing device that learns the function or classifier by deep learning. The method of claim 5,
The function is configured to include parameters and coefficients related to numerical values for the data acquired by the data acquisition unit,
The learning unit is an information processing device that adjusts the coefficients by deep learning. The method according to any one of claims 1, 2, 6 and 7,
The data acquisition unit acquires the environmental data and the biometric data,
The environmental data includes temperature and humidity around the person or animal,
The biometric data is an information processing device including a pulse rate, respiratory rate, and body temperature of the human or animal. The method of claim 8,
An information processing device comprising a data storage unit in which the environmental data and the biometric data and information on the behavioral and psychological symptoms in the human or animal dementia corresponding to the environmental data and the biometric data are associated and stored . The method of claim 8,
Information including a storage unit in which a correlation table is stored in which the environmental data and the biometric data and information on behavior and psychological symptoms in the human or animal dementia corresponding to the environmental data and the biometric data are correlated and configured Processing device. The method according to any one of claims 1, 2, 6, 7, 9 and 10,
An information processing apparatus comprising a coping method derivation unit for generating or selecting a coping method based on the prediction contents of the behavior and psychological symptoms in the human or animal dementia predicted by the prediction unit. The method of claim 5,
The actual behavior and psychological symptoms of the human or animal are determined according to a judgment algorithm by the behavior and psychological symptom determination unit,
The learning unit evaluates the difference between the determination result by the behavioral/psychological symptom determination unit and the determination result of the behavioral/psychological symptom of the person or animal actually judged by a person, and learning the determination algorithm based on the evaluation result. An information processing device that performs. The data acquisition unit is selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. The process of acquiring at least one type,
The information processing method, comprising a step of predicting, based on the data acquired by the data acquisition unit, for predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or a timing of the onset of the onset of the human or animal. Computer
The data acquisition unit is selected from the group of environment data of a person or animal, biometric data of the person or animal, behavior data of the person or animal, image or video data of the person or animal, and audio data of the person or animal. Acquiring at least one species, and
A program for causing a prediction unit to execute a step of predicting the onset of behavioral/psychological symptoms in the dementia of the human or animal or a timing of onset of the onset based on the data acquired by the data acquisition unit.

Images