WO2019221252A1

WO2019221252A1 - Information processing device, information processing method, and program

Info

Publication number: WO2019221252A1
Application number: PCT/JP2019/019591
Authority: WO
Inventors: 政治羽田野; 直久皆川; 孝雄内藤
Original assignee: 一般社団法人認知症高齢者研究所
Priority date: 2018-05-17
Filing date: 2019-05-16
Publication date: 2019-11-21
Also published as: CN112135568A; KR20210010548A; JP7285046B2; JPWO2019221252A1

Abstract

Provided are an information processing device, an information processing method, and a program capable of easily and accurately performing prediction about dementia of a human or an animal. An information processing device 10 comprises: a data acquisition unit 40 that acquires at least one selected from the group consisting of surrounding environment data, biological data, behavior data, image or video data, and sound data about a human or an animal; and a prediction unit 30b that predicts onset of behavioral and psychological symptoms of dementia (BPSD) or the onset timing thereof on the basis of the data acquired by the data acquisition unit 40. The prediction unit 30b can perform BPSD-related prediction by using at least one of reasoning analysis, regression analysis, HotSpot analysis, proximity analysis, and time space analysis. A learning unit 30c that evaluates the difference between the BPSD and the BPSD-related prediction performed by the prediction unit 30b and that learns a function or a classifier on the basis of the evaluation result, is included.

Description

Information processing apparatus, 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.

An apparatus that supports diagnosis of dementia has been proposed (see Patent Document 1).

JP 2017-217052 A

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.

1. Information processing apparatus Information processing apparatus of the present invention,
At least selected from the group of environmental data around a person or animal, biological 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 for acquiring one type;
A predicting unit that predicts the onset of the behavioral / psychological symptoms in the dementia of the person or animal or the onset time based on the data acquired by the data acquiring unit.

In the present invention,
The said prediction part can perform the prediction regarding the behavioral / psychological symptom in the dementia of the said person or animal using at least 1 sort (s) of reasoning analysis, regression analysis, HotSpot analysis, proximity analysis, and spatiotemporal analysis.

In the present invention,
Including an indexing unit for indexing the data acquired by the data acquisition unit,
The prediction unit receives the index value indexed by the indexing unit and performs prediction using a function or a classifier that outputs a prediction value related to behavioral / psychological symptoms in dementia of the person or animal be able to.

In the present invention, the classifier may be an SVM (Support Vector Machine), a neural network, or a linear regression model.

In the present invention, the prediction unit evaluates the difference between the prediction of behavior / psychological symptoms in dementia of the person or animal and the actual behavior / psychological symptoms of the person or animal, and based on the evaluation result, the function or A learning unit that performs learning of the classifier can be included.

In the present invention, the learning unit evaluates the difference between the prediction of the behavior / psychological symptoms in the human or animal dementia predicted by the prediction unit and the actual behavior / psychological symptoms of the human or animal,
Prediction regarding behavior / psychological symptoms in dementia of the person or animal of the prediction unit, evaluation contents of difference between actual behavior / psychological symptoms of the person or animal, information on data acquired by the data acquisition unit, and In this context, the function or classifier can be learned by deep learning.

In the present invention, the function includes a parameter relating to a numerical value related to data acquired by the data acquisition unit, and a coefficient,
The learning unit can adjust the coefficient by deep learning.

In the present invention,
The data acquisition unit acquires the environmental data and the biological data,
The environmental data may include temperature and humidity around the person or animal, and the biological data may include pulse, respiratory rate, and body temperature of the person or animal.

The present invention includes a data storage unit in which the environmental data and the biological data are stored in association with information related to behavioral and psychological symptoms in dementia of the person or animal corresponding to the environmental data and the biological data. be able to.

In the present invention, a correlation table configured by associating the environmental data and the biometric data with information related to behavioral and psychological symptoms in dementia of the person or animal corresponding to the environmental data and the biometric data is stored. Can include a storage unit.

In the present invention, a coping method deriving unit that generates or selects a coping method can be included based on the prediction content related to the behavioral / psychological symptoms in dementia of the human or animal predicted by the prediction unit.

In the present invention, the actual behavior / psychological symptom of the person or animal is determined by a behavior / psychological symptom determination unit according to a determination algorithm,
The learning unit evaluates a difference between the determination result of the behavior / psychological symptom determination unit and the determination result of the behavior / psychological symptom of the person or animal actually determined by the person, and the determination algorithm based on the evaluation result Can learn.

2. Information processing method The information processing method of the present invention comprises:
A data acquisition unit includes a group of environmental data around a person or animal, biological 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. Obtaining at least one selected from:
A prediction unit including a step of predicting the onset or the onset time of the behavior / psychological symptoms in dementia of the person or animal based on the data acquired by the data acquisition unit.

3. Program The program of the present invention
In the computer, the data acquisition unit includes environmental data around the person or animal, biological 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. Obtaining at least one selected from the group of:
The predicting unit is configured to execute a step of predicting the onset or the onset time of the behavior / psychological symptoms in the dementia of the person or animal based on the data acquired by the data acquiring unit.

・ Predicting the onset of BPSD or the time of onset, and taking early measures to prevent the onset of BPSD and to greatly reduce the occurrence of BPSD itself. Since BPSD can be predicted in advance, the burden on the caregiver can be reduced.

It is a figure which shows the whole structure of the flow of information processing. It is a figure which shows the structural example of information processing apparatus. It is a figure which shows the structural example of an information processing system. It is a figure explaining the processing flow of information processing. It is a figure which shows the example of data of five sense data. It is a figure which shows the example of data of environmental data. It is a figure which shows the example of data of biometric data. It is a figure which shows the example of data of a care record. It is a figure which shows a neural network typically. It is a figure which shows the correlation of action and BPSD. It is the figure which arranged the output content of each step in data processing. It is a figure which shows the correspondence table of BPSD and a coping method. It is a flow outline figure about judgment of BPSD and learning of a judgment algorithm.

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

The embodiment will be described by taking predictions regarding behavioral and psychological symptoms in human dementia as an example. Hereinafter, the behavioral / psychological symptoms in human dementia are simply referred to as “BPSD”.
1. Overview of Information Processing Device (1) Information Processing Device As shown in FIG. 1, the information processing device 10 has a function F1 for acquiring data, a function F2 for indexing the data, and a BPSD (human The function F3 for predicting the possibility and content of the onset and the timing of the onset of the behavior / psychological symptoms in dementia, the function F4 for outputting the prediction result about the BPSD, and the generation / output of the coping method of the predicted BPSD A function F5, a function F6 that determines whether or not BPSD has occurred, a function F7 that is verified after symptom confirmation and execution, and a function F7 that learns a prediction algorithm based on the verification result are included.

(2) BPSD (behavior / psychological symptoms of dementia)
Behavioral symptoms are symptoms revealed by observation. Evaluation items for behavioral symptoms include, for example, dredging, dangerous operation, concealment, temporary work, trouble, collection, screaming / excitement. Psychological symptoms are symptoms revealed by interviews. As evaluation items for psychological symptoms, for example, depression, delusion, and the like can be given. Behavioral and psychological symptoms may or may not appear depending on the person.

2. Configuration of Information Processing Device (1) Functional Configuration of Information Processing Device Information processing device 10 is connected to devices connected to various communication networks via a communication network (for example, the Internet) as shown in FIG. Yes. As the device, for example, various IoT sensors can be applied. Specifically, the environmental sensor 60a, the biological sensor 60b, the camera 60c, the microphone 60d, the position detection sensor (BLE tag, etc.) 60e, the door open / close sensor 60f, the intravesical An example is a urine volume sensor 60 g.

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

The data acquisition unit 40 acquires at least one selected from the group of human environmental data, human biological data, human behavior data, human image or video data, and human voice data. The data acquisition unit 40 acquires, for example, environmental data and biological data. The environmental data can include the ambient temperature and humidity around the person, and the biological data can include the person's pulse, respiratory rate, and body temperature.

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

The indexing unit 30a is for indexing the data acquired by the data acquiring unit 40.

The prediction unit 30b predicts the onset of BPSD or its onset based on the data acquired by the data acquisition unit 40. The prediction unit 30b may perform prediction regarding BPSD using at least one of inference analysis, regression analysis, HotSpot analysis, proximity analysis, and spatiotemporal analysis.

The prediction unit 30b can input the index value indexed by the indexing unit 30a and perform prediction using at least one of the function 20a or the classifier 20b that outputs a prediction value related to BPSD.

The classifier 20b can be an SVM (Support Vector Machine), a neural network, or a linear regression model. The function 20a can include a parameter related to a numerical value related to data acquired by the data acquisition unit 40 and a coefficient. For example, as shown in FIG. 9, the neural network can be a three-layer feedforward neural network including an input layer, an intermediate layer, and an output layer.

The learning unit 30c evaluates the difference between the prediction regarding the BPSD of the prediction unit 30b and the actual behavior / psychological symptoms of the person, and learns at least one of the function 20a or the classifier 20b based on the evaluation result. 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 person's behavior / psychological symptoms, and the prediction about the BPSD of the prediction unit 30b and the actual person's behavior. The function 20a or the classifier 20b can be learned by deep learning in relation to the evaluation content of the difference from the psychological symptom and the information related to the data acquired by the data acquisition unit 40. The learning unit 30c may adjust the coefficient by deep learning.

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

The BPSD determination unit 30e determines whether the BPSD is based on the input data.

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

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 can be stored.

The correlation table 20c can be composed of a database configured by associating environmental data and biometric data with information related to BPSD corresponding to the environmental data and biometric data. The correspondence table 20d can be made up of a database configured by associating the predicted presence / absence of BPSD and the timing of the occurrence with the coping method. The data storage unit 20e is the data acquired by the data acquisition unit 40, the data obtained by processing the data by the processing unit 30, the data input by the input unit 52, and is previously referenced for various determinations by the processing unit 30. Data to be stored is stored.

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

(A) Five sense data As shown in FIG. 5, for example, the five sense data detects sound source detection, sound type identification, olfactory recognition, face authentication, motion detection, lateral distance function, stereoscopic detection, and the like. The five senses data can be measured with a camera or a microphone. In addition, if it is a device which can acquire such data, it will not specifically limit.

(B) Environmental data As shown in FIG. 6, for example, the environmental data measures temperature, humidity, illuminance, amount of water, and atmospheric pressure, and detects a moving distance. The environmental data can be measured from a temperature sensor, a humidity sensor, an illuminance sensor, an atmospheric pressure sensor, or the like. In addition, if it is a device which can acquire such data, it will not specifically limit.

(C) Biometric data The biometric data detects, for example, heart rate, respiration, getting out of bed, sleep (non-REM sleep, detection of REM sleep), awakening, excretion, momentum, body temperature, etc., as shown in FIG. The sensor for detecting the biometric data is not particularly limited as long as it can perform such detection, and can be configured by one or a plurality of sensors. Taking sleep as an example, a non-REM sleep state and a REM sleep state can be grasped by a Doppler sensor. About excretion, detect opening / closing sensor of toilet door and human sensor in toilet, or detect excretion timing by analyzing communication function tag (BLE tag) and camera video Can do. Regarding 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 burned, calculated from the distance traveled by a tag (BLE tag) that has a communication function, or from the caregiver's record, such as exercise, cleaning, washing, etc. You can calculate the amount of exercise and calorie consumption. In addition, if it is a device which can acquire such data, it will not specifically limit.

(D) Care Record The items of the care record include, for example, as shown in FIG. 8, the needs of the person receiving the care or the caregiver, the concern of the person receiving the care or the caregiver, the subjective symptoms of the person receiving the care, The caregiver's observation, the condition evaluation of the person receiving the care, the impression of the person receiving the care or the caregiver, the caregiver's response, the call of the caregiver, what the caregiver did, etc. can be mentioned.

In the care record, in order to improve the prediction accuracy of BPSD, it is preferable to have the following record.
1) The occurrence of BPSD 2) The time (time zone) at which the BPSD occurred and the location (the role of the location, the positional relationship of the location, etc.) are clearly recorded 3) Environmental factors at the time of occurrence of the BPSD (Temperature, humidity, illuminance, barometric pressure, noise, off-flavor, presence of person, content of conversation, behavior and attitude of subject) 4) The behavior and attitude of the subject before and after the occurrence of BPSD Be recorded. In addition, the environmental factors must be recorded.

Combining these, the role of the occurrence of BPSD, the role of the place, the distance in 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, the relationship with people before and after that From this point of view, it is possible to analogize a situation where the possibility of occurrence of BPSD is high with high accuracy.

(E) Data acquisition method In the data which an acquisition part acquires, you may acquire as follows, for example.
1) Periodic acquisition of environmental data (temperature, humidity, barometric pressure, illuminance) 2) Periodic acquisition of biometric data (heart rate, respiratory rate, sleep state, getting out of bed) 3) Location of target user by BLE tag Acquire information and calculate travel distance 4) Acquire the target user's location information in conjunction with the AI camera and calculate travel distance 5) Link the above acquired data with the user ID and send it to the information processing device periodically Send

1) In the above periodic acquisition of temperature and humidity, the temperature / humidity sensor can transmit the acquisition data at a cycle of 10 minutes from real time by default. Each gateway can acquire the data of the target sensor in association with the user ID. A single gateway can acquire multiple sensor data and support multiple users simultaneously, enabling analysis.

In the above 2) periodic acquisition of atmospheric pressure and illuminance, an expansion module equipped with an atmospheric pressure sensor and an illuminance sensor can be connected to the IoT gateway, and information on atmospheric pressure and illuminance can be updated in a cycle of 10 minutes from real time. A single gateway can support multiple users' data and can support each analysis.

3) In the periodic acquisition of heart rate, respiratory rate, sleeping state, and bed leaving state, a Doppler sensor unit is installed on the wall around the bed, and the heart rate, breathing rate, sleeping state and bed leaving / landing when landing Time event notification can be performed. Heart rate, respiratory rate, and sleep state are measured every 10 minutes from the real time and can be analyzed by notifying the IoT gateway. The bed leaving / landing state may be notified immediately when an event occurs.

(2) Physical configuration of information processing apparatus The information processing apparatus 10 includes an electronic computer. For example, a large electronic computer such as a supercomputer, an electronic computer equipped with a GPU (Graphics Processing Unit), or a plurality of personal computers. It consists of a quantum computer.

The processing unit 30 can be realized by an arithmetic 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.). The storage unit 20 may be configured separately from the arithmetic device or may be configured in the arithmetic device. The information processing apparatus 10 may be composed of one electronic computer or a plurality of electronic computers.

The program that causes the information processing apparatus 10 to execute information processing can be stored in a storage device (for example, ROM, hard disk) included in the information processing apparatus 10.

The data acquisition unit 40 can be constituted by, for example, a reception unit that can receive data from a communication network.

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

3. Information Processing Method (1) Outline of Information Processing Method The information processing method will be described with reference to FIG.

The data acquisition unit 40 acquires, from an IoT device or the like, five sense data, environmental data, biological data, behavior data, and care records for a person receiving care (S1).

The data obtained by the indexing unit 30a is indexed and an evaluation index parameter is calculated (S2).

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

The coping method deriving unit 30d generates a coping method based on the occurrence and timing of occurrence of BPSD (S4).

The data acquisition unit 40 acquires behavior data and the like, and the BPSD determination unit 30e evaluates whether BPSD symptoms actually occur (S5).

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

Fig. 11 shows the measurement data and processing data obtained at each stage. Each data of FIG. 11 can be stored in the data storage unit 20e of the storage unit 20 in association with each other.

(2) Example of indexing process (a) Environmental index For the environmental index, the degree of discomfort calculated for each temperature + humidity, atmospheric pressure, and illuminance can be used as an index. First, regarding temperature + humidity, a graph indicating a known comfort range is used to determine whether it is uncomfortable depending on whether it is within the comfort range.

As a process for this, it is first determined whether or not the humidity is between 40% and 70%.

If the humidity is between 40% and 70%, it is determined whether the temperature is appropriate using the following formula.

It is assumed that when the temperature and humidity are within the range of the mathematical formula, it is comfortable, and when it deviates from this range, it is uncomfortable. When it is determined that it is uncomfortable, the distance from the intersection of the current temperature / humidity line and the comfort range is simply used as an index of the discomfort degree from the center point of the comfort range.

Next, for the barometric discomfort index, the following formula is applied with the standard pressure 1013 hPa as a standard value.

Here, when the atmospheric pressure is P, the discomfort index is P ′.

Next, although it is a discomfort index of illuminance, this is a reference illuminance that varies depending on the state of the user. The following formula is used with the reference illuminance as I ₀ .

As the illuminance I, I ′, which is a discomfort index of illuminance, is calculated. The reference illuminance is calculated as 200 lux during activity, 50 lux during rest, and 20 lux during sleep.

(B) Biometric index As for the heart rate, the biometric index P ′ is obtained by the following calculation formula, where P is a heart rate per minute and the average heart rate is P ₀ .

As for the respiration rate, the biometric index B ′ is obtained by the following calculation formula, where B is the respiration rate per minute and B ₀ is the average respiration rate.

(3) Prediction process (a) Prediction logic When the prediction logic is P, matrix calculation can be performed based on the following formula to calculate P.

Here, p1 to p5 represent parameters, and α to ε ′ represent the degree of items related to the behavioral / psychological symptoms (BPSD) of dementia.

(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 times are listed as candidates as reference points.
1) 0:00 (date change point)
2) Hiji (sunset)
3) Waking up 4) Meals

∙ With respect to the occurrence of BPSD, the elapsed time from each reference point can be calculated, and the correlation can be derived from the deviation. Regression analysis can be performed on the timing of occurrence of BPSD and the elapsed time from the reference point to derive the relationship. At this time, wake-up may be classified according to whether it is nighttime sleep or nap, and meals may be classified according to types such as breakfast, lunch, dinner, and snacks. If the time when BPSD occurs is point A, 0:00, wake up at night, the elapsed time of breakfast / lunch / dinner close, and if it is after a nap, the elapsed time from a nap, if after a snack The elapsed time from snacking may be used as an additional indicator.

新しい When a new event occurs, there is a tendency that a memory failure that forgets the event just before that increases. When you become depressed and you don't know what to do, you tend to do the habit of wandering unintentionally, doing the same act of repeating the same action as searching for something. There is a tendency to make angry voices that are irritating and annoying because of excessive involvement by directing instructions. These flows have unique time differences, and can be predicted in consideration of time series.

In the prediction of BPSD, unlike an image analysis such as a photograph, it is preferable to use a function 20a or a classifier 20b that takes into account the influence of time series. The learning of the function 20a or the classifier 20b is similar to language / audio / video analysis in deep learning (DeepLarning), so it is preferable to learn using LSTM (Longshort-termmemory).

(C) Prediction from the environmental factors of BPSD Since the influence of the surrounding environment on the occurrence of BPSD is large, it is possible to analyze the environmental factors at the time of BPSD occurrence, especially temperature, humidity, atmospheric pressure, illuminance, smell, and noise. it can. As for the temperature and humidity, the degree of comfort can be diagnosed based on the comfort index graph for the relationship between temperature and humidity and can be used as an index. For atmospheric pressure, illuminance, odor, and noise, the influence and degree of each index on BPSD can be calculated by analyzing the relationship with the occurrence of BPSD as a single index and performing regression analysis.

(D) Prediction from natural language analysis BPSD can also be analyzed by natural language analysis. This is a method in which a part of speech included in a question or inquiry is identified by natural language processing, a hypothesis is generated, and then the hypothesis is supported or searched for evidence. A response method is assigned based on the onset status of BPSD according to the statistical model method of the evidence weight score.

(E) Prediction from behavior As shown in FIG. 10, the correlation of BPSD is known from the behavior of a dementia patient. Thereby, the onset of BPSD can be predicted from the behavior of the dementia patient.

(4) Coping method Based on the BPSD symptoms, a coping method can be generated based on the function 20a or the classifier 20b. Further, a BPSD coping method may be selected based on the correspondence table 20d between BPSD symptoms and coping methods. A correspondence table 20 d as shown in FIG. 12 may be stored in the storage unit 20.

In the coping method, when it is necessary to create the menu contents of a meal, the nutritional calories may be calculated based on the amount of exercise and the calorie consumption and used for the menu.

(5) Verification (a) Evaluation of Occurrence Verification of the onset of BPSD can be performed by a BPSD problem behavior evaluation scale (TBS: Troublesome Behavior Scale). TBS describes the destructive behavior and burden of elderly people with dementia.

The TBS defines fifteen items describing the destructive behavior and burden behavior of elderly dementia patients and their frequency, and the applicant has reliability and validity as a measure for assessing the transfer of behavior of dementia patients. I confirmed that there was. In addition, problem behaviors that are observed relatively well by people with dementia are evaluated (for example, a five-step evaluation) based on the frequency that the caregiver observed in the past predetermined period (for example, the past month), and the prediction results are verified. can do. The frequency can be classified into “at least once a day”, “several times a week”, “several times a month”, and “none”.

BPSD analysis can be performed by voice pathological analysis. Using emotion recognition technology (Sensibility Technology), it is possible to analyze the ratio of the four emotions of “anger”, “joy”, “sadness”, and “normal” during utterance and the degree of “excitement” and display the degree. Verification of the onset of BPSD may be performed by brain image diagnosis.

As shown in FIG. 13, in the evaluation of whether or not BPSD has occurred, and when it has occurred, the content and timing of its onset, the data acquired by the data acquisition unit 40 as to whether or not BPSD has occurred in a demented patient or Based on the data input by the input unit 52, the BPSD determination unit 30e can make a determination. This BPSD determination algorithm can be learned by the learning unit. The learning of 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 indexing unit 30a indexes the data, determines whether the BPSD determination unit 30e is BPSD (F13), and determines whether the data is BPSD. Is output. Learning of the judgment algorithm used when the input actual BPSD status and the result determined by the BPSD determination unit 30e are verified (F15), and the learning unit 30c determines whether the BPSD determination unit 30e is BPSD. (F16).
(B) Evaluation of coping method Through the evaluation of coping with BPSD, the degree of reliability is evaluated according to how much evaluation the derivation of the care method has acquired. By running analytics on the massive amount of data collected from care sites, collecting insights and converting them into inspiration can help derive appropriate care methods. In the evaluation of the BPSD coping method, whether the dementia patient has developed BPSD or not is verified by the processing unit 30 based on the data acquired by the data acquiring unit 40 or the data input by the input unit 52. May be. Based on the verification result, the learning unit 30d can learn the derivation algorithm of the coping method derivation unit 30d and update the derivation algorithm. 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.

3. Effects The occurrence of BPSD or the timing of the onset is predicted, and early treatment can prevent the onset of BPSD in advance and greatly reduce the occurrence of BPSD itself. Since BPSD can be predicted in advance, the burden on the caregiver can be reduced.

The above embodiment can be variously modified within the scope of the gist of the present invention. In the above embodiment, the behavior / psychological symptoms of human dementia have been described. However, the present invention is not limited to this and can be widely applied to behaviors / psychological symptoms of animal dementia.

4). Application Example (1) AI Artificial Intelligence System The information processing apparatus according to the present embodiment can be applied to the following AI (Artificial Intelligence) artificial intelligence system. In other words, the AI artificial intelligence system is an information processing device formed by a data storage unit, an analysis unit, a sensitivity processing unit, and a planning unit based on a knowledge expression method, and data collected by automatic identification, automatic handling, and automatic notification from an IoT gateway. Furthermore, the care record data collected by human interface, natural language analysis, and speech recognition can be merged, and further, a data mining unit can be configured by statistical analysis.

Based on knowledge data that can be used for care, it is possible to continuously grasp the state of elderly people with dementia and predict the onset of specific BPSD from changes in the state, suggesting an appropriate pre-stored response method . Has the following effects.
(A) Focus not only on caregivers but also on people with dementia (b) Multidimensionally and flexibly to meet the needs of caregivers and people with dementia (c) In the case of drug therapy indications (D) Information sharing can be analyzed in correlation with data obtained from IoT information while paying attention to the following information (a) to (c).
1) Information to deal with specific BPSD 2) Information to ensure physical safety and well-being of people with dementia 3) Information to deal with difficult ADL (daily life movement) information

基礎 Basic research on AI can be applied to responding to behavioral and psychological symptoms (BPSD) of dementia based on inference and learning.

(2) Functions of AI Artificial Intelligence System The AI artificial intelligence system can have the following functions.
(A) Expert system A system that accumulates expert knowledge as rules and solves problems using inference techniques.
(B) Voice recognition Speak to a smartphone or tablet. The computer understands what was spoken and understood what was spoken, and it is written.
(C) Natural language processing The meaning and content of the documented information is sorted and recorded so that the computer can understand the information and the information can be retrieved by the life support recording method of F-SOAIP.
(D) Sensitivity processing Based on knowledge of cognitive science and ergonomics, a feeling that the feeling is warm or cold is received from the environmental sensor and realized on the computer.
(E) Image recognition Lets the computer understand the contents taken by the camera, etc., and sorts the brightness and color tone of the living room pleasantly and unpleasantly.
(F) Machine learning A system that finds consistent rules (patterns) from data collected by IoT sensors and nursing care records. It is strongly related to the field of mathematics statistics. For example, it is analyzed and arranged using the following statistical methods.

1) FTA (Fault Tree Analysis) analysis This is an analysis method that searches for the cause from the results, and analyzes the path, cause, and probability of occurrence of an undesirable event. In order to analyze the occurrence frequency of BPSD, the potential danger (fault) of the cause is logically traced (here, “fault” refers to an event such as an environment or a human error). The probability of occurrence is added to calculate the probability that a basic event can occur.
2) ETA (Event tree analysis) analysis It is also called risk prediction analysis, and analyzes the process up to the occurrence of BPSD based on the probability of occurrence and the response of its countermeasure (care) based on success or failure.
3) HAZOP (Hazard and Operability Study) analysis This is an analysis based on empirical rules, and in situations where behavioral and psychological symptoms occur, the condition is appropriate for the parameters that represent the behavior and personality characteristics of the subject. It is a method of analyzing whether it was born as a result of the effect of not doing.
The parameters here are analyzed including data such as body temperature, breathing, pulse, temperature, humidity, barometric pressure, and sleep time obtained from the IoT gateway, and a guide word such as whether the event has occurred repeatedly.
4) Information search It is a system to find out what is needed for dementia care from the accumulated data.

(G) Inference This is a system for deriving consistent answers by integrating laws from various patterns.
1) Hot Spot Analysis
This is an analysis method that regards a space (place) where a past behavior / psychological symptom has occurred as a space (place) where the possibility of occurrence of a behavior / psychological symptom is high.
2) Regression methods
In addition to the past BPSD, other variables related to the BPSD such as environment and human relations are used as independent variables, and the future BPSD is predicted by regression analysis.
3) Near-Repeat Methods
A future BPSD is predicted based on the spatio-temporal proximity of one BPSD and the next BPSD.
4) Spatiotemporal Analysis
The behavioral / psychological symptoms are predicted from the movement pattern accompanying the time change of the occurrence of BPSD and various factors affecting it.
5) Risk Terrain Analysis
A risk aspect is created based on the proximity to the spatial factors that affect BPSD, and the future occurrence of BPSD is predicted.

(H) Data mining A system that links database technology and machine learning to find useful information from a large amount of unorganized data. Predicts and responds to BPSD by data mining techniques such as classification and clustering. Is derived.
(I) Human interface Smartphones and tablet devices that allow caregivers to more easily operate devices such as computers can be applied.
(J) Planning A system for determining the order in which care should be performed can be applied in deriving an appropriate care response of BPSD.
(K) Multi-agent This is a system that proposes a care method by gathering caregivers who solve BPSD and examining the occurrence of BPSD again with F-SOAIP when information on complicated problems is solved at the care site.

して Utilizing the above contents, it can be applied to support BPSD for dementia based on inference and learning.

(3) Care methods Dementia requires care methods tailored to different behavioral and psychological symptoms. This is because some care methods have different effects on patients with dementia. The information processing apparatus is capable of accumulating various documents and past information covering information on standard care methods and the best interpersonal assistance methods. All of these can identify which care methods provide the best options for caregivers to take in caring for patients with dementia. Under the guidance of care professionals, the IoT service for dementia may collect the knowledge necessary to acquire the skills in the care setting. This is defined as “a corpus of knowledge in dementia care”. The creation of a corpus can be started by loading many relevant documents into a dementia-aware AI.

Also, the creation of the corpus can involve the intervention of professional staff to select information or to eliminate all old, inferior, and information unrelated to the problem area. This is defined as “curation of dementia-compatible content”.

Indexes and other metadata that are pre-processed by curation and can be linked more efficiently with content can be built from the field using the Life Support Recording Method (F-SOAIP). A dementia-aware AI trained with question-answer pairs can continue to learn by continuously interacting with the bot. In addition, when new information is released, the AI for dementia is also updated, constantly adapting to changes in knowledge and linguistic interpretation in a given field, and identifying new insights and patterns hidden in the information Can be ready to do. It is a method in which a part of speech included in a question or inquiry is identified by natural language processing, a hypothesis is generated, and then the hypothesis is supported or searched for evidence.

対応 According to the statistical model method of evidence weighted score, a response method can be assigned based on the onset state of behavior / psychological symptoms. Dementia-responsive AI can be evaluated for its reliability by the degree of evaluation that the derivation of care methods has gained through successful cases of behavioral and psychological symptoms. In short, AI for dementia enables analytics to be performed on large amounts of data collected from nursing care sites, collecting insights and converting them into inspiration, so that appropriate care methods can always be derived.

The above embodiment can be variously modified 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 and animals suffering from dementia.

DESCRIPTION OF SYMBOLS 10 Information processing apparatus 20 Storage part 20a Function 20b Classifier 20c Correlation table 20d Correspondence table 20e Data storage part 30 Processing part 30a Indexing part

30b Prediction part

30c Learning part 30d Countermeasure generation part 30e BPSD judgment part 40 Data acquisition part 52 Input Unit 54 display unit 56 transmission unit 70 communication network

Claims

At least selected from the group of environmental data around a person or animal, biological 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 for acquiring one type;
An information processing apparatus, comprising: a prediction unit that predicts the onset or the onset time of behavior / psychological symptoms in dementia of the person or animal based on the data acquired by the data acquisition unit.
In claim 1,
The information processing apparatus that predicts behavior / psychological symptoms in dementia of the human or animal using at least one of inference analysis, regression analysis, HotSpot analysis, proximity analysis, and spatiotemporal analysis.
In claim 1 or 2,
Including an indexing unit for indexing the data acquired by the data acquisition unit,
The prediction unit receives the index value indexed by the indexing unit and performs prediction using a function or a classifier that outputs a prediction value related to behavioral / psychological symptoms in dementia of the person or animal Information processing device.
In claim 1 or 2,
The classifier is an information processing apparatus that is an SVM (Support Vector Machine), a neural network, or a linear regression model.
In claim 1 or 2,
The prediction unit evaluates the difference between the prediction of the behavior or psychological symptoms of the person or animal dementia and the actual behavior or psychological symptoms of the person or animal, and learns the function or classifier based on the evaluation result An information processing apparatus including a learning unit for performing.
In claim 5,
The learning unit evaluates the difference between the prediction of behavior / psychological symptoms in the human or animal dementia predicted by the prediction unit and the actual behavior / psychological symptoms of the human or animal,
Prediction regarding behavior / psychological symptoms in dementia of the person or animal of the prediction unit, evaluation contents of difference between actual behavior / psychological symptoms of the person or animal, information on data acquired by the data acquisition unit, and An information processing apparatus that learns the function or classifier by deep learning in relation to the above.
In claim 5,
The function includes a parameter related to a numerical value related to data acquired by the data acquisition unit, and a coefficient,
The learning unit is an information processing apparatus that adjusts the coefficient by deep learning.
In claim 1, 2, 6 or 7,
The data acquisition unit acquires the environmental data and the biological data,
The environmental data includes an ambient temperature and humidity around the person or animal, and the biological data includes an information processing apparatus including a pulse, respiratory rate, and body temperature of the person or animal.
In claim 8,
An information processing apparatus comprising: a data storage unit in which the environmental data and the biological data, and information related to behavioral and psychological symptoms in dementia of the person or animal corresponding to the environmental data and the biological data are stored in association with each other.
In claim 8,
A storage unit storing a correlation table configured by associating the environmental data and the biological data with information related to behavioral and psychological symptoms in dementia of the person or animal corresponding to the environmental data and the biological data; Including information processing apparatus.
In claim 1, 2, 6, 7, 9 or 10,
An information processing apparatus including a coping method derivation unit that generates or selects a coping method based on the prediction content related to the behavioral / psychological symptoms in dementia of the human or animal predicted by the prediction unit.
In claim 5,
The actual behavior / psychological symptoms of the person or animal are determined according to a determination algorithm by the behavior / psychological symptom determination unit,
The learning unit evaluates a difference between the determination result of the behavior / psychological symptom determination unit and the determination result of the behavior / psychological symptom of the person or animal actually determined by the person, and the determination algorithm based on the evaluation result Information processing device that learns.
A data acquisition unit includes a group of environmental data around a person or animal, biological 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. Obtaining at least one selected from:
A prediction unit predicting the onset or the onset time of the behavioral / psychological symptoms in dementia of the person or animal based on the data acquired by the data acquisition unit.
In the computer, the data acquisition unit includes environmental data around the person or animal, biological 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. Obtaining at least one selected from the group of:
A program for causing the prediction unit to execute the step of predicting the onset or the onset time of behavioral / psychological symptoms in dementia of the person or animal based on the data acquired by the data acquisition unit.