TWI744512B - Computer program, terminal, prediction method and server - Google Patents

Abstract

Improve the prediction of symptoms or severe illnesses based on meteorological information The accuracy of the risk of change.

The computer program is used to make the terminal realize the following functions : The function of accepting the input of the user’s attributes; and notifying the user: the model selected from the multiple models representing the relationship between disease and weather and corresponding to the input attributes, and weather forecast information The function of the obtained predictive information about the disease.

Description

Computer program, terminal, prediction method and server

The present invention relates to computer programs, terminals, methods and servers.

Regarding the relationship between health and weather, if the low air pressure approaches, head or feet will be painful; hay fever will be more serious on sunny days after rain; heatstroke will increase on sunny days; if dry, the skin will be rougher. It has been known as the wisdom of life in the past. In addition, it is known to use a discrimination algorithm that combines the factors of subjective information such as specific papers or questionnaires, and biological information/environmental information obtained in restricted places such as meteorological laboratories to target the general public. , To provide averaged information.

For example, Non-Patent Document 1 provides pollen information indicating the state of pollen scattering for each region. The user can estimate the risk of symptomatic or severe hay fever by referring to the pollen information.

[Prior Technical Literature] [Non-Patent Literature]

[Non-Patent Document 1] https: //tenki.jp/pollen/, May 18, 2017 Day search

However, for people who do not have hay fever, the importance of pollen information is low. In addition, people with hay fever are also different in their susceptibility to pollen. Even if the amount of flying is the same, there are people who will become severely ill and those who will not. This is the same for other diseases, not limited to hay fever.

The present invention has been completed in view of such a problem, and its purpose is to provide a technology that can predict the risk of symptoms or aggravation of the disease according to the attributes of the user such as gender, age, medication status, etc., by providing meteorological information, thereby improving the accuracy of the prediction .

One aspect of the present invention relates to computer programs. The computer program is used to enable the terminal to realize the following functions: the function of accepting the input of the user's attributes; and the notification to the user: corresponding to the input attributes from the plural models representing the relationship between disease and weather The selected model, and the function of predicting information about the disease obtained from the information of weather forecast.

Another aspect of the present invention is a server. The server is equipped with: a holding means for holding different plural models representing the relationship between disease and weather; a receiver that receives the attributes of the user through the network through the terminal Segment; from among the plural models held in the holding means, the selection means for selecting the model corresponding to the acquired attributes; the means for obtaining weather forecast information from the weather server through the network; the selected model , And the obtained meteorological forecast information, the derivation means of deriving the forecast information about the disease; and the sending means of sending the derived forecast information about the disease through the network to the terminal.

Among them, any combination of the above constituent elements, or the constituent elements or expressions of the present invention, which are interchangeable among the devices, methods, systems, computer programs, recording media storing the computer programs, etc., are also effective as the present invention. State.

With the present invention, it is possible to improve the accuracy of predicting the occurrence of symptoms or the risk of severe disease based on meteorological information.

100: Disease prediction system

102: Disease prediction server

104: Medical Data Server

106: Weather Data Server

108: portable terminal

110: Network

112: Medical Big Data

114: Weather Big Data

121: memory

122: processor

123: Communication interface

124: display

125: input interface

126: Bus

302: Data Acquisition Department

304: Data Analysis Department

306: Model Holding Department

308: Property reception department

310: Model Selection Department

312: Forecast Acquisition Department

314: Risk Derivation Department

316: Communication Department

318: User Information Retention Department

320: Performance Information Maintenance Department

400: Initial login screen

402: Initial login information input area

404: Send button

406: login screen

408: ID area

410: PW area

412: Send button

414: Risk prediction screen

416: Disease Selection Area

418: Area display area

420: Date display area

422: Risk level display area

424: Move display area

426: Setting button

428: Comment display area

430: Temperature Chart

432: Weather display area

434: Medium Risk Area

436: Low Risk Area

438: Date and time display area

440: Information registration screen

442: Health status login area

444: Voice login area

446: Examination history registration area

448: Medication Record Area

450: Memo area

452: Login button

454: Risk Forecast Screen

456: Date display area

458: Setting button

460: High-risk area

462: Move display area

464: Risk level display area

466: Risk Forecast Screen

468: Date display area

470: Move display area

472: Setting button

474: Login data area

476: Risk level display area

Fig. 1 is a schematic diagram showing an example of the system configuration of the disease prediction system of the embodiment.

Fig. 2 is a hardware configuration diagram of the portable terminal of Fig. 1.

Fig. 3 is a block diagram showing the function and structure of the disease prediction server of Fig. 1.

FIG. 4 is a data structure diagram showing an example of medical input data generated by the data acquisition unit in FIG. 3.

Fig. 5 is a data structure diagram showing an example of weather input data generated by the data acquisition unit in Fig. 3.

Fig. 6 is a graph for explaining an example of multiple regression analysis in the data analysis unit of Fig. 3.

Fig. 7 is a data structure diagram showing an example of the model holding section of Fig. 3.

Figure 8 (a) ~ (d) are schematic diagrams showing display examples when the risk index is classified into 5 risk levels.

Fig. 9 is a flowchart showing the processing flow in the disease prediction system of Fig. 1.

Figure 10 is a representative screen diagram of the initial login screen.

Figure 11 is a representative screen diagram of the login screen.

Fig. 12 is a representative screen diagram of the risk prediction screen showing the current risk level.

Fig. 13 is a representative screen diagram of the information registration screen.

FIG. 14 is a representative screen diagram of the risk prediction screen showing the future risk level.

Fig. 15 is a representative screen diagram of a risk prediction screen showing past risk levels.

Hereinafter, the same or equivalent constituent elements, members, and processes shown in the respective drawings are denoted by the same reference numerals, and repeated descriptions are appropriately omitted. In addition, a part of the members that are not important in the description in each figure is omitted from the display.

In the implementation form, new value is created by multiplying medical data and weather data. By comparing medical data and meteorological data, hypotheses about the relationship between disease and weather can be verified, and a highly accurate model can be developed. The model system developed as shown above can provide new services or value in various fields such as media, manufacturers, finance, or autonomous organizations. For example, by using the developed model, it is possible to see the causal relationship between weather and disease (illness), and to predict the occurrence area or period of the disease.

The implementation form supports the improvement of health literacy through services using the proposed model, and promotes the distortion of action, thereby contributing to the realization of a healthy society. In addition, the implementation mode promotes research on the impact of weather or the environment on living organisms through services using the proposed model, and contributes to the realization of a society that is in harmony with the natural world.

FIG. 1 is a schematic diagram showing an example of the system configuration of the disease prediction system 100 of the embodiment. The disease prediction system 100 includes a disease prediction server 102, a medical data server 104, a weather data server 106, and a portable terminal 108. The disease prediction server 102, the medical data server 104, the weather data server 106, and the portable terminal 108 are all connected to a network 110 such as the Internet, and communicate through the network 110. Among them, Figure 1 shows an example, and the number of elements is not limited.

The medical data server 104 is operated by a company that analyzes medical-related data (hereinafter referred to as a medical data analysis company), and holds medical big data 112. The user is accepted at a medical institution on a regular basis, such as once a year Health checkup (regular health checkup, comprehensive health checkup, etc.). The user receives a health diagnosis at a medical institution cooperating with the health insurance association or other insurance institution he has joined. After receiving the health diagnosis, the user receives the result of the health diagnosis in writing or electronically (hereinafter referred to as the health examination result). Each medical institution informs the user of the result of the health check to the insurer or the user’s business owner that the user has joined. Among them, in addition to the health check results, the prescription data generated when the user receives general medical treatment in a medical institution is also provided to the insurer of the health insurance that the user has joined. As shown above, the insurer or business owner accumulates the results of health examinations or prescriptions of their insureds or practitioners.

Medical data analysis companies can sign contracts with most insurers or business owners, and the insurers or business owners after signing the contract can obtain the results of health examinations or prescriptions and anonymize them to construct medical big data 112. This intake may be performed collectively, for example, on a regular basis (e.g., monthly). Medical data analysis companies can also directly obtain health check results or prescription data from users through portable terminals 108 such as smartphones, tablet terminals, or mobile phones.

The weather data server 106 is operated by a company that analyzes weather-related data (hereinafter referred to as a weather data analysis company), and maintains big weather data 114. The meteorological data analysis company conducts survey analysis or information provision on weather, environment, disaster prevention, etc. The meteorological data analysis company collects wind direction, wind speed, temperature, daily range, humidity, solar radiation, radiation budget, and rainfall for each region (e.g. 250m mesh unit) and time period (e.g. minute unit). Wait for meteorological data, construct meteorological big data 114. The weather data server 106 is based on the weather big data 114 The past weather data is used to generate weather forecast information according to a predetermined forecast algorithm. The information of meteorological prediction includes: for example, the predicted value of wind direction, wind speed, temperature, daily difference, average annual difference, or humidity in a certain future time period at a certain place.

The disease prediction server 102 obtains the medical data held in the big medical data 112 from the medical data server 104, and obtains the meteorological data held in the big weather data 114 from the weather data server 106. The disease prediction server 102 compares the acquired medical data and weather data to generate a different complex model representing the relationship between the disease and the weather. Models vary according to the attributes of users. For example, the model for men in their 10s is different from the model for women in their 40s. The generation of the model will be described later.

The disease prediction server 102 obtains the attributes of the user of the portable terminal 108 through the network 110, and selects a model corresponding to the obtained attributes. The disease prediction server 102 obtains the weather prediction information from the weather data server 106 through the network 110, and derives the disease prediction information from the obtained weather prediction information and the selected model. The disease prediction server 102 transmits the disease prediction information to the portable terminal 108 via the network 110.

The user of the portable terminal 108 downloads and installs the disease prediction application program (hereinafter referred to as the disease prediction APP program) on the portable terminal 108 via the network 110 from the download website. Or in addition, the disease prediction APP program can also be pre-installed on the portable terminal 108. The disease prediction APP program is based on the target level specified by disease/region/age/sex, etc. Distribution becomes the valuable information at the beginning of action distortion. The disease prediction APP program can also be provided by any one of a medical data analysis company, a weather data analysis company, or a management body of the disease prediction server 102. The portable terminal 108 executes the disease prediction APP program, whereby the portable terminal 108 communicates with the disease prediction server 102 through the network 110 and realizes various functions. Hereinafter, the functions implemented by the portable terminal 108 (CPU (Central Processing Unit, central processing unit) and other processing units) executing the disease prediction APP program as the function of the portable terminal 108 will be described. These functions are actually implemented by the portable terminal 108 by the disease prediction APP program.

Fig. 2 is a hardware configuration diagram of the portable terminal 108 of Fig. 1. If the portable terminal 108 can be executed by installing a disease prediction APP program, it can also be any terminal. The portable terminal 108 includes a memory 121, a processor 122, a communication interface 123, a display 124, and an input interface 125. These elements are respectively connected to the bus 126, and communicate with each other through the bus 126.

The memory 121 is a memory area for storing data or programs. Data or programs can be permanently stored in the memory 121 or temporarily. In particular, the memory 121 is an APP program for predicting memory diseases. The processor 122 implements various functions in the portable terminal 108 by executing programs stored in the memory 121, especially the disease prediction APP program. The communication interface 123 is an interface for transmitting and receiving data with the outside of the portable terminal 108. For example, the communication interface 123 includes: an interface used to access the wireless communication network of a mobile phone, or an interface used to LAN (Local Area Network, local area network) access interface, etc. In addition, the communication interface 123 may also include a wired network interface such as USB (Universal Serial Bus). The display 124 is a device used to display various information, such as a liquid crystal display or an organic EL (Electroluminescence, electroluminescence) display. The input interface 125 is a component for receiving input from the user. The input interface 125 includes, for example, a touch panel provided on the display 124, or various input keys.

FIG. 3 is a block diagram showing the function and structure of the disease prediction server 102 in FIG. 1. The blocks shown here can be realized by components or mechanical devices led by a computer’s CPU in hardware, and can be realized in software by computer programs, etc. However, it is depicted here by The functional blocks realized by the cooperation of others. Therefore, these functional blocks can be realized in various forms by a combination of hardware and software, which can be understood by those who are familiar with the technology in the field related to this specification.

The disease prediction server 102 includes: a data acquisition unit 302, a data analysis unit 304, a model holding unit 306, an attribute receiving unit 308, a model selection unit 310, a prediction acquisition unit 312, a risk derivation unit 314, a sending unit 316, The user information holding unit 318 and the actual performance information holding unit 320.

(Model generation state (phase))

When generating the model, the data obtaining unit 302 obtains the prescription data from the medical data server 104 through the network 110, or directly obtains the prescription data from the medical data server 104 without going through the network 110. Package of prescription information obtained Including: medical prescriptions, and dispensing prescriptions. The items of the prescription information include: patient information such as age or gender, prescription information such as the year and month of diagnosis and treatment, or the actual number of days of treatment or total points (medical expenses), the name of the injury or illness, or the date and month of the start of diagnosis and treatment or the transition period, and other injury and illness information, and medicine. Drug information such as product name or dosage, and medical treatment information such as the date of implementation or medical treatment (surgery, radiotherapy, etc.).

The data obtaining unit 302 obtains the general book data from the medical data server 104 through the network 110, or directly obtains the data from the medical data server 104 without going through the network 110. The data of the total booklet includes: my family (blood relationship), the start year and month of the observation, the end year and month of the observation, and the reason for the end. In this way, the overall observation can be set, and the matrix becomes clearer.

The data obtaining unit 302 obtains the regional data from the medical data server 104 through the network 110, or directly obtains it from the medical data server 104 without going through the network 110. The area data includes: the location of the medical institution where the patient was diagnosed (hence the prescription has been made), and the patient's place of residence. If an individual cannot be identified, the first three digits of the postal code are used.

The data acquisition unit 302 systematically combines the prescription data, the general book data, and the area data obtained by the system to generate medical input data. The medical input data indicates when, where, and what kind of disease the user has. The items of medical input data include, for example, age, gender, region, year, month, day, illness, severity, hospitalization, surgery, frequency of consultation, department of consultation, medical expenses, and prescription drugs.

Fig. 4 shows the medical treatment generated by the data acquisition unit 302 The data structure diagram of an example of the input data. The medical input data corresponds to the area ID of the specific area, the patient's age (year), the patient's gender, the patient's illness, the date of diagnosis, the severity of the illness, the presence or absence of the onset, the presence or absence of hospitalization, the presence or absence of surgery, and the number of patients to maintain. The area ID may be, for example, a prefectural code or a national local public organization code prescribed in JIS X 0401. Severe severity is represented by five stages such as 1-5, with 5 being the most severely injured. The severity of the disease is determined based on, for example, the type and amount of drugs prescribed or used. For example, the data in the top row of Figure 4 represents severe degree 4 atopic dermatitis that requires hospitalization and surgery. There were 5 female patients aged 0 to 4 who were diagnosed in District 13 on April 1.

Returning to FIG. 3, the data acquisition unit 302 acquires weather data from the weather data server 106 through the network 110. The items of the acquired weather data include, for example, the following.

Meteorological related items: cloud cover, precipitation, wind direction, wind speed, temperature, humidity, sunshine time, snow depth, snowfall, air pressure.

Environmental related items: SO ₂ (sulfur dioxide), NO (nitrogen monoxide), NO ₂ (nitrogen dioxide), NO _X (nitrogen oxide), CO (carbon monoxide), O _X (photochemical oxidant), NMHC (non-methane) Hydrocarbons), CH ₄ (methane), THC (total hydrocarbons), SPM (plankton particulate matter), PM2.5 (fine particulate matter), SP (plankton dust), yellow sand, cedar pollen, cypress pollen , Volcanic ash, UV (ultraviolet rays).

The area of the acquired weather data is designated by the observation location or prediction area (for example, a 200m grid). The time interval of the acquired weather data is such as minute unit, hour unit, day unit, week unit, month unit. The data acquisition unit 302 generates weather input data by matching the format of the acquired weather data with the format of the medical input data.

FIG. 5 is a data structure diagram showing an example of weather input data generated by the data acquisition unit 302. The weather input data is maintained by matching the area ID, date and time (time period), air pressure, precipitation, air temperature, dew point temperature, vapor pressure, humidity, and wind speed.

Returning to FIG. 3, the data analysis unit 304 system combines the medical input data and weather input data generated by the data acquisition unit 302 to generate analysis target data. The data analysis unit 304 uses, for example, the area ID and time zone (test date, date and time) as the key value (key) at the time of integration. The data analysis unit 304 subdivides the analysis target data according to the attributes and regions of the user. For example, the data analysis unit 304 distinguishes analysis target data based on the user's gender, age, and region. This generates analysis target data for 0-4 year old women in area 1, analysis target data for 0 to 4 year old men in area 1, analysis target data for 5 to 9 year old women in area 1, and information about area 2. Analysis target data for women aged 0 to 4 years, etc.

The data analysis unit 304 is for each of the analysis target data for gender, age, and place, and among the analysis target data, data items derived from medical input data are used as target variables, and data items derived from meteorological input data are used as target variables. Illustrate multiple regression analysis of variables. In this way, models are generated according to gender, age, disease, and region. For example, through multiple regression analysis, determine the reason for the index used to calculate the symptom risk of a certain disease for users of a certain age/gender in a certain area The data item derived from the meteorological input data and the model coefficients accompanying the data item. The data analysis unit 304 registers the information of the generated model in the model holding unit 306.

FIG. 6 is a graph for explaining an example of multiple regression analysis in the data analysis unit 304. In the graph shown in Figure 6, the X-axis, Y-axis, and Z-axis systems respectively represent temperature, average annual difference, and risk index. The risk index indicates the risk of symptoms or attacks of asthma, and is calculated based on the number of asthma patients and severity of asthma based on medical input data. The data points in the graph correspond to the entry of analysis target data for males under 10 years old in Kyushu/Okinawa. The data analysis unit 304 performs multiple regression analysis on the set of data points in the graph shown in FIG. 6 with the risk index of asthma as the objective variable, and temperature and the mean annual difference as the explanatory variable, and determines the model coefficient for each explanatory variable.

In the model, the disease risk index R is calculated by the following formula 1.

R=a×temperature+b×average annual difference...(Equation 1)

Here, a is the model coefficient of the temperature (hereinafter referred to as the temperature coefficient), and b is the model coefficient of the average annual difference (hereinafter referred to as the average annual difference coefficient).

FIG. 7 is a data structure diagram showing an example of the model holding unit 306. The model maintaining unit 306 is aimed at asthma, and maintains the temperature coefficient a and the average annual difference coefficient b by gender/year and by region. Among them, in this example, the temperature and the mean annual difference are used as the explanatory variables, but the result of the analysis is As a result, other items are also considered, such as the situation where the day is worse or the humidity becomes a better explanatory variable.

(Applicable model)

Return to Figure 3 to explain the value provided to users using the model generated as described above. The attribute receiving unit 308 receives the attribute, target disease, and area of the user of the portable terminal 108 from the portable terminal 108 through the network 110. The attributes of the received user include: the age and gender of the user.

The model selection unit 310 selects a model corresponding to the attribute, disease, and region acquired by the attribute receiving unit 308 from among the plural models held in the model holding unit 306. For example, the model selection unit 310 refers to the model holding unit 306 to obtain model coefficients corresponding to the acquired gender, age, disease, and region, that is, the temperature coefficient and the average annual difference coefficient.

The forecast acquisition unit 312 acquires the weather forecast information in the acquired area from the weather data server 106 through the network 110.

The risk derivation unit 314 derives the forecast information about the disease from the model selected by the model selection unit 310 and the weather forecast information acquired by the forecast acquisition unit 312. For example, the risk derivation unit 314 substitutes the obtained temperature coefficient and average annual difference coefficient and the obtained predicted temperature and predicted average annual difference into Equation 1, thereby calculating the risk index of a certain disease. The risk derivation unit 314 classifies the calculated risk index into plural levels (hereinafter referred to as risk levels).

Figure 8(a)~(d) shows that the risk index is classified into 5 winds A schematic diagram of the display example of the risk level. Figure 8(a) shows a situation where the size of the risk index is represented by five texts. Figure 8(b) shows a graph showing the time change of the risk index. Figure 8(c) shows a situation where 5 graphs (in this case, avatars and their numbers) are used to express the size of the risk index. Figure 8(d) shows a distribution diagram showing the geographical distribution of the risk index.

Or in addition, the risk derivation unit 314 may also classify the risk index into three risk levels. At this time, the risk derivation unit 314 uses, for example, two threshold values RH and RL (RH>RL), and classifies the risk index based on the following judgment criteria.

(1) If the risk index is R≧RH, it is risk level 3 (dangerous)

(2) If RH>risk index R≧RL, it is risk level 2 (note)

(3) If RL>risk index R, it is risk level 1 (safe)

Returning to FIG. 3, the transmitting unit 316 transmits the disease-related prediction information derived by the risk deriving unit 314 to the portable terminal 108 via the network 110. For example, the transmitting unit 316 transmits the derived time-series data of the future risk level to the portable terminal 108.

The operation of the disease prediction system 100 constructed by the above configuration is explained.

FIG. 9 is a flowchart showing the processing flow in the disease prediction system 100 of FIG. 1. When the portable terminal 108 receives an instruction to activate the disease prediction APP program through the input interface 125, the disease prediction APP program is activated (S902). For example, when the portable terminal 108 detects a tap on the icon of the disease prediction APP program displayed on the display 124, it starts the disease prediction APP program. If you log in for the first time (Y in S904), the processing proceeds to step S908. If it is not the completion of the initial registration (N in S906), the processing proceeds to step S906.

In step S906, the portable terminal 108 communicates with the disease prediction server 102 to perform the initial registration process. The portable terminal 108 generates an initial login screen (described later) and displays it on the display 124, whereby the user accepts the input of the initial login information, that is, login ID, password, gender, age, birth date, and prescription drugs. The portable terminal 108 registers the input initial registration information in the memory 121 and transmits it to the disease prediction server 102 via the network 110. The disease prediction server 102 receives the initial registration information and registers it in the user information holding unit 318 of the disease prediction server 102. In the initial registration process, the portable terminal 108 may display the usage agreement on the display 124, and accept the user's approval/non-approval of the usage agreement. The terms of use include: for example, regulations on personal information processing or regulations on data processing.

In step S908, the portable terminal 108 generates a login screen (described later) and displays it on the display 124, whereby the user accepts the input of the login information, that is, the user ID and password. The portable terminal 108 transmits the entered user ID and password to the disease prediction server 102 via the network 110. The disease prediction server 102 uses the received user ID and password to perform user authentication. The disease prediction server 102 transmits the result of user authentication to the portable terminal 108 via the network 110. Among them, user authentication may also be performed on the portable terminal 108.

If the result of the user authentication indicates failure, the portable terminal 108 notifies the user of the information, and the figure asks for the user ID and password to be entered again. If the result of user authentication indicates that the portable terminal 108 is successful, the gender/age of the user, the target disease, and the positioning function of the portable terminal 108 (such as GPS) will be retained. Etc.) The acquired position of the portable terminal 108 is transmitted to the disease prediction server 102 via the network 110 (S910). Here, instead of transmitting the gender/age in step S910, the gender/age of the user registered in the user information holding unit 318 of the disease prediction server 102 in step S906 may be used. In addition, the portable terminal 108 can also obtain the position of the portable terminal 108 by allowing the user to input.

The disease prediction server 102 receives the age/sex of the user and the target disease and the location of the portable terminal 108, and selects the corresponding model coefficients from the model holding unit 306 (S912). The disease prediction server 102 obtains weather prediction information from the weather data server 106 (S914). The disease prediction server 102 derives the future risk level of the target disease from the selected model coefficients and the acquired weather prediction information (S916). The disease prediction server 102 transmits the derived future risk level to the portable terminal 108 via the network 110 (S918).

The portable terminal 108 generates a risk prediction screen (described later) based on the received future risk level, and displays it on the display 124 (S920). The risk prediction screen is a screen showing the future risk level of the target disease in time series. The risk prediction picture is the past risk The level is expressed in a time series in line with the future risk level. The past risk level can also be obtained by the portable terminal 108 reading out the past risk level received by the disease prediction server 102 and maintained. Or in addition, the disease prediction server 102 may also use the selected model coefficients and the past meteorological information obtained by the meteorological data server 106 (for example, AMeDAS (Automated Meteorological Data Acquisition System) observations). Value or ground truth), the risk level is derived, and the derived past risk level is provided to the portable terminal 108.

The portable terminal 108 is on the risk prediction screen, and the user accepts the request for information registration. In accordance with the request, the portable terminal 108 generates an information registration screen (described later) and displays it on the display 124, thereby accepting input of health status/medical information at the past or present time (S922). Health status/medical information includes: information indicating health status, information about examinations, and information about dosing.

The portable terminal 108 transmits the data corresponding to the inputted health status/medical information and the user ID to the disease prediction server 102 via the network 110 (S924). The disease prediction server 102 associates the received user ID with the health status/medical information and registers it in the actual performance information holding unit 320 (S926). The disease prediction server 102 refers to the information held in the actual performance information holding unit 320, and updates the model for each user (S928). The updated model is used in the next risk level derivation.

The actual performance information holding unit 320 holds the actual performance value of the user's health state at a certain point in the past. On the other hand, the risk level at the same point in the past can be derived from the model selected by the user and the weather information at the same point in the past. In step S928, the model coefficients are adjusted to reduce the deviation between the risk level derived as described above and the actual performance value of the health state. This adjustment is done for each user, so even if the same model was originally used by two users, there are cases where the model coefficients of the two users become different according to subsequent adjustments.

FIG. 10 is a representative screen diagram of the initial login screen 400. The initial login screen 400 is provided with an initial login information input area 402 for accepting input of initial login information, and a sending button 404. The user enters information in the initial login information input area 402 and taps the send button 404. In this way, the portable terminal 108 obtains the information input into the initial registration information input area 402 as the initial registration information, and sends it to the disease prediction server 102.

FIG. 11 is a representative screen diagram of the login screen 406. The login screen 406 has an ID area 408 that accepts input of a login ID, a PW area 410 that accepts input of a password, and a sending button 412. The user inputs information in the ID area 408 and the PW area 410, and taps the send button 412. In this way, the portable terminal 108 obtains the information input into the ID area 408 and the PW area 410 as the login ID and password, respectively, and sends them to the disease prediction server 102.

FIG. 12 is a representative screen diagram of the risk prediction screen 414 showing the current risk level. Risk prediction screen 414 has: disease selection area Domain 416; the area display area 418 corresponding to the location where the message was sent in step S910; the date display area 420 that displays the current date; the risk level display area 422 that displays the risk level slightly later than the present, for example, 30 minutes later; and The time transition display area 424 of the weather and the risk level is displayed.

In the disease selection area 416, the currently selected target disease (asthma in the case of FIG. 12) is displayed. The disease selection area 416 is configured to select other diseases. For example, if the user designates the drop-down mark of the disease selection area 416, a list of selectable diseases is displayed, and the user selects the desired disease from the list. The portable terminal 108 acquires the selected disease as the target disease, and for the acquired new target disease, step S910 is executed again, and in step S918, the future risk level for the target disease is acquired, based on the acquired disease. The risk level constitutes the risk prediction picture 414 again.

In the risk level display area 422, a risk level slightly later than the present (that is, in the near future) is displayed. By viewing the display, the user can grasp the symptom risk or the severe risk of the target disease in the near future. Even if the display system is the same disease, the same area, and the same time, it varies from user to user.

The transition display area 424 includes: a setting button 426, a comment display area 428 showing comments about future risk levels or weather forecasts, a temperature graph 430 showing the temporal transition of temperature, a weather display area 432 showing changes in weather, and The medium-risk area 434 and the low-risk area 436 of the time transition of the risk level, and the date displaying the date and time Period time display area 438. In the transition display area 424, the arrow displayed along with the setting button 426 indicates the present, the past weather and past risk level information is displayed on the left side, and the future weather and future risk level information is displayed on the right side.

If the user designates (for example, taps) the setting button 426, the portable terminal 108 accepts the designation as a request for information registration from the user, and displays the information registration screen 440 on the display 124. FIG. 13 is a representative screen diagram of the information registration screen 440. The information registration screen 440 is, for example, displayed in the form of a pop-up window with respect to the transition display area 424. The information registration screen 440 has: a health status registration area 442 for inputting the user's current health status in a selection form; a voice registration area 444 for stimulating the user's voice input; and a medical history registration area 446 for registering the medical history The medication record area 448 for registering the medication history; the memo area 450 for accepting the input of the memo; and the registration button 452. In the health status registration area 442, the health status of the target disease (asthma in the case of FIG. 13) is entered. If the user taps the voice registration area 444, a predetermined voice input interface is activated. The portable terminal 108 obtains the user's speech through the interface, and detects the fatigue degree from the obtained speech. If the user taps the registration button 452, the portable terminal 108 obtains the information entered into the health status registration area 442, the medical history registration area 446, the medication record area 448, and the memo area 450 and the detected fatigue level as The health status/medical information is stored in the memory 121 and sent to the disease prediction server 102.

Returning to Figure 12, the comment display area 428 shows: risk comments about future risk levels (the situation in Figure 12 is "Watch out tonight"), and weather comments related to weather forecasts (the situation in Figure 12 is "Watch for temperature drops" ). The weather comment is generated based on the information of the weather forecast. Risk comments are generated based on future risk levels. The risk comment system includes: the time period part (the situation in Figure 12 is "tonight") and the risk part (the situation in Figure 12 is "attention"). The time period part corresponds to the time or time period corresponding to the position on the display area 424 when the comment display area 428 is displayed. The risk part corresponds to the risk level at that moment or time period. For example, if the risk level is expressed in five stages, ranging from 1 to 5, the risk part is selected from among "danger, severe alert, vigilance, attention, and safety" corresponding to the risk level.

The transition display area 424 visually informs the time transition of the risk level. In FIG. 12, the transition display area 424 includes a medium-risk area 434 and a low-risk area 436. The medium-risk area 434 is to add network cables or add colors (red, blue, green, etc.). The low-risk area 436 is a thicker mesh line or a lighter color or a colorless area than the medium-risk area 434. The medium-risk area 434 indicates that the risk level of asthma is 5 stage 2, which is the attention level, and the low-risk area 436 indicates that the risk level is 1, which is the safe level.

FIG. 14 is a representative screen diagram of the risk prediction screen 454 showing the future risk level. If the user swipes the risk prediction screen 414 shown in FIG. 12 to the left, the portable terminal 108 scrolls the screen horizontally to the left to display the risk prediction screen 454 shown in FIG. 14. Risk prediction screen The 454 series has: a disease selection area 416; an area display area 418; a date display area 456 that displays a future date; a risk level display area 464 that displays a risk level in the future; and a display that displays the change in time of weather and risk levels Area 462.

In the risk level display area 464, the future risk level is displayed. By viewing the display, the user can grasp the risk of symptomatic or severe disease of the target disease in the future. Even if the display system is the same disease, the same region, and the same time, it varies from user to user.

The transition display area 462 includes a setting button 458, a comment display area 428, a temperature graph 430, a weather display area 432, a high-risk area 460 that indicates the temporal transition of the risk level, and a date and time display area 438. In the transition display area 462, the arrow displayed along with the setting button 458 indicates the future time point of the display object.

When the user designates the setting button 458, the portable terminal 108 accepts the designation as a request for registration of a location from the user, and displays a location registration screen (not shown) on the display 124. The user can specify the desired area through the location registration screen. The portable terminal 108 executes step S910 again for the designated area, obtains the future risk level for the area in step S918, and re-constructs the risk prediction screen 454 based on the obtained risk level.

The transition display area 462 visually informs the time transition of the risk level. In FIG. 14, the transition display area 462 includes the high-risk area 460. The high-risk area 460 is thicker or darker than the medium-risk area 434. High-risk area 460 indicates asthma The risk level is 3 of the five stages, that is, the alert level.

FIG. 15 is a representative screen diagram of the risk prediction screen 466 showing the past risk level. If the user swipes to the right on the risk prediction screen 414 shown in FIG. 12, the portable terminal 108 scrolls the screen horizontally to the right to display the risk prediction screen 466 shown in FIG. 15. The risk prediction screen 466 has: a disease selection area 416; an area display area 418; a date display area 468 that displays past dates; a risk level display area 476 that displays past risk levels; and time transitions that display weather and risk levels The transition display area 470.

In the risk level display area 476, the past risk level is displayed. The transition display area 470 has a setting button 472, a temperature graph 430, a weather display area 432, a low risk area 436, a date and time display area 438, and a registration data area 474. In the transition display area 470, the arrow displayed along with the setting button 472 indicates the past time of the display object.

If the user designates the setting button 472, the portable terminal 108 accepts the designation as a request for information registration from the user, and displays the information registration screen on the display 124. This information registration screen has the same structure as the information registration screen 440 shown in FIG. 13 except that it does not have a voice registration area 444.

The registration data area 474 indicates that the health status/medical information is registered through the information registration screen at the time or time period corresponding to the part on the transition display area 470 where the area is displayed. If the user designates the login data area 474, the portable terminal 108 will accept the designation as Health status/requirements for viewing of medical information. The portable terminal 108 refers to the memory 121 to read out the health status/medical information corresponding to the registered data area 474 and displays it on the display 124.

In the above-mentioned embodiment, examples of the holding portion are a hard disk or a semiconductor memory. In addition, according to the description in this manual, the CPU, or the module of the installed application program, or the module of the system program, or the semiconductor memory that temporarily stores the contents of the data read from the hard disk, can be used It is understandable for those who are familiar with the technology in the field related to this specification to realize each part.

With the disease prediction system 100 of the present embodiment, a model corresponding to the attributes of the user is selected, and the user is notified of the risk level of the target disease in the future derived from the selected model and weather forecast information. Therefore, the notified risk level corresponds to the attributes of the user, and therefore the accuracy of the risk level can be improved.

Children and adults, men and women have different susceptibility to disease. In addition, according to the intensive research of the present inventors, there is a significant correlation between disease and weather. In this embodiment, a disease prediction APP program that combines these two items and provides personalized disease prediction to the user is realized.

If a parent has a child, by registering the child's attributes in the parent's portable terminal 108 instead of his own, he can know the risk of the child's asthma attack or becoming severe in the future, so that countermeasures can be taken in advance.

In addition, in the disease prediction system 100 of this embodiment, the disease prediction APP program is configured so that the user can select the target disease. In this way, according to the user, the disease concerned is different, and the corresponding information can be provided according to the user's needs.

In addition, in the disease prediction system 100 of the present embodiment, the risk prediction screen is configured so that the user can accept instructions for information registration. In addition, the risk prediction screen is configured so that the registered health status/medical information can be viewed later. Therefore, the user can view the past health status/medical information registered by him, and the user convenience is improved.

In addition, in the disease prediction system 100 of this embodiment, the user's model is adjusted based on the registered health status/medical information. Therefore, the user model can be more personalized, and the prediction accuracy of the risk level can be further improved. In addition, the detailed content of the adjusted model, such as the description of the change of the variable or the increase or decrease of the model coefficient, etc., is notified to the user, so as to provide the user with new insights. For example, if the user is concerned about his own asthma and believes that the cause is a temperature change, if the user is notified that the model coefficient of humidity is greater than the model coefficient of temperature due to model adjustment, the user can perceive that the humidity must actually be paid attention to. In this way, the user's actions are distorted.

In addition, in the disease prediction system 100 of this embodiment, the registered health status/medical information is transmitted to the disease prediction server 102 via the network 110, and is accumulated there. Therefore, the administrator of the disease prediction server 102 can obtain data that records the user's health status or drug administration performance in a time series. The manager can use the accumulated data as shown above in, for example, epidemiological research or medical research. Follow-up investigation on utility, etc. Or, in addition, if the user takes predetermined preventive measures against the disease, it is possible to verify the extent of the effectiveness of the preventive measures based on the accumulated data.

The configuration and operation of the disease prediction system 100 of the embodiment have been described above. This embodiment is an example, and various modifications are possible in the combination of each component or each process, and such modifications are also within the scope of the present invention, and are understood by those skilled in the art.

In the embodiment, the case where the disease prediction APP program is installed in the portable terminal is described, but it is not limited to this, and it can also be implemented by, for example, an HTML program through the browser of the portable terminal. Or in addition, a recording medium with a disease prediction APP program can also be provided.

In the embodiment, it is explained that the disease prediction server 102 performs model selection, weather prediction acquisition, and risk derivation, but it is not limited to this, and at least one of these functions may be provided in the portable terminal 108 . At this time, the required information such as model coefficients is transmitted from the disease prediction server 102 to the portable terminal 108 via the network 110.

In the embodiment, the case where the disease prediction server 102, the medical data server 104, and the weather data server 106 are formed as different individuals is explained, but it is not limited to this, and a disease prediction server can also be provided. The function of the medical data server 104 and the function of the medical data server 104 can also be provided with a server that has the function of the disease prediction server 102 and the function of the weather data server 106. Device.

In the embodiment, it is explained that the future risk level and its time transition are displayed on the display 124, but it is not limited to this. If the information about the future risk level is notified to the user, it can also be borrowed. The user is notified by, for example, sound or vibration.

In the embodiment, it is mainly explained that the risk of occurrence of symptoms of asthma is derived from a model representing the relationship between asthma and weather, but it is not limited to this. For example, a model representing the relationship between the degree of air pollution and respiratory system diseases, a model representing the relationship between climate and skin disorders, or a model representing the relationship between weather and mental disorders, etc. may be generated and used. Or, in terms of disease, not only asthma, but also colds, atopic dermatitis, headaches, strokes, cerebral infarctions, or myocardial infarctions.

In the embodiment, it is explained that the target disease can be selected, but it is not limited to this. For example, the portable terminal obtains the risk level of various diseases, and informs the user of the information of the disease with a relatively high risk level. Even under the same weather conditions, the risk level of the disease may vary according to the attributes of the user. For example, when the temperature is low, the risk level of asthma for men under 10 is higher than that of other diseases, and for women aged 80 to 89, the risk of myocardial infarction is higher than that of other diseases (including asthma). The risk level situation. In the situation shown above, the disease prediction APP program determines and informs the disease with a relatively high risk level according to the user's attributes, thereby providing the user with more useful information.

In the embodiment, the case where the disease prediction server 102 obtains medical data such as prescription data from the medical data server 104 through the network 110 is explained, but it is not limited to this. It is also possible, for example, that the administrator of the disease prediction server obtains the prescription data offline when the model is generated, and compares it with the weather data. Therefore, the created model is static, and the created model is stored in the disease prediction server. Among them, as shown in the embodiment, the disease prediction server 102 cooperates with the medical data server 104 online. If the model is dynamically created, the model can be automatically updated regularly, so it is more suitable.

100: Disease prediction system

102: Disease prediction server

104: Medical Data Server

106: Weather Data Server

108: portable terminal

110: Network

112: Medical Big Data

114: Weather Big Data

Claims (8)

A computer program that is used to make a computer execute each project in a method executed in a terminal. The aforementioned method includes: a project that accepts the input of the user's attributes; Send to the server through the network. Through the network, the server obtains the model selected from the multiple models representing the relationship between the disease and the weather corresponding to the input attributes, and the weather forecast The project of predicting disease information obtained from the information; the project of displaying the predicted information about the disease obtained in the foregoing on the display in a time series to notify the user; through the aforementioned screen, receiving information about the user’s health The process of inputting information; and the process of updating the selected model by sending the inputted information about the user's health to the server through the foregoing network. For example, the computer program of item 1 of the scope of patent application, in which the predictive information about the disease includes information indicating the risk of symptoms. For example, the computer program of the first item of the scope of patent application, which additionally includes: the project obtained by the region, the aforementioned project obtained by the server includes: by additionally combining the aforementioned The obtained area is sent to the server through the network, and the server obtains the model selected from the different plural models that represent the relationship between disease and weather, and the attribute corresponding to the input through the network. The project of predicting information about diseases obtained from the information of weather forecasts in the obtained area. For example, the computer program of any one of items 1 to 3 of the scope of patent application, wherein the aforementioned screen is coordinated with: predictive information about the disease, and information about the disease obtained from the selected model and past weather information The past information is expressed in time series. The aforementioned screen includes an area for accepting instructions from the user. The aforementioned project for accepting input of information about the user’s health is to accept the past corresponding to the aforementioned area if the aforementioned area is designated. Or the input of information about the user’s health at the current point in time. For example, in the computer program of item 4 of the scope of patent application, the health-related information includes at least one of: information indicating the health status, information about the examination, information about the administration of drugs, and the user's words. A terminal machine is equipped with: a means of accepting the input of the user’s attributes; by sending the input attributes to a server through the network, through the aforementioned network, the aforementioned server obtains information indicating diseases and Among the different complex models of meteorological relations, the selected model corresponding to the input attribute Means of predicting information about diseases based on information from weather forecasts; a means of notifying the user by displaying the previously obtained predictive information about diseases on a screen in a time series on the monitor; accepting through the aforementioned screens A means of inputting information about the user’s health; and a means of updating the selected model by sending the inputted information about the user’s health to the server through the aforementioned network . A prediction method, which is a prediction method executed in a terminal, which includes: a project that accepts the input of the user's attributes; by sending the inputted attributes to the server through the network, through the aforementioned network , The project of obtaining, from the aforementioned server, the predicted information about the disease obtained from the model selected from the multiple models representing the different relationship between the disease and the weather corresponding to the input attributes, and the information of the weather forecast; The acquired predictive information about the disease is displayed on the display in a time-series screen to notify the user of the process; the process of accepting the input of information about the user’s health through the foregoing screen; and by combining the foregoing The input information about the user’s health is sent to the aforementioned server through the aforementioned network, so that the aforementioned selected model The updated project. A server is provided with: holding means for holding plural models representing the difference in the relationship between disease and weather; first receiving means for receiving the attributes of the user through a network through a terminal; Among the plural models of methods, the selection method of selecting the model corresponding to the acquired attributes; the method of obtaining weather forecast information from the weather server through the network; the selected model and the acquired weather forecast Information about the disease, the means of deriving predictive information about the disease; the means of sending the derived predictive information about the disease to the terminal through the network; through the aforementioned network, the aforementioned terminal receives information about the user Health information; and, based on the aforementioned received information about the user's health, an update method for updating the aforementioned selected model.

Images