KR20230106493A - A health prediction system having a personalized health prediction learning model and a method for controlling the system - Google Patents

Abstract

The present invention relates to a communication unit that performs wireless communication with a collection terminal that collects user's biometric information measured by at least one measurement device, and when characteristic information of the at least one measurement device is received through the communication unit, the received measurement device An artificial intelligence unit that generates a predictive model for a specific disease by reflecting the characteristic information of the user and performs learning on the predictive model based on the collected biometric information of the user, and when the predictive model is learned, Controls the artificial intelligence unit to calculate a predictive value related to the specific disease in response to the biometric information collected in the collection terminal from the learned predictive model, and based on the calculated predictive value and a preset threshold, the specific disease and a control unit for discriminating a prediction result for the prediction result and transmitting the prediction result to the collection terminal so that the determined prediction result is displayed through the collection terminal.

Description

Health prediction system with personalized health prediction learning model and control method of the system

The present invention relates to a health prediction system for diagnosing a user's health condition and predicting a disease, and more particularly, to a health prediction system having a personalized health prediction learning model specialized for a user through learning.

Currently, with the development of science and technology, various medical devices are appearing. In particular, with the advent of wearable devices worn by users, many wearable devices having a medical function of diagnosing a user's health state are appearing in consideration of the fact that wearable devices are worn by users.

As these medical devices appeared en masse, a healthcare platform providing health information of users collected from the medical devices appeared. Such a healthcare platform collects collected medical information and provides the collected information to users so that users can check information about their own health conditions.

However, as medical devices are diversified as medical devices appear en masse, devices that measure the same health condition are manufactured by different manufacturers or have different attachment forms or measurement forms. In addition, there are problems in that different diagnosis results are output according to the device characteristics of the measuring device despite checking the same health status of the user due to differences in manufacturing, attachment, or measurement. Accordingly, there is a problem in that the reliability of the diagnosis result is lowered because the diagnosis result of the health care platform varies depending on which manufacturer, which type, and which method is checked, despite the same user's same health condition.

In addition, in the case of medical diagnosis information, the influence of the user's individual biometric characteristics may have a very large effect. Therefore, a personalized diagnosis system is required for users, but current health care platforms provide only diagnosis results determined based on predetermined threshold values of collected data, and do not provide such customized diagnosis services.

In addition, in the case of medical information, it is information about a user's individual physical characteristics, and may be very sensitive personal information. Therefore, it is very important to maintain the security of such medical information to protect personal privacy.

However, such medical information should be disclosed to designated third parties if necessary. For example, in the case of family members, medical information should be able to be shared, and in the case of a third party, such as a doctor or fitness trainer, who needs to check the user's health condition, the medical information may need to be shared. That is, it is necessary to share medical information with an authorized third party.

Therefore, there is a need for a medical information record management method that strictly manages the user's medical information records stored in the healthcare platform so as not to be leaked and provides them only to authorized third parties.

An object of the present invention is to solve the above-described problems and other problems, and the user's biometric information is obtained from the user's biometric information through a user-customized health prediction learning model in which the characteristics of a measuring device used for measuring biometric information for medical diagnosis are reflected. An object of the present invention is to provide a health prediction system capable of diagnosing a person's health status and predictability of diseases, etc., and a control method of the system.

Another object of the present invention is to provide a health prediction system capable of dynamically determining whether or not a disease occurs according to a user's health condition according to a user's situation and a method for controlling the system.

Another object of the present invention is to provide a health prediction system and a control method of the system, which can provide diagnosis results for biometric information measured from a user in real time simultaneously with measurement of the biometric information.

In addition, the present invention prevents the medical information measured and stored from the user from being provided to a third party, and allows the authorized third party to view the medical information so that the authorized third party can check the user's medical information. Its object is to provide a health prediction system and a control method of the system.

According to one aspect of the present invention to achieve the above or other object, the server of the health prediction system according to an embodiment of the present invention wirelessly communicates with a collection terminal that collects user's biometric information measured by at least one measuring device. When the characteristic information of the at least one measurement device is received through the communication unit, a predictive model for a specific disease is created by reflecting the characteristic information of the measurement device, and the collected biometric information of the user is generated. an artificial intelligence unit that learns the prediction model based on the prediction model, and when the prediction model is learned, the prediction related to the specific disease is predicted in response to the biometric information collected in the collection terminal from the learned prediction model. The artificial intelligence unit controls the artificial intelligence unit to calculate a degree, determines a prediction result for the specific disease based on the calculated predictive degree and a preset threshold, and displays the determined prediction result through the collection terminal. and a controller for transmitting to the collection terminal.

In one embodiment, the collection terminal obtains information related to the user's situation from at least one peripheral device around the user, analyzes the user's situation based on the obtained information, and provides situational information about the analyzed user's situation. It is transmitted to the server, and the server changes a threshold value based on the situation information, and determines a prediction result for the specific disease based on the changed threshold value and the predictive value.

In one embodiment, the at least one peripheral device is characterized in that information is at least one mobile terminal located near a user or an access point (AP) installed in an area where the user is located.

In one embodiment, the predictive model is applied to at least one exclusive layer that is a hidden layer including weights applied to input biometric information classified into a specific group and to all input biometric information. An artificial neural network including at least one common layer that is a hidden layer including weights, and the biometric information is input to either the exclusive layer or the common layer according to whether it is classified into the specific group, The prediction diagram output as a result of the predictive model is characterized in that it is output from any one of the at least one common layer.

In one embodiment, the control unit may input the biometric information into the predictive model as the biometric information of the specific group according to the characteristic information of the measuring device in which the biometric information is measured.

In an embodiment, the characteristic information of the measuring device may include a manufacturer of the measuring device, a type of biometric information measured by the measuring device, a type of the measuring device, a mounting method for a user to mount the measuring device, and the measurement device. It is characterized in that the device includes at least one of measurement methods for measuring the user's biometric information.

In an embodiment, the characteristic information of the measurement device may include quantified clinical accuracy obtained by quantifying the clinical accuracy of the measurement device.

In one embodiment, the control unit updates the predictive model when a preset update condition is satisfied, and the preset update condition is a measurement device that measures a user's biometric information when a preset update period expires. It is characterized in that is satisfied when is added, replaced, or any one measuring device is removed.

In one embodiment, the control unit changes a weight assigned to at least one node among hidden layers constituting the prediction model, removes at least one of the hidden layers, or creates a new value including nodes to which a new weight is assigned. Updating the predictive model to further include a hidden layer, and determining the changed weight based on device characteristic information of the added, replaced, or removed measuring device, or determining the hidden layer to be removed or the new hidden layer. to be

In one embodiment, the control unit generates at least one new prediction model according to the device characteristic information of the added, replaced, or removed measuring device, and the existing prediction model and the at least one new prediction model for the collected biometric information of the user. It is characterized in that a statistical value of predictive degrees calculated in each of one new predictive model is calculated as the predictive degree.

In one embodiment, the control unit combines at least one predictive model derived in a process of learning the predictive model based on the collected biometric information of the user with the learned predictive model to obtain the learned predictive model. It is characterized by updating.

In one embodiment, the control unit compares weights of hidden layers corresponding to each other from the learned prediction model and the at least one derived model, and adds nodes to which different weights have been assigned to obtain the learned prediction model and It is characterized in that combining the at least one derivative model.

In one embodiment, when the measuring device is a measuring device for measuring the user's blood sugar, the predictive model is a model obtained by learning the user's blood sugar change amount varying according to each nutrient, and the control unit is a collection terminal for the user. If a virtual diet is entered through, the nutrients of each food included in the input diet are analyzed, and based on the change in blood sugar of the user according to each analyzed nutrient and the intake of each food input together with the virtual diet, , The user's total blood sugar change corresponding to the virtual menu is calculated and transmitted to the collection terminal in response to the input of the virtual menu.

In one embodiment, the control unit, when learning of the predictive model is completed, determines a range of a minimum value and a maximum value of the user's biometric information that can be detected through the measurement device from the basic biometric characteristics of the user, and determines the determined Controls the artificial intelligence unit so that predictive degrees of the predictive model corresponding to each biometric information included in the range of minimum and maximum biometric information are calculated, and some of the calculated predictive degrees are stored as a caching table. Characterized in that it is transmitted to the collection terminal.

In one embodiment, when the caching table is received from the server, the collection terminal detects a predictive value corresponding to the biosignal collected from the measurement device from the caching table, and detects the predicted degree from the detected caching table. It is characterized in that the predicted result for the specific disease is determined based on the predictive value and the threshold value according to the situation information, and the determined predicted result is output.

In one embodiment, the server further includes a memory for storing the user's biometric information collected through the communication unit by dividing it according to a time at which the biometric information was measured and a type of the biometric information, and the control unit controls the It is characterized in that address information related to a storage area in which the divided biometric information is stored in the memory is provided as a response to a user's request.

In one embodiment, the control unit registers a token indicating the authority of a third party to request the user's biometric information through a blockchain network, and responds to the user's request based on the registered token. Provides the address information to the third party according to the above, and limits the scope of the user's biometric information that can be viewed by the third party and the valid period during which the user's biometric information can be viewed based on the token registered in the blockchain network characterized by

According to one aspect of the present invention to achieve the above or other objects, a control method of a health prediction system according to an embodiment of the present invention includes, in a collection terminal of the health prediction system, at least one device for measuring user's biometric information. connecting communication with a measuring device and collecting device characteristic information of the at least one measuring device; receiving, by a cloud server, the characteristic information of the measuring device collected from the collection terminal; Generating a predictive model for a specific disease in which characteristic information is reflected; receiving, by the collection terminal, biometric information of a user collected by the at least one measurement device; and, by the cloud server, the collection terminal collects receiving biometric information and learning the predictive model based on the received biometric information; and when the learning of the predictive model is completed by the cloud server, the biometric information received from the collection terminal from the learned predictive model. Calculating a predictive value corresponding to , and determining, by the cloud server or the collection terminal, a prediction result for the specific disease based on a predetermined threshold value and the predictive value, and outputting the determined prediction result. It is characterized in that it includes.

In an embodiment, the characteristic information of the measuring device may include a manufacturer of the measuring device, a type of biometric information measured by the measuring device, a type of the measuring device, a mounting method for a user to mount the measuring device, and the measurement device. It is characterized in that the device includes at least one of measurement methods for measuring the user's biometric information.

In one embodiment, the calculating of the predictive value may include determining, by the cloud server, a range of a minimum value and a maximum value of the user's biometric information that can be detected through the measurement device from the basic biometric characteristics of the user. and calculating, by the cloud server, predictive degrees of the learned prediction model corresponding to each of the biometric information included in the range of the minimum and maximum values of the biometric information, and the calculated predictive degrees by the cloud server. and transmitting a part of the prediction result as a caching table to the collection terminal, and the step of outputting the prediction result by the cloud server or the collection terminal includes: Detecting one of the predictive degrees corresponding to the biometric information received from the measurement device, and the collection terminal predicting the specific disease based on the detected predictive degree and a predetermined threshold value. and outputting the determined prediction result.

A health prediction system and a control method of the system according to the present invention is a user's health condition through a learning model learned according to the characteristics of a measuring device used to measure biometric information for a user's medical diagnosis and the user's biological characteristics. Since it diagnoses, there is an effect that a customized medical diagnosis service specialized for the user can be provided.

In addition, the present invention detects the situation around the user based on at least one other device and analyzes the diagnosis result of the pre-learned health prediction model by reflecting different threshold values according to the detected user situation, thereby determining whether or not a disease has occurred. It can be determined dynamically according to the user's situation. Therefore, when the user can receive medical services, it is possible to more sensitively determine whether or not a disease has occurred, so that the user can more quickly cope with an outbreak and prepare for an outbreak situation in advance.

In addition, the present invention provides an authorized third party with only the address information of the storage where the medical information is stored, not the previously stored medical information of the user, and allows only viewing of the medical information through the address information, thereby providing access to the third party. There is an effect of preventing medical information from being transmitted and minimizing leakage of medical information. In addition, by distributing and storing the address information in a blockchain method, security of medical information according to the address information can be enhanced and leakage of the address information can be prevented.

1 is a block diagram showing the configuration of a health prediction system according to an embodiment of the present invention.
2 is a block diagram showing the configuration of a server of a health prediction system according to an embodiment of the present invention.
3 is a block diagram illustrating the configuration of a collection terminal that serves as a personal health gateway (PHG) of a health prediction system according to an embodiment of the present invention.
4 is a block diagram showing the configuration of a biometric information measurement device (PHD, Personal Health Device) of a health prediction system according to an embodiment of the present invention.
5 is a flowchart illustrating an operation process in which a health prediction system according to an embodiment of the present invention learns a predictive model for diagnosing a user's health state according to characteristics of a measuring device used by the user.
6A to 6B are exemplary diagrams for explaining characteristics of a measuring device.
7 is an exemplary diagram illustrating an example of a predictive model for diagnosing a health state of a user in a health prediction system according to an embodiment of the present invention.
8 is a flowchart illustrating an operation process in which a result of predicting whether or not to have a specific disease by reflecting a user's situation is predicted in the health prediction system according to an embodiment of the present invention.
9 is an exemplary diagram illustrating an example in which a threshold value is set differently according to a detected user's situation in the health prediction system according to an embodiment of the present invention.
10 is a flowchart illustrating an operation process of learning a change in blood sugar according to a user's diet through a health prediction system according to an embodiment of the present invention.
FIG. 11 is a flowchart illustrating an operation process for providing a result of predicting a change in blood sugar according to a user's diet through the prediction model learned in FIG. 10 .
12A to 12C are conceptual diagrams illustrating methods of updating a prediction model in a health prediction system according to an embodiment of the present invention.
13 is a flowchart illustrating an operation process of providing a caching table including diagnosis prediction results according to user's biometric information in the health prediction system according to an embodiment of the present invention.
14 is an exemplary diagram illustrating an example in which a caching interval for extracting the caching table of FIG. 13 is determined according to user's biometric information and characteristics.
15 is a flowchart illustrating an operation process of providing diagnosis prediction results using a caching table in the health prediction system according to an embodiment of the present invention.
16 is a flowchart illustrating an operation process of allowing an authorized third party to view medical information of a user in the health prediction system according to an embodiment of the present invention.

It should be noted that the technical terms used in this specification are only used to describe specific embodiments and are not intended to limit the present invention. Also, singular expressions used in this specification include plural expressions unless the context clearly indicates otherwise. The suffixes "module" and "unit" for components used in the following description are given or used interchangeably in consideration of ease of writing the specification, and do not in themselves have a meaning or role distinct from each other.

As used herein, “consists of.” or "includes." Such terms should not be construed as necessarily including all of the various components or steps described in the specification, and some of the components or steps may not be included, or additional components or steps may be included. It should be interpreted as being more inclusive.

In addition, in describing the technology disclosed in this specification, if it is determined that a detailed description of a related known technology may obscure the gist of the technology disclosed in this specification, the detailed description will be omitted.

In addition, the accompanying drawings are only for easy understanding of the embodiments disclosed in this specification, the technical idea disclosed in this specification is not limited by the accompanying drawings, and all changes included in the spirit and technical scope of the present invention , it should be understood to include equivalents or substitutes. In addition, each of the embodiments described below, as well as combinations of embodiments, are changes, equivalents, or substitutes included in the spirit and technical scope of the present invention, and may fall within the spirit and technical scope of the present invention. .

1 is a block diagram showing the configuration of a health prediction system according to an embodiment of the present invention.

Referring to FIG. 1 , the health prediction system according to an embodiment of the present invention includes at least one measuring device (PHD) 30 capable of measuring user's biometric information, and the at least one measuring device 30 It may be configured to include a collecting terminal (PHG) 20 that collects and stores the biometric information of the user to be measured, and a server 10 that receives the measurement information stored in the collection terminal 20 .

First, the measurement device 30 may be a device that measures user's biometric information. The measurement device may be classified according to user's biometric information (eg, blood pressure, weight, blood sugar, heart rate, etc.) to be measured. In addition, even if the biometric information to be measured is the same, it may be distinguished from each other according to a manufacturer, a wearing method, or a method of measuring the biometric information.

The measurement device 30 may automatically measure the user's biometric information according to a user's selection or according to a preset period, and transmit the measured biometric information to the collection terminal 20 in a preset communication method. A more detailed configuration of the measuring device will be described in more detail with reference to FIG. 4 below.

Also, the collection terminal 20 may be connected to the at least one measurement device 30 through wireless communication. In addition, biometric information measured from each of the at least one measuring device 30 may be received and stored. In this case, the collection terminal 20 may request the measurement device to measure the biometric information at a preset period, and the measurement device measures the user's biometric information in response to the request and sends the measured information to the collection terminal 20. can be sent to That is, the collection terminal 20 is a gateway between the server 10 and the measurement device 30, PHG ( Personal Health Gateway).

Meanwhile, the collection terminal 20 may be connected to the server 10 through wireless communication in a preset manner. In addition, the measurement information collected by the at least one measurement device 30 may be transmitted to the server 10 according to a preset period or immediately after the measurement information is collected. In response to the transmitted measurement information, a diagnosis prediction result of diagnosing the user's health condition may be received. In addition, the received diagnostic prediction result may be displayed so that the user can identify it.

Meanwhile, a wireless communication method in which the collection terminal 20 and the server 10 are connected (hereinafter referred to as a first communication method) and a wireless communication method in which the collection terminal 20 and the at least one measurement device 30 are connected (hereinafter referred to as a first communication method) The second communication method) may be different from each other.

For example, the first communication method may be a communication method performed by transmitting and receiving a radio signal through a communication network based on wireless Internet technologies. In this case, wireless Internet technologies include WLAN (Wireless LAN), Wi-Fi (Wireless-Fidelity), Wi-Fi (Wireless Fidelity) Direct, DLNA (Digital Living Network Alliance), WiBro (Wireless Broadband), WiMAX (World Interoperability for Microwave Access), High Speed Downlink Packet Access (HSDPA), High Speed Uplink Packet Access (HSUPA), Long Term Evolution (LTE), Long Term Evolution-Advanced (LTE-A), and 5G communication technologies.

In this case, the server 10 may form a cloud network through a communication network based on the wireless Internet technologies. And the server 10 may be connected through the cloud network. In this case, the server 10 may be a cloud server.

Meanwhile, the second communication method may be a communication method according to a short-range communication technology. For example, the short-range communication technology includes Bluetooth™, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, Near Field Communication (NFC), Wi-Fi (Wireless-Fidelity), Wi-Fi Direct, Wireless USB (Wireless Universal Serial Bus), etc. More preferably, the second communication method may be a Bluetooth communication method, in which case the collection terminal 20 and at least one measurement device 30 use Bluetooth Low Energy (BLE), generic ATTribute profile (GATT), which is a Bluetooth Low Energy protocol. A communication connection may be made based on a protocol.

Meanwhile, the collection terminal 20 may be a user's mobile terminal. In this case, the mobile terminal includes a mobile phone, a smart phone, a laptop computer, a digital broadcast terminal, a personal digital assistant (PDA), a portable multimedia player (PMP), a navigation device, a slate PC, and a tablet. It may include a PC (tablet PC), an ultrabook, a wearable device (eg, a watch type terminal (smartwatch), a glass type terminal (smart glass), a head mounted display (HMD)), and the like. .

In addition, the collection terminal 20 may be connected to at least one peripheral device other than the at least one measurement device 30 . In addition, information sensed by the at least one connected peripheral device may be collected.

For example, the peripheral device may be at least one wearable device 21 worn by a user. In this case, the user's location information or current time information collected in the wearable device may be collected in the collection terminal 20 . Also, the peripheral device may be an access point (AP) disposed in a predetermined specific area. In this case, the AP may include information about an area in which it is deployed. Therefore, the aggregation terminal 20 can obtain information about the currently located area based on the information collected by the AP.

As such, the information on the peripheral devices collected by the collection terminal 20 may be stored as context information related to the user's current context. Then, the collection terminal 20 can infer the user's current context based on the collected context information. In addition, the inferred user's situation information may be transmitted to the server 10 together with the biometric information measured by the at least one measuring device 30 . A more detailed configuration of the collection terminal 20 will be described in more detail with reference to FIG. 3 below.

The server 10 may be connected to the collection terminal 20 through a cloud network. In addition, collected and stored information may be transmitted from the collection terminal 20 . In addition, the user's health condition may be diagnosed based on the information transmitted from the collection terminal 20, and the diagnosed prediction result may be provided to the collection terminal 20 as a response to the received information.

Here, the prediction result may be a prediction result for a specific disease or health condition. For example, the specific disease or health condition may be diabetes or obesity. In this case, the server 10 may calculate the predictive value of diabetes or obesity of the user based on the information transmitted from the collection terminal 20, and predict the possibility of developing diabetes or obesity based on the calculated predictive value. In addition, the predicted result may be provided to the collection terminal 20 as the diagnostic prediction result.

To predict health state diagnosis, the server 10 may have a predictive model capable of predicting the probability of occurrence of the specific disease or health state based on input biometric information.

For example, the predictive model may be configured in the form of an artificial neural network in which learning using machine learning technology is performed. In this case, the predictive model may include weights set for each node constituting each hidden layer of a plurality of hidden layers. In addition, weights set for each node may be optimized through learning performed according to the machine learning technique.

Meanwhile, the server 10 may allow the prediction model to be learned by reflecting the characteristics of the measurement device 30 used by the user to obtain biometric information. For example, the server 10 may change weights applied to each hidden layer of the initial prediction model according to the characteristics of the measurement device 30 provided by the user. That is, as described above, when the measuring device 30 is classified according to a manufacturer, a wearing method, or a method for measuring biometric information, the server 10 determines the initial prediction model according to the manufacturer, wearing method, or measurement method. Weights assigned to each hidden layer may be set differently. To this end, the server 10 may request information on measurement devices that provide measurement information from the collection terminal 20, and the collection terminal 20 may identify a device collected from a user's input or a measurement device connected to the server 10. Device characteristic information identified based on the information, that is, information including the manufacturer, wearing method, or measurement method of the corresponding measuring device may be transmitted to the server 10 .

Meanwhile, the server 10 may classify inputs according to transmitted biometric information and input them to the predictive model. For example, the server 10 may input biometric information measured from devices having different characteristics according to the device characteristic information into the predictive model as input information of different groups. Alternatively, the server 10 may input different groups of input information to the predictive model according to the type of measured user's biometric information (eg, blood pressure, weight, blood sugar, heart rate, etc.). In this case, input information of different groups may be input to hidden layers of different levels or may be input to different nodes in the same level of hidden layers.

Meanwhile, the information transmitted from the collection terminal 20 may include not only the user's biometric information measured by the measurement device 30 but also information related to the user's situation (circumstance information).

Then, the server 10 may predict the health state of the user based on the received condition information with the predictive degree calculated by the predictive model. For example, the server 10 calculates the predictability of diabetes or obesity of the user based on the information transmitted from the collection terminal 20, and calculates the possibility of onset of a specific disease corresponding to the predictability calculated based on the situation information. Predictable.

In this case, the probability of developing the specific disease is the true positive rate (TPR), the probability of being diagnosed as positive in the presence of the disease, the false positive rate (FPR), the probability of being diagnosed as positive in the absence of the disease, or the negative rate in the absence of the disease. Probability of being diagnosed as True Negative Rate (TNR), Probability of being diagnosed as negative while having a disease can be predicted as False Negative Rate (FNR) (in this case, TPR = 1- FPR, TNR = 1- FNR ).

In this case, the true positive rate (TPR) or true negative rate (TNR) and the false positive rate (FPR) or false negative rate (FNR) may be determined based on a predetermined threshold value. That is, when the threshold value increases, the true positive rate or true negative rate increases (false positive rate or false negative rate decreases), and when the threshold value decreases, the false positive rate or false negative rate may increase (true positive rate or true negative rate decreases). Therefore, the lower the threshold value, the higher the probability of outputting a prediction result by determining that a patient has a disease even when there is no disease.

In this case, the server 10 may increase or decrease the threshold based on the situation information. For example, when the user is in a hospital, the server 10 may detect that the user is in the hospital through situation information and lower the threshold. In this case, since the false positive rate is high, even if the predictive value calculated according to the biometric information detected from the user does not actually reach the disease level (determination result according to the normal threshold), it can be detected as having the disease according to the lowered threshold. Accordingly, even when the user's biometric information changes slightly, a predicted result can be transmitted as having an onset, and the collection terminal 20 can output an alarm based on the predicted result. Therefore, according to the user's biometric information, the presence or absence of an onset can be more sensitively predicted, and as a result, it is possible to more quickly cope with an actual onset or to prevent an onset of an onset.

Meanwhile, the server 10 may store the user's biometric information provided from the collection terminal 20 . Further, address information of the storage area where the biometric information is stored may be provided to an authorized third party so that the authorized third party may view the stored user's biometric information.

In this case, the address information may be distributed and stored in a plurality of terminals constituting a block chain network. In this case, the address information may be distributed and stored in a plurality of data blocks. In addition, in the case of blockchain algorithms, security can be enhanced because access to data blocks is limited to those who have a token method. To this end, the server 10 may be connected to the blockchain network. A more detailed configuration of the server 10 will be described in more detail with reference to FIG. 2 .

Meanwhile, in the above description, an example in which the user's biometric information measured by at least one measuring device 30 is provided to the server 10 via the collection terminal 20 has been described. However, if the at least one measurement device 30 is provided with a communication module enabling communication connection with the server 10 and if the measured biometric information can be stored in the measurement device, the collection Of course, biometric information may be directly transmitted to the server 10 without passing through the terminal 20 . In this case, the collection terminal 20 may transmit to the server 10 only situation information corresponding to the current situation of the user determined based on information collected from at least one peripheral device.

2 is a block diagram showing the configuration of the server 10 of the health prediction system according to an embodiment of the present invention.

2, the server 10 of the health prediction system according to an embodiment of the present invention includes a controller 100 and a communication unit 110 connected to the controller 100 and controlled by the controller 100; It may include an authentication unit 120, a memory 130, and an artificial intelligence unit 150. The components shown in FIG. 2 are not essential to implement the server 10, so the server 10 described in this specification may have more or fewer components than the components listed above. there is.

First, the communication unit 110 may include one or more modules that allow the server 10 to perform wireless communication with at least one other device using a preset communication technology. For example, the communication unit 110 is at least one configured to transmit and receive wireless signals in a communication network according to WLAN, Wi-Fi, Wi-Fi Direct, DLNA, WiBro, WiMAX, HSDPA, HSUPA, LTE, LTE-A, 5G communication technology, etc. may include a wireless Internet module of Also, a communication network according to at least one of the wireless communication technologies may form a cloud network.

Further, the authentication unit 120 may store and authenticate various types of information for authenticating the reading authority of the biometric information stored in the server 10 . For example, the authentication unit 120 may store a third party's authentication key for authenticating an authorized third party, and if there is a requester to view biometric information, the authentication key of the requestor is received from the blockchain network and stored in advance. Through the authentication key, it is possible to authenticate whether the requestor is an authorized third party. Here, the third party's authentication key stored in the authentication unit 120 may be stored when the third party registers an account with the server 10 .

Further, if the requester is an authorized third party, a token allowing the requester to view the stored biometric information of the user may be issued and the issued token may be stored. In this case, the requestor, i.e., an authorized third party, receives the address information of the storage area where the biometric information is stored through the token stored in the authentication unit 120, and reads the user's biometric information stored in the server 10. can do.

Meanwhile, the token may have a valid time set, and may be automatically discarded when the set valid time elapses. Therefore, even an authorized third party can view the user's biometric information only after the token is reissued through an authentication process using the authentication key when the validity time elapses.

Meanwhile, the artificial intelligence unit 150 serves to process information based on artificial intelligence technology, and includes one or more modules that perform at least one of information learning, information reasoning, information perception, and natural language processing. can include

The artificial intelligence unit 150 uses machine running technology to generate a vast amount of information (big data, big data) such as information stored in the server 10 and information stored in an external storage capable of communicating with the server 10. At least one of learning, reasoning, and processing may be performed.

Here, learning may be performed through the machine learning technology. The machine learning technology is a technology that collects and learns large-scale information based on at least one algorithm, and determines and predicts information based on the learned information. Learning of information is an operation of identifying characteristics, rules, criteria, etc. of information, quantifying the relationship between information and predicting new data using quantified patterns.

The algorithms used by these machine learning technologies can be algorithms based on statistics, for example, decision trees that use the tree structure as a predictive model, artificial algorithms that mimic the structure and function of neural networks in living things. Neural networks, genetic programming based on biological evolution algorithms, clustering that distributes observed examples into subsets called clusters, Monte Carlo that calculates function values as probabilities through randomly extracted random numbers method (Montercarlo method) and the like.

As one field of the machine learning technology, the deep learning technology is a technology that performs at least one of learning, judgment, and processing of information using an artificial neural network algorithm. An artificial neural network may have a structure in which layers (hidden layers) are connected to each other and data is transferred between layers. This deep learning technology can learn a vast amount of information through an artificial neural network using a central processing unit (CPU) optimized for parallel computation.

Meanwhile, in this specification, the artificial intelligence unit 150 and the control unit 100 may be understood as the same component. In this case, the functions performed by the control unit 100 described in this specification can be expressed as being performed by the artificial intelligence unit 150, and the control unit 100 is named the artificial intelligence unit 150 or, conversely, the artificial intelligence unit 150. The intelligence unit 150 may also be referred to as the control unit 100.

Also, differently from this, in this specification, the artificial intelligence unit 150 and the control unit 100 may be understood as separate components. In this case, the artificial intelligence unit 150 and the control unit 100 may perform various controls on the server 10 through data exchange with each other. The control unit 100 may perform at least one function among functions executable in the server 10 or control at least one of the components of the server 10 based on the result derived from the artificial intelligence unit 150. can Furthermore, the artificial intelligence unit 150 may operate under the control of the control unit 100 .

Meanwhile, the memory 130 stores data supporting various functions of the server 10 . The memory 130 includes a plurality of application programs (application programs or applications) running in the server 10, data and instructions for the operation of the server 10, and for the operation of the artificial intelligence unit 150. Data (eg, at least one algorithm information for updating or learning a machine learning or predictive model) may be stored.

Also, information on at least one different biometric information measuring device 30 may be stored in the memory 130 . For example, in the memory 130, weight information corresponding to each characteristic of the different measurement devices 30 and information about hidden layers and nodes to which the corresponding weights are applied are in the form of a database (DB) (device characteristic DB 131). can be saved as Accordingly, different weights for each characteristic of the measuring device are applied to different hidden layers, and different initial prediction models can be generated for each measuring device used by the user, and learning is performed through the initial prediction model in which the characteristics of the measuring device are reflected. can

Here, the characteristics of the measurement device 30 may be a manufacturer of the measurement device 30, a method in which the measurement device is worn to measure biometric information, or a method in which the biometric information is measured. In this case, different weights to be applied and at least one hidden layer to which the weights are applied may be determined according to the manufacturer, wearing method, measurement method, and the like.

In this case, different weights and hidden layers may be determined for each characteristic of the measurement device 30 . That is, the controller 100 of the server 10 may primarily determine weights and hidden layers corresponding to the manufacturer of the input measuring device. In addition, other weights and hidden layers may be additionally determined (secondary) according to the wearing method of the input measuring device. In addition, other weights and hidden layers may be further determined (third order) according to the measurement method of the input measuring device. In addition, an initial prediction model may be generated by applying the weights determined in the first to third steps to the determined hidden layers. And the generated predictive model may be stored in the model data database (DB, Database) 132 .

In this case, if a user simultaneously uses a plurality of measuring devices (eg, first and second measuring devices), the control unit 100 of the server 10 corresponds to the characteristics of the first and second measuring devices, respectively. weights and hidden layers can be detected. In addition, an initial prediction model customized to device characteristics of the plurality of measurement devices may be generated by applying the detected weights to the detected hidden layers. In this case, if opposite weights are applied to a specific hidden layer, it goes without saying that the weights applied to the corresponding hidden layer may conflict due to the application of the weights. And the generated predictive model may be stored in the model data database (DB, Database) 132 .

To this end, the memory 130 of the server 10 may include a device characteristic DB 131 storing various characteristic information of different measurement devices in the form of a database. In addition, a model data DB 132 may be provided in which an initial prediction model generated according to the characteristics of the measuring device detected from the device characteristic DB 131 is stored. Here, the initial prediction model may be a prediction model that has not been learned, and may be gradually learned by the artificial intelligence unit 150.

Meanwhile, the memory 130 of the server 10 may further include a database (hereinafter referred to as biometric information DB 133 ) storing measured biometric information received through the communication unit 110 . The biometric information DB 133 may classify and store received biometric information according to the stored time or the type of stored biometric information, and address information about a region on the memory 130 in which each of the separated biometric information is stored, e.g. For example, pointer information can be stored. In this case, the pointer information may be transmitted through the communication unit 110 under the control of the controller 100 of the server 10 .

Meanwhile, the control unit 100 may control each connected component and control the overall operation of the server 10 .

When information of at least one measuring device 30 that measures user's biometric information is received through the communication unit 110, the control unit 100 obtains information of weights and hidden layers according to the characteristics of the received measuring device 30. It can be detected from the device characteristic DB 131. In addition, an initial predictive model customized for measurement devices used by the user may be generated based on the detected weights and information of the hidden layers. In this case, the initial prediction model may be an artificial neural network model having a plurality of hidden layers each having a plurality of nodes. The generated initial prediction model may be stored in the model data DB 132 as an initial prediction model.

When the user's biometric information measured by the measuring device 30 is received through the communication unit 110, the controller 100 can learn the initial prediction model based on the received biometric information. Such learning may be performed until a preset learning completion condition is satisfied.

For example, the learning completion condition may be satisfied through an accuracy test using the number of learning data, the learning time, or preset virtual biometric information.

Here, the accuracy test may be a test of comparing a predictive value calculated when the virtual biometric information of which predictive value is known in advance is input to the predictive model and the previously known predictive value. In this case, if the calculated predictive value is within an error range from the predicted predictive value corresponding to the previously known virtual biometric information, it may be determined that learning is completed. However, if this is not the case, it may be judged that the learning is not sufficient.

Meanwhile, when learning is completed, the controller 100 may input received biometric information to the learned prediction model and compare a predictive value output from the learned prediction model with a preset threshold value. And, according to the comparison result with the threshold value, the likelihood of developing or not developing a specific disease is predicted (positive prediction), and the predicted likelihood of developing or not developing (negative prediction) is predicted as a response to the biometric information. can transmit

Meanwhile, the controller 100 may change the threshold value based on user context information. In this case, when the threshold value increases, the true positive rate or true negative rate increases (false positive rate or false negative rate decreases), and when the threshold value decreases, the false positive rate or false negative rate may increase (true positive rate or true negative rate decreases). Therefore, the lower the threshold value, the higher the probability of outputting a prediction result by determining that a patient has a disease even when there is no disease.

Therefore, as a result of determining the user's situation based on the situation information, when the user is in a hospital or the like, a threshold value may be set to more sensitively predict whether or not an onset occurs according to the user's biometric information. In this case, the controller 100 may further increase the false positive rate by lowering the threshold value. Therefore, it can be predicted that the user's biometric information deteriorates even slightly, and the prediction result can be transmitted to the collection terminal 20 . Therefore, when an actual outbreak occurs, it can be dealt with more quickly, and an actual outbreak can be prevented in advance by taking measures before the actual invention.

Meanwhile, the controller 100 may update the trained prediction model when a preset update condition is satisfied. For example, the update condition may be satisfied when a preset update period expires (cycle-based update). Alternatively, it can be satisfied when a specific event occurs (event-based update). Here, the specific event may be a case in which new measurement device information is received or accuracy is lowered to a predetermined level or less as a result of a self-accuracy test.

Here, the accuracy test may be the same as or similar to the accuracy test for determining whether or not the learning has been completed. For example, the control unit 100 may periodically perform the self-accuracy test, and may determine whether to update the prediction model according to a result of performing the accuracy test.

Updating the predictive model can be done in a variety of ways. For example, at least a portion of each hidden layer constituting the predictive model and at least a portion of weights set in each node constituting the hidden layer may be changed to new values. In this case, since the set weights are predictive models, such an update method may correspond to replacing the predictive model with a new predictive model.

Here, new values of the weights may be weights of a new predictive model generated according to characteristics of new measurement device information. Alternatively, they may be weights of a predictive model generated in a process of learning the predictive model. That is, they may be weights of a predictive model before a recent learning result is reflected.

Alternatively, the update of the predictive model may be performed through a combination or ensemble according to a plurality of predictive models including the currently learned predictive model.

For example, the controller 100 may update the predictive model through a combination or ensemble between at least one new predictive model generated according to the characteristics of the new measurement device information and the currently learned predictive model. Alternatively, the predictive model may be updated through a combination or ensemble of a plurality of predictive models further including a predictive model generated in the process of learning the predictive model. In this case, the predictive model generated in the process of learning the predictive model may be a predictive model before a recent learning result is reflected.

Meanwhile, the updated predictive model may be trained until a predetermined learning completion condition is satisfied for the received biometric information. In this case, the control unit 100 may provide, as a response to the received biometric information, the predictive value calculated according to the previously learned prediction model and the prediction result according to the threshold until the learning of the updated prediction model is completed. there is. After the learning of the updated prediction model is completed, the predictive value calculated according to the updated prediction model and the prediction result according to the threshold may be provided as a response to the received biometric information.

Meanwhile, the controller 100 may transmit the address information when there is a request for address information indicating an address of a region where specific biometric information of a user is stored from the collection terminal 20 through the communication unit 110 . Here, the specific biometric information may be a specific type of biometric information stored at a specific time. Then, the address information can be transmitted to a terminal of an authorized third party through the collecting terminal 20 . When a request for biometric information corresponding to the address information is received from the authorized third party terminal, the controller 100 determines that the third party terminal stores the user's specific biometric information stored in the storage area corresponding to the address information. Reading may be permitted.

Meanwhile, FIG. 3 is a block diagram showing the configuration of the collection terminal 20 serving as a personal health gateway (PHG) of the health prediction system according to an embodiment of the present invention.

The collection terminal 20 includes a wireless communication unit 210, an input unit 220, an authentication unit 230, a sensing unit 240, an output unit 250, an interface unit 260, a memory 270, and a control unit 200. ) may be included. The components shown in FIG. 3 are not essential to implement the collection terminal 20, so the collection terminal 20 described in this specification may have more or fewer components than the components listed above. there is.

More specifically, among the components, the wireless communication unit 210 is between the collection terminal 20 and the wireless communication system, between the collection terminal 20 and at least one peripheral device, or between the collection terminal 20 and an external server. may include one or more modules enabling wireless communication of

The wireless communication unit 210 may include at least one of a wireless Internet module 211, a short-distance communication module 212, and a location information module 213.

The wireless Internet module 211 refers to a module for wireless Internet access, and may be built into or external to the collection terminal 20 . The wireless Internet module 211 is configured to transmit and receive radio signals in a communication network based on wireless Internet technologies.

Wireless Internet technologies include WLAN, Wi-Fi, Wi-Fi Direct, DLNA, WiBro, WiMAX, HSDPA, HSUPA, LTE, LTE-A, and 5G communication technologies, and the wireless Internet module 211 is Data may be transmitted and received according to at least one wireless Internet technology within a range including unlisted Internet technologies.

The short-range communication module 212 is for short-range communication, using at least one of Bluetooth, RFID, infrared communication, UWB, ZigBee, NFC, Wi-Fi, Wi-Fi Direct, and Wireless USB technology, It can support short-distance communication. The short-range communication module 212 may be used between the aggregation terminal 20 and a wireless communication system, between the aggregation terminal 20 and a peripheral device, or between the aggregation terminal 20 and an external server through wireless area networks. Wireless communication between located networks may be supported.

Here, the peripheral device is a wearable device capable of exchanging (or interlocking) data with the collection terminal 20 according to the present invention, for example, a smart watch or smart glass. ), or a head mounted display (HMD). The short-distance communication module 212 may detect (or recognize) a wearable device capable of communicating with the collection terminal 20 around the collection terminal 20 .

The location information module 213 is a module for acquiring the location (or current location) of the collection terminal 20, and representative examples thereof include a Global Positioning System (GPS) module or a Wireless Fidelity (WiFi) module. For example, the aggregation terminal may obtain the location of the aggregation terminal 20 using a signal transmitted from a GPS satellite by using a GPS module. As another example, when the aggregation terminal 20 utilizes the Wi-Fi module, the location of the aggregation terminal 20 is determined based on information of the Wi-Fi module and a wireless access point (AP) that transmits or receives a wireless signal. can be obtained. If necessary, the location information module 213 may perform the function of any one of the other modules of the wireless communication unit 210 to obtain data on the location of the collection terminal 20 in substitution or addition. The location information module 213 is a module used to acquire the location (or current location) of the collection terminal 20, and is not limited to a module that directly calculates or acquires the location of the collection terminal 20.

The input unit 220 includes a camera 221 or video input unit for inputting a video signal, a microphone 222 for inputting an audio signal, or a user input unit 223 for receiving information from a user, for example , a touch key, a push key (mechanical key, etc.). Voice data or image data collected by the input unit 220 may be analyzed and processed as a user's control command.

In addition, the authentication unit 230 may store and authenticate various types of information for authenticating the right to read biometric information stored in the server 10 . For example, the authentication unit 230 may store address information transmitted from the server 10 according to a request of an authorized third party. The address information may be pointer information indicating address information corresponding to a region of the memory 130 of the server 10 in which the user's specific biometric information is stored.

In addition, identification information for identifying a user and authenticating usage rights may be stored, and a user identification module (UIM), a subscriber identity module (SIM), and a universal subscriber identity module (universal subscriber identity module) may be stored. ; USIM) and the like.

The sensing unit 240 may include one or more sensors for sensing at least one of information within the collection terminal 20 , surrounding environment information surrounding the collection terminal 20 , and user information. For example, the sensing unit 240 may include a proximity sensor 241, an illumination sensor 242, a touch sensor, an acceleration sensor, a magnetic sensor, and gravity. Sensors (G-sensor), gyroscope sensor, motion sensor, RGB sensor, infrared sensor (IR sensor), ultrasonic sensor, optical sensor (e.g. For example, cameras (see 221), microphones (see 222), battery gauges, environmental sensors (eg, barometers, hygrometers, thermometers, radiation sensors, heat sensors, gas sensors, etc.) , chemical sensors (eg, an electronic nose, a healthcare sensor, a biometric sensor, etc.). Meanwhile, the collection terminal 20 disclosed in this specification may combine and utilize information sensed by at least two or more of these sensors.

The output unit 250 is for generating an output related to sight, hearing, or touch, and may include a display unit 251 and a sound output unit 252. The display unit 251 may implement a touch screen by forming a mutual layer structure or integrally with the touch sensor. Such a touch screen may function as a user input unit 223 providing an input interface between the collection terminal 20 and the user, and may provide an output interface between the collection terminal 20 and the user.

The interface unit 260 serves as a passage for various types of external devices connected to the collection terminal 20 . The interface unit 260 connects a device having a wired/wireless headset port, an external charger port, a wired/wireless data port, a memory card port, and an identification module. It may include at least one of a port, an audio input/output (I/O) port, a video input/output (I/O) port, and an earphone port. In response to an external device being connected to the interface unit 260, the collection terminal 20 may perform appropriate control related to the connected external device.

In addition, the memory 270 stores data supporting various functions of the collection terminal 20 . The memory 270 may store a plurality of application programs (applications) running in the collection terminal 20 , data for operation of the collection terminal 20 , and commands. At least some of these application programs may be downloaded from an external server through wireless communication. In addition, at least some of these application programs may exist on the collection terminal 20 from the time of shipment for basic functions of the collection terminal 20 (eg, measurement information collection and transmission functions, output of received prediction results). . Meanwhile, the application program may be stored in the memory 270, installed on the collection terminal 20, and driven by the control unit 200 to perform an operation (or function) of the collection terminal 20.

Meanwhile, the memory 270 may store information of different measurement devices that may be connected to the collection terminal 20 . For example, the memory 270 may store identification information of measurement devices manufactured by a plurality of different manufacturers and information about characteristics of each measurement device corresponding to the identification information. Here, the characteristics of the measuring device may include manufacturer information, the type of biometric information measured by the measuring device, and the shape of the measuring device (eg, forearm type or watch type in the case of a blood pressure monitor). In addition, the method in which the measuring device is attached to the user to measure biometric information (hereafter, mounting method, e.g., contact, adhesive, pressurized, etc.) invasive or non-invasive, continuous measurement or single measurement, etc.).

Therefore, when a measuring device is connected, the collection terminal 20 can identify various characteristic information of the measuring device by identifying the connected measuring device. In addition, characteristic information of the identified measuring device may be transmitted to the server 10 . Accordingly, the server 10 determines weights to be set for the hidden layer and the nodes of the hidden layer from the device characteristic DB 131 based on the characteristic information of the measuring device, and generates an initial prediction model corresponding to the characteristics of the received measuring device. can do.

On the other hand, if the measurement device is not identified based on the identification information received from the measurement device or if identification information is not received from the measurement device, the control unit 200 of the collection terminal 20 provides characteristic information of the currently connected measurement device from the user. can be entered directly. In this case, the user may directly input measurement device characteristic information including at least one of the manufacturer of the measurement device, the type of biometric information to be measured, the shape of the measurement device, a mounting method, and a measurement method. Then, the collection terminal 20 may transmit the collected measurement device characteristic information to the server 10 .

Meanwhile, the control unit 200 controls overall operations of the collection terminal 20 in addition to operations related to the application programs. The control unit 200 may provide or process appropriate information or functions to the user by processing signals, data, information, etc. input or output through the components described above or by running an application program stored in the memory 170.

In addition, the controller 200 may control at least some of the components described in FIG. 3 to drive an application program stored in the memory 270 . Furthermore, the controller 200 may combine and operate at least two or more of the components included in the collection terminal 20 to drive the application program.

Meanwhile, the control unit 200 may perform a communication connection with at least one measuring device 30 . Here, the at least one measurement device 30 may be connected through the short-range communication module 212 . More preferably, the at least one measurement device may be connected in a Bluetooth Low Energy (BLE) communication method. In addition, the user's biometric information received from the at least one connected measuring device 30 may be received and the received biometric information may be transmitted to an external server, that is, the server 10 connected through the wireless Internet module 211. In response to the transmitted biometric information, a diagnostic prediction result for the user's health condition, that is, a prediction result for the probability of developing at least one predetermined disease is received from the server 10, and the received prediction result is displayed. It may be output through at least one of the unit 251 and the sound output unit 252 .

Also, the controller 200 may perform a communication connection with at least one peripheral device 21 . The context of the user may be analyzed based on at least one connected peripheral device, a detection result of the sensing unit 240, or location information of the location information module 213. In addition, information of the analyzed situation, that is, context information, may be transmitted to the server 10 separately from the measured biometric information. Then, the server 10 may perform diagnostic prediction on the health state of the user based on the received situation information.

In addition, the controller 200 receives a biometric information viewing request from an authorized third party, that is, a third party having a token allowing viewing of the user's biometric information, so that specific biometric information of the user from the server 10 When address information of the stored storage area is received, it may be stored in the authentication unit 230 . When receiving a request for the address information from the authorized third party, the stored address information may be transmitted to the terminal of the third party in response to the request.

Meanwhile, the controller 200 may process biometric information received from at least one measuring device 30 into information in a form requested by the server 10 . For example, the controller 200 may convert biometric information received from at least one measurement device 30 into information in a format requested by the server 10 . Alternatively, the controller 200 may transmit secondary biometric information calculated from primary biometric information received from at least one measuring device 30 to the server 10 .

For example, the controller 200 may calculate the oxygen saturation level and calorie consumption of the user based on the number of steps and movement received from biometric information received from at least one measuring device 30 . In addition, information such as the calculated oxygen saturation level and calorie consumption may be transmitted to the server 10 as biometric information received from the at least one measurement device 30 . Then, the server 10 may input the secondary biometric information to the pre-learned prediction model. That is, the controller 200 converts the biometric information received from the at least one measuring device or calculates other biometric information from the biometric information received from the at least one measuring device, so that the at least one measuring device 30 The collected biometric information can be processed into a form of biometric information that can be input to the pre-learned prediction model. The processed biometric information may be transmitted to the server 10 .

4 is a block diagram showing the configuration of the health information measuring device 30 of the health prediction system according to an embodiment of the present invention.

Referring to FIG. 4 , in the health prediction system according to an embodiment of the present invention, the measuring device 30 for measuring the user's biometric information is connected to the control unit 300 and the control unit 300, and the control unit 300 ) may be provided with a communication unit 310, a measurement unit 320, a memory 330 and an interface unit 340 controlled by. In addition, although not shown, a display unit and a sound output unit may be further included. The components shown in FIG. 4 are not essential to implement the measuring device 30, so the measuring device 30 described herein may have more or fewer components than the components listed above. there is.

First, the communication unit 310 may include at least one communication module for wireless communication with the collection terminal 20 . For example, the communication unit 310 may include at least one module to which short-distance communication technology is applied. More preferably, the communication unit 310 may be a communication module for Bluetooth low energy communication (BLE).

Meanwhile, the communication unit 310 may further include a wireless Internet module enabling wireless Internet access. In this case, if the wireless Internet module is used, the measuring device 30 can be directly connected to the server 10 without going through the collection terminal 20, and the server 10 can be measured through the measuring device 30. The user's biometric information may be transmitted.

The measurement unit 320 may include at least one sensor for detecting a specific type of user biometric information. For example, the measuring unit 320 may include a blood glucose sensor that detects blood sugar of the user by a measurement method such as invasiveness, or a sensor such as a blood pressure sensor, a pulse sensor, or a body temperature sensor that measures body temperature. Alternatively, when the measuring device 30 is a weight scale or height gauge, it may include a weight sensor for detecting the user's weight or a length sensor for detecting the user's height.

Characteristics of the measurement device may be determined according to the type of sensor provided in the measurement unit 320 . For example, a measurement method (eg, an invasive method) in which the measurement device 30 measures the user's biometric information may be determined according to a sensor provided in the measurement device 30 . Also, a wearing method for the user to wear the measurement device 30 may be determined according to a method for fixing the sensor to the user for the measurement.

The memory 330 stores data supporting functions of the measuring device 30 . The memory 330 may store application programs driven by the measurement device 30 , data for operation of the measurement device 30 , and commands. In addition, the measurement result of the measurement unit 320 may be stored.

The interface unit 340 serves as a passage for various types of external devices connected to the measurement device 30 . The controller 300 may perform appropriate control related to the connected external device in response to an external device being connected to the interface unit 340 .

Also, the control unit 300 controls overall operations of the measurement device 30 in general. The control unit 300 processes signals, data, information, etc. input or output through the components described above or runs an application program stored in the memory 330 to display information measured by the measurement unit 320 or to display information measured by the measurement unit 320. Information may be provided to the server 10 or the collection terminal 20 .

For example, the control unit 300 may detect the user's biometric information by controlling the measurement unit 320 according to a preset period. Alternatively, if there is a request from the collection terminal 20, the measurement unit 320 may be controlled to detect the user's biometric information. Also, the detected biometric information may be transmitted to the collection terminal 20 . Alternatively, when direct wireless communication with the server 10 is possible, the measured biometric information may be directly transmitted to the server 10 .

Hereinafter, embodiments related to a control method that can be implemented in the health prediction system composed of the server 10, the collection terminal 20, and the measuring device 30 will be described with reference to the accompanying drawings. It is apparent to those skilled in the art that the present invention can be embodied in other specific forms without departing from the spirit and essential characteristics of the present invention.

In the following description, the server 10 is a server 10 that can be connected through a cloud network, and will be described on the assumption that it is a cloud server 10 for convenience.

5 is a flowchart illustrating an operation process in which a health prediction system according to an embodiment of the present invention learns a predictive model for diagnosing a user's health state according to characteristics of a measuring device used by the user. 6A to 6B are exemplary diagrams for explaining characteristics of a measuring device. 7 is an exemplary diagram illustrating an example of a predictive model for diagnosing a health state of a user in a health prediction system according to an embodiment of the present invention.

First, referring to FIG. 5 , the collection terminal 20 of the health prediction system according to an embodiment of the present invention may first transmit information of at least one connected measuring device 30 to the cloud server 10 . Here, the information on the measurement device 30 may be identification information of the measurement device 30 or information on characteristics of the measurement device 30 .

For example, when the measuring device 30 is connected, the collection terminal 20 detects the characteristic information of the measuring device from the device information DB 271 based on the identification information received from the connected measuring device to detect the characteristic information of the measuring device. can be obtained. Alternatively, the collection terminal 20 may acquire characteristic information of the connected measuring device based on a user's input for the connected measuring device (S500). In addition, the obtained measurement device characteristic information may be transmitted to the cloud server 10 (S502).

Here, the characteristic information of the measuring device may include a quantified characteristic of the clinical accuracy of the measuring device, that is, the quantified clinical accuracy. For example, when the measuring device is a blood glucose meter, it may include characteristic information according to the Clarke Error Grid shown in FIG. 6A.

The Clark error grid is intended to quantify the clinical accuracy of the patient's estimate of current blood glucose compared to the blood glucose value obtained from the meter. As quantified characteristic information, each area is an error from the measured value of the reference sensor (area A), an area that is out of the error range but does not lead to inappropriate treatment (area B), an area that leads to unnecessary treatment (area C), hypoglycemia or It shows a region where detection of hypertension fails (region D) and a region in which hypoglycemic treatment and hyperglycemic treatment are confused (region E). In addition, the state in which blood glucose estimation clinical test results of a specific measuring device (blood glucose meter) are distributed in each region of the Clark error grid may represent quantified clinical accuracy of the measuring device. In addition, the quantified clinical accuracy of the measurement device (blood glucose meter) analyzed through the Clark error grid may be detected as characteristic information of the blood glucose meter.

Meanwhile, the characteristic information of the measurement device may include information about a wearing method and a measurement method of the measurement device.

For example, in the case of a blood glucose meter, there may be an attachment type method as shown in (a) of FIG. 6b and a contact type method as shown in (b) of FIG. 6b. In this case, the blood glucose meters shown in (a) and (b) of FIG. 6B may each have different characteristic information with respect to a wearing method. On the other hand, as shown in (a) and (b) of FIG. 6B, blood glucose meters that measure blood sugar in an invasive manner may have the same characteristic information about the measurement method.

Meanwhile, the cloud server 10 that has received the characteristic information of the measurement device from the collection terminal 20 may detect weights of hidden layers and nodes corresponding to each of the received characteristic information of the measurement device from the device characteristic DB 131. In addition, an initial prediction model in which the received characteristic information of the measurement device is reflected may be generated based on the weights of the detected hidden layer and nodes (S504).

Here, the initial prediction model may be a model for predicting the likelihood of developing at least one specific disease. The at least one specific disease (hereinafter referred to as specific disease) may be a disease corresponding to the biometric information obtained in step S504. That is, a predictive model for predicting the possibility of developing a specific disease may be determined based on the biometric information measured by the measurement device. In this case, the predictive model may be a model capable of predicting the possibility of onset of the at least one disease based on input biometric information.

Alternatively, the specific disease may be a disease directly selected by the user. For example, if there are multiple diseases related to the biometric information acquired in step S504, a predictive model for predicting the onset probability of at least one disease selected by the user among the diseases may be determined.

Then, the cloud server 10 sets the weights of the hidden layer and the node corresponding to each characteristic information of the measuring device with respect to the determined predictive model for the specific disease, so that the initial value in which the characteristic information of the measuring device is reflected for the specific disease A predictive model may be generated (S504).

The initial prediction model may be an artificial neural network model including a plurality of hidden layers and weights assigned to nodes included in each of the plurality of hidden layers. Such an artificial neural network model may be a model designed in which the number of hidden layers passed through is different from each other according to input information input through an input layer. 7 illustrates an example of an artificial neural network model (prediction model) according to an embodiment of the present invention.

Referring to FIG. 7 , the prediction model of the health prediction system according to an embodiment of the present invention may include at least one exclusive layer 710 and at least one common layer 720 and 730 . Here, the exclusive layer 710 may be a hidden layer including weights applied only to inputs classified into specific groups. Also, the common layers 720 and 730 may be hidden layers including weights applied to inputs of all groups.

That is, as shown in FIG. 7 , when inputs classified as group A are applied, weights according to the exclusive layer 710 may be assigned via the exclusive layer 710 . In addition, values to which weights of respective nodes set in the exclusive layer 710 are applied may be input to the common layers 720 and 730 . That is, both the weights assigned by the exclusive layer 710 and the weights assigned by the common layers 720 and 730 may be applied to the inputs classified into the group A.

On the other hand, when inputs classified as group B are applied, they may be directly input to the common layers 720 and 730 without passing through the exclusive layer 710 . That is, only the weights assigned to the common layers 720 and 730 may be applied to the inputs classified into the group B in a state in which the weights assigned to the exclusive layer 710 are not applied.

Here, input information input to the predictive model may be biometric information. Also, the group of input information may be a characteristic of a measuring device that measures the biometric information. That is, the cloud server 10 classifies biometric information received from the measurement device into a specific group according to the characteristics of the received measurement device, and sets an exclusive layer 710 corresponding to the specific group or an exclusive layer 710. ) may not be set. Accordingly, biometric information received from measurement devices classified into a specific group is input to the common layers 720 and 730 via the exclusive layer 710 in which weights corresponding to the specific group are set, and thus collected from the collection terminal 20. A predictive model reflecting the characteristics of the measured instrument can be created.

Meanwhile, the collection terminal 20 may receive biometric information detected by the measuring device from the connected measuring device (S506). Then, the received biometric information may be transmitted to the cloud server 10 (S508).

Then, the cloud server 10 may determine whether learning of the initial prediction model generated in step S504 is completed (S510). Here, whether or not the learning is completed may be determined in various ways. For example, the cloud server 10 may determine whether the learning has been completed based on whether learning has been performed on a sufficient number of learning data (biometric information). Alternatively, the cloud server 10 may determine whether or not the learning has been completed based on a period in which the learning has been completed. Alternatively, the cloud server 10 may determine whether or not the learning has been completed through an accuracy test using biometric information of which the predictive value is known in advance. In this case, the cloud server 10 inputs biometric information of which the predictive value is known in advance to the predictive model currently being learned, and learns based on the error between the predictive value calculated from the predictive model being learned and the previously known predictive value. completion can be determined.

As a result of determining that learning is complete in step S510, if learning is not complete, the cloud server 10 may perform learning for a predictive model based on the currently received biometric information (S520). Then, the collection terminal 20 determines whether the preset biometric information collection period has expired (S522), and if the biometric information collection period has expired, proceeds to step S506 again to measure the biometric information measured by the measuring device. can be received from the device. In this case, the process from step S506 to step S510 may be performed again.

On the other hand, if it is determined that the learning of the predictive model is completed as a result of the determination in step S510, the cloud server 10 may input the biometric information received in step S504 into the learned predictive model. Then, a predictive value corresponding to the biometric information may be calculated through the learned predictive model (S512).

Meanwhile, as described above, the predictive model may be a predictive model for predicting the possibility of developing a specific disease. Therefore, the predictive value calculated by the prediction model may be the possibility of developing a specific disease. Then, the cloud server 10 may determine the prediction result of the specific disease, that is, whether or not the disease occurs, based on the possibility of the disease. In this case, the cloud server 10 may predict the onset of the specific disease according to whether the predictive value exceeds the threshold based on a preset threshold (S514). In addition, the predicted result, that is, whether or not the specific disease occurs can be transmitted to the collection terminal 20 (S516).

Then, the collection terminal 20 may output the received prediction result, that is, whether or not a specific disease occurs through the output unit 250 (S518). In this case, the collection terminal 20 may output whether or not the disease occurs through at least one of visual information displayed through the display unit 251 and sound information output through the audio output unit 252 .

Meanwhile, the collection terminal 20 may determine whether the preset biometric information collection period has expired (S522), and if the biometric information collection period has expired, proceed to step S506 of receiving biometric information from the measuring device again. Accordingly, the process from step S506 to step S518 may be performed again.

Meanwhile, the cloud server 10 of the health prediction system according to an embodiment of the present invention may determine a threshold value for determining whether the specific disease occurs based on the user's situation. In this case, if the threshold value is different, even if the predictive value is the same, whether or not the specific disease occurs may be determined differently. That is, when the predictive value calculated through the learned prediction model is 50% and the threshold value is 30%, the cloud server 10 determines that the user has a specific disease and sends the predicted result to the collection terminal 20. can transmit On the other hand, if the threshold value is 70%, the cloud server 10 may determine that the user does not have a specific disease and transmit the prediction result to the collection terminal 20 .

The cloud server 10 may set the threshold differently based on the user's circumstances. For example, when an urgent action is required even if the user's health condition deteriorates even slightly, such as when the user is admitted to a hospital, the cloud server 10 may set the threshold lower than a normal case. Then, even if the predictive value is lower than the normal case, the prediction result is displayed as the occurrence of the specific disease, so that measures for the specific disease can be quickly received. That is, it is possible to more sensitively detect the deterioration of the user's health condition, and accordingly, to receive more urgent medical treatment according to the deterioration of the health condition or to provide medical services corresponding to the deteriorating health condition before the specific disease actually develops. By receiving it, it is possible to prevent the occurrence of the specific disease in advance.

FIG. 8 is a flowchart illustrating an operation process in which a result of predicting whether or not to have a specific disease by reflecting a user's situation is predicted in the health prediction system according to an embodiment of the present invention. 9 is an exemplary diagram illustrating an example in which a threshold value is set differently according to a detected user's situation in the health prediction system according to an embodiment of the present invention.

First, referring to FIG. 8 , the collection terminal 20 may determine a prediction target for predicting an outbreak (S800). Here, the predicted target may be a specific disease or specific condition. An initial threshold may be determined according to the determined prediction target.

The prediction target may be directly selected by the user. For example, when a user inputs a specific disease through the collection terminal 20, a prediction target may be determined as the specific disease. Alternatively, the prediction target may be automatically set to a specific disease corresponding to a currently set prediction model.

Meanwhile, according to the above description, the specific disease may be multiple. That is, the prediction model learned in the cloud server 10 may be a model capable of predicting the onset of a plurality of diseases based on input biometric information. In this case, different threshold values may be set for each disease, and whether or not each disease occurs may be determined based on the different threshold values set for each disease and the predictive value calculated from the prediction model. In this case, the prediction target may mean any one disease among a plurality of diseases that can be predicted by the prediction model.

When the prediction target is determined in step S600, the collection terminal 20 may analyze the current situation of the user (S602). To this end, the collection terminal 20 may be connected to at least one peripheral device, and may collect information related to a user's situation from the connected peripheral devices.

For example, the collection terminal 20 may collect at least one location information related to its own location. For example, the collection terminal 20 may detect a location calculated from the location information module 213 . Alternatively, the collection terminal 20 may collect information about the current location from an AP installed in the area where it is located. In addition, the user's situation can be inferred based on the collected location information.

Alternatively, the collection terminal 20 may collect information on at least one mobile terminal located nearby. For example, the collection terminal 20 may compare identification information of at least one mobile terminal located around itself with identification information of mobile terminals of pre-registered acquaintances. In addition, if at least one of the identification information of pre-registered acquaintances is located in the vicinity, the user's situation may be inferred based on the at least one pre-registered acquaintance. For example, the collection terminal 20 may infer that the user is located near family members based on mobile terminal identification information of family members located near the user.

Meanwhile, when the user's surrounding situation and location are detected, the collection terminal 20 may analyze whether the user is in a preset situation. For example, the collection terminal 20 may determine whether the user is in a hospital as a result of analyzing the user's location. In addition, when the user is located in the hospital, it may be detected that the user is in the preset situation. In this way, when it is determined that the user is in a preset situation, the collection terminal 20 may generate information about the user's situation, that is, situation information.

Meanwhile, the collecting terminal 20 may receive the user's biometric information measured by the connected measuring device according to a preset period or according to the user's request, independently of the information on the user's surroundings (S804). Then, when biometric information is received in step S804, the received biometric information, the situation information, and information on the currently selected prediction target may be transmitted to the cloud server 10 (S806).

Then, the cloud server 10 may first input the received biometric information into the learned prediction model to calculate a predictive value. A threshold value of a disease corresponding to the prediction target may be determined from among threshold values corresponding to each of diseases predictable by the prediction model.

Meanwhile, the cloud server 10 may change the determined threshold based on the situation information received from the collection terminal 20 (S808). For example, when the situation of the user corresponding to the conditional information received from the collection terminal 20 is a preset situation (eg, when the user is in a hospital), the cloud server 10 determines the severity of the disease corresponding to the prediction target. The threshold value may be changed by a preset displacement. To this end, the memory 130 of the cloud server 10 may include situation information corresponding to a plurality of different situations and information about change displacement of different threshold values corresponding to each of the plurality of situation information. For example, when the user is in a general ward, in an emergency ward, and at home, the change displacement of the threshold value for each situation may be different.

Meanwhile, when the threshold value is determined according to the predicted object and the analyzed situation in step S808, the cloud server 10 may calculate a predictive value corresponding to the received biometric information according to the learned prediction model (S810). In addition, based on the calculated predictive value in step S808 of the user's current situation and the threshold value determined according to the prediction target, an onset or non-onset prediction result of the disease to be predicted can be derived (S812). Then, the derived prediction result may be transmitted to the collection terminal 20 (S814). Then, the collection terminal 20 may output the received prediction result through the output unit 250 (S816).

9, when the predictive value is 0 (0%) to 1.0 (100%), the positive predictive value (PR) and the negative predictive value (PR) determined according to each predictive value and a threshold (threshold) It is a diagram showing the correlation of Negative, NR).

Referring to FIG. 9 , it can be seen that the negative predictive value decreases when the positive predictive value increases, and the negative predictive value increases when the positive predictive value decreases. Here, the positive predictive value is a predictive degree indicating that the user has developed a disease to be predicted, and the negative predictive degree is a predictive degree indicating that the user has not developed a disease to be predicted. That is, the positive predictive value and the negative predictive value may have values that are opposite to each other.

Also, when a specific threshold is set, a positive predictive value criterion and a negative predictive value criterion corresponding to the set threshold may be determined. That is, as the threshold value increases, the positive predictive value criterion, that is, the reference value of the predictive degree for determining whether or not the disease to be predicted increases, and conversely, the prediction to determine whether the disease to be predicted does not occur. The reference value of the degree (negative predictive degree) may be lowered. Conversely, as the threshold value decreases, the reference value of the predictive value (positive predictive value) for determining whether or not the disease to be predicted has occurred, and conversely, the predictive value for determining whether the disease to be predicted has not occurred ( The reference value of the negative predictive value) may be increased.

In this state, when the user is located in a hospital, the cloud server 10 may lower the threshold value by a preset change displacement. Therefore, as shown in FIG. 9 , the threshold value 900 may be changed from the first value 901 to the second value 902 .

When the threshold value 900 is changed in this way, the positive predictive value reference value corresponding to the changed threshold value 900 may be lowered. Conversely, the reference value of the voice predictive value corresponding to the changed threshold value 900 may be higher. Therefore, when the predictive value is calculated from the learned predictive model, the probability of being judged positive, that is, the probability of being judged as having the disease may be higher, and the probability of being judged as not having the disease may be lower. Therefore, it can be more easily determined that the user has a disease corresponding to the prediction target according to the prediction result determined through the received biometric information.

Meanwhile, when it is determined that a disease corresponding to the predicted target has occurred, since the user is in a hospital, the user can receive medical services for the predicted target disease. Therefore, when a user is located in a hospital or the like, medical services for the predicted target disease can be provided more sensitively according to changes in health status according to biometric information. In addition, even if it is a predictive degree that is not judged to have an onset in a normal case, it is determined to be onset by a change in the threshold value and medical service can be provided. Therefore, the present invention has an effect of preventing the onset of the predicted disease in advance by enabling the user to receive medical services before the onset of the predicted disease.

On the other hand, since the health prediction system according to an embodiment of the present invention uses a prediction model learned from the user, the user may input virtual information to simulate a change in the user's health state according to the input virtual information in advance. Of course there is.

For example, if the health prediction system according to an embodiment of the present invention predicts the user's health condition according to blood sugar detected through a blood glucose meter, the user inputs information on food that the user has not yet consumed, thereby providing the user with the food. When the blood sugar is ingested, the change in blood sugar of the user can be predicted and the predicted result can be provided to the user.

10 and 11 show such an example.

Referring first to FIG. 10, FIG. 10 is a flowchart illustrating an operation process of learning a change in blood sugar according to a user's diet through a health prediction system according to an embodiment of the present invention.

In this case, the collection terminal 20 may receive information on the diet consumed by the user from the user (S1000). In addition, information on the diet input by the user may be transmitted to the cloud server 10 (S1002). Here, the diet may include information on the name of at least one food and an intake amount of the food (eg, weight or quantity, for example, 1 serving, 1.5 servings, or 2 servings).

Then, the cloud server 10 may analyze nutrients of each food included in the received diet information. Then, based on the analyzed nutrients, it is possible to analyze the nutrients corresponding to the input diet and the amount of each nutrient consumed by the user (S1004).

In step S1004, the cloud server 10 may detect nutrients corresponding to each type of food included in the diet. To this end, the memory 130 of the cloud server 10 contains nutritional analysis information for each of a plurality of different foods, that is, nutrients (eg, carbohydrates, proteins, ...) included in each food and each nutrient for the reference weight. It may include information about the composition ratio of

Further, the cloud server 10 may calculate the amount of each nutrient consumed by the user based on the amount of each food consumed. For example, if the intake input by the user is weight, the amount of nutrients contained in each food consumed by the user may be calculated by reflecting the difference in ratio between the standard weight and the intake weight in the nutritional analysis result of each food. In addition, if the intake input by the user is a portion (intake), the weight of the food corresponding to the intake is calculated based on the weight of the standard portion (one serving), and the difference in ratio between the calculated weight of the food and the standard weight Based on this, it is possible to calculate the amount of nutrients contained in each food consumed by the user. In addition, nutrients consumed by the user and the amount of the nutrients may be calculated with respect to the entire diet by summing the intakes of nutrients of each food.

Meanwhile, when a user consumes food and a predetermined time elapses, the measuring device may measure the user's blood sugar level. The blood glucose measured by the measuring device is transmitted to the collection terminal 20, and the collection terminal 20 can receive it (S1006). Then, the collection terminal 20 may transmit the received blood glucose measurement information to the cloud server 10 (S1008).

Upon receiving the blood glucose measurement information, the cloud server 10 compares the previously received blood glucose measurement result, that is, the blood glucose measurement result measured before the user took food, and the blood glucose measurement result received in step S1008 to determine the blood glucose level. The amount of change can be calculated (S1010). In this case, the change in blood sugar may be the change in blood sugar with respect to a diet currently consumed by the user.

In addition, the cloud server 10 may detect nutrients of the previously collected diet and a change in blood glucose measured for the diet (S1012). And based on the changes in nutrients and blood sugar of the previous diet detected in step S1012 and the changes in nutrients and blood sugar in the diet currently consumed by the user measured in step S1010, the amount of change in blood sugar according to the change in nutrients is calculated. It can (S1014). That is, the cloud server 10 detects the difference between the nutrients in the previous diet and the nutrients in the current diet consumed by the user, and determines the change in blood sugar detected with respect to the nutrients in the previous diet and the nutrients in the current diet consumed by the user. A difference in the amount of change in the detected blood sugar may be detected. Accordingly, it is possible to detect a difference in blood glucose change amount for a difference in nutrients ingested by the user.

Then, the cloud server 10 may learn a blood glucose change prediction model for each nutrient based on the difference in blood glucose change according to the calculated difference in nutrients (S1016).

This learning process may be repeated more than a predetermined number of times, and when the repetition is completed more than the predetermined number of times, the cloud server 10 may determine that the learning of the predictive model of the blood glucose change amount for each nutrient has been completed. Alternatively, the cloud server 10 determines that the learning of the blood glucose change prediction model for each nutrient has been completed, when the magnitude of the change in blood sugar changed according to the learning result of step S1016 is less than a preset level for at least one nutrient. can That is, when it is difficult to expect the effect of learning any longer, the cloud server 10 may determine that the learning of the blood glucose change prediction model for each nutrient has been completed.

Meanwhile, FIG. 11 is a flowchart illustrating an operation process of providing a result of predicting a change in blood sugar according to a virtual diet input by a user through the prediction model learned in FIG. .

Referring to FIG. 11 , the user may input virtual diet information through the collection terminal 20 according to an embodiment of the present invention (S1100). In this case, the virtual menu information is information on a menu that the user has not consumed, and may be information on a menu including foods to be consumed by the user in the future. In addition, the collection terminal 20 may transmit the virtual diet information input by the user to the cloud server 10 (S1102). Here, the virtual diet information may include information on virtual intake.

Meanwhile, the cloud server 10 may analyze nutrients of each food included in the received virtual diet (S1104). In this case, nutrients for each food may be analyzed, and based on the virtual intake, intake of each nutrient for each food may be analyzed.

Then, the cloud server 10 may calculate the blood glucose change amount corresponding to each nutrient through the blood glucose change amount prediction model for each nutrient learned in FIG. 10 (S1106). In this case, the blood glucose change amount calculated through the learned blood glucose change amount prediction model for each nutrient may be the blood sugar change amount according to the reference weight of each nutrient. Therefore, the cloud server 10 may calculate a blood sugar change corresponding to the intake amount of the nutrient according to the calculated amount of blood sugar change according to the reference weight for each nutrient and the intake amount of the nutrient analyzed in step S1104.

In addition, the cloud server 10 may sum the blood glucose change amount corresponding to the intake amount of each nutrient for each nutrient. The summed result may be predicted as a change in blood sugar corresponding to the virtual diet received in step S1102 (S1108).

The predicted change in blood sugar may be provided to the collection terminal 20 in response to the virtual diet information received in step S1102 (S1110). Then, the collection terminal 20 may output the received information on the predicted change in blood glucose through the output unit 250 (S1112). In this case, the information of the input virtual diet and the information of the predicted change in blood sugar may be displayed on the display unit 251 together.

Meanwhile, the health prediction system according to an embodiment of the present invention may recommend an optimal diet to the user based on the learned model for predicting changes in blood sugar for each nutrient. In this case, the learned blood glucose change prediction model for each nutrient may be a blood glucose change characteristic for each nutrient of the user learned according to the user's blood glucose measurement result. Accordingly, the cloud server 10 may recommend an optimal diet based on the learned user's blood glucose change characteristics for each nutrient.

For example, the cloud server 10 may detect at least one other user having similar biological characteristics to the user and recommend the user's diet based on diet information input by the detected at least one other user (collaborate filtering method). That is, the cloud server 10 may compare each user's blood glucose change prediction model for each nutrient, and recommend the user's diet based on diet information input by at least one other user having a similar blood sugar change prediction model for each nutrient. there is.

In this case, the cloud server 10 may recommend the menu information to the user if the user has a similar taste among the menu information input by the at least one other user (content based filtering method). In this case, whether the tastes are similar, it may be determined that the tastes are similar to each other when the types of foods included in the diet match more than a predetermined ratio.

Meanwhile, when a recommended menu is determined based on menu information input by at least one other user, the cloud server 10 may provide information on the recommended menu to the collection terminal 20 . In this case, the cloud server 10 provides each food constituting the recommended diet, the result of predicting the change in blood sugar for each food calculated according to the nutrient analysis result of the food, and the glycemic index (GI) information corresponding to each food. can provide

Here, the glycemic index information (GI) is a value obtained by comparing the level of blood sugar rise after consuming a certain amount of the food with the level of blood sugar rise after consuming the same amount of standard carbohydrate food. It may mean an index representing the degree of increase.

[Table 1] below shows examples of recommended diet information provided from the cloud server 10.

[Table 1]

Meanwhile, the cloud server 10 of the health prediction system according to an embodiment of the present invention may update the learned prediction model when a preset update condition is satisfied.

In this case, the preset update condition may be satisfied when the preset update cycle expires. Alternatively, when a measurement device is added, changed, or removed, the preset update condition may be satisfied.

Alternatively, when accuracy is out of the error range as a result of a self-accuracy test according to preset biometric information of which predictability is known in advance, the update condition may be satisfied. This is because the accuracy of the learned predictive model may decrease over time.

12A to 12C are conceptual diagrams for explaining methods of updating a prediction model in the health prediction system according to an embodiment of the present invention.

Referring first to FIG. 12A, (a) of FIG. 12A illustrates an example of a predictive model before being updated, and (b) of FIG. 12A illustrates an example of a predictive model after being updated. As shown in FIG. 12A , the cloud server 10 may change at least some weights of at least some hidden layers of the currently learned prediction model.

In this case, since the weights set for the nodes of the hidden layer are the prediction model, in the case of FIG. 12A, at least a part of the currently learned prediction model may be replaced according to new data. In this case, the new data may be data according to the newly added or changed characteristics of the measuring device.

That is, when a measurement device is newly added or changed, as shown in (a) and (b) of FIG. 12A, a new hidden layer, for example, at least one exclusive layer or a common layer may be further added. Alternatively, at least one exclusive layer or common layer may be changed to another layer. In this case, the other layers may be layers having different weights assigned to at least some of the nodes constituting the layer.

On the other hand, when the measurement device is removed, some exclusive layers or common layers constituting the currently learned predictive model may be removed. In this case, the model shown in (b) of FIG. 12A may be changed to the model shown in (a) of FIG. 12A through an update.

Meanwhile, when the prediction model is updated by replacing, adding, or removing at least some of the hidden layers, the cloud server 10 may perform re-learning on the updated prediction model. To this end, the cloud server 10 and the collection terminal 20 may repeatedly perform the process from step S506 to step S522 of FIG. 5 .

Meanwhile, unlike the case of FIG. 12A, the cloud server 10 may further generate a new prediction model without changing the existing prediction model, and may use the existing model and the new model in combination.

For example, as shown in (a) of FIG. 12B, when a predictive degree is calculated using one predictive model (model A, 1200) before the update, if the event condition is satisfied, the cloud server ( 10) may further generate at least one new model that meets the event condition.

For example, when a new measuring device is added and an existing measuring device is replaced with another device, the cloud server 10 displays a first new model (model B, 1201) according to the device characteristics of the newly added measuring device and the replaced measurement A second new model (model C, 1202) according to device characteristics of the device may be newly created.

Here, the first and second new models 1201 and 1202 may be prediction models specialized for specific input information. That is, it may be a prediction model specialized for a specific measurement device, a specific type of biometric information, a specific biometric information measurement method, a wearing method of a specific measurement device, or a prediction model specialized for a specific measurement device of a specific manufacturer.

Meanwhile, the cloud server 10 may input biosignals measured from the user to each of the existing model 1200 , the first new model 1201 , and the second new model 1202 . In addition, prediction figures output from each of the prediction models 1200, 1201, and 1202 may be merged.

In this case, the cloud server 10 may calculate statistical values for predictive degrees calculated from each of the prediction models 1200 , 1201 , and 1202 . For example, the statistical value may be any one of an average value, a maximum value, and a minimum value. In addition, the calculated statistical value may be transmitted to the collection terminal 20 as a predicted degree calculated for the input biometric information.

Here, the cloud server 10 may assign different weight ratios to the prediction calculated from the existing prediction model 1200 and the prediction calculated from the first and second new models 1201 and 1202, respectively. For example, the weight ratio may be set to 50% for the predictive value of the existing prediction model 1200, and the weight ratio may be set to 25% for the predictive value of the first new model 1201. In addition, the weight ratio may be set to 25% in the predictive value of the second new model 1202 . In addition, the predictive value corresponding to the input biometric information may be calculated by summing the predictive value according to the weight.

In this case, the weight ratio may be determined according to the learning state of each model. Here, the first and second new models 1201 and 1202 may be in a state in which learning has not been completed. Therefore, lower weight ratios may be set for the predictive degrees of the first and second new models 1201 and 1202 for which learning is not completed, compared to the existing model 1200 for which learning is completed. Then, the cloud server 10 may perform learning on each of the first and second new models 1201 and 1202 in parallel.

On the other hand, if the first and second new models 1201 and 1202 are models specialized for specific input information, specific new models are provided according to input information provided from the collection terminal 20, that is, biometric information (or information of a measuring device). The weight ratio of the model can be high. For example, if the first new model 1201 is a specialized model specialized for the currently received biometric information, the cloud server 10 sets the weight ratio of the existing model 1200 and the second new model 1202 to 0%. can In addition, the weight ratio of the first new model 1201 may be set to 100%. In this case, the predictive value of the first new model 1201 may be transmitted to the collecting terminal 20 as a predictive value corresponding to the input biometric information.

Meanwhile, when a plurality of predictive models are used as shown in FIG. 12B, the calculation time of the predictive value may be delayed. Accordingly, the cloud server 10 may preliminarily limit the maximum number of usable models. In addition, when updating the predictive models, if the number of predictive models exceeds the limited maximum number of uses, the predictive models may be removed in order of oldest elapsed time since creation. Accordingly, it is possible to prevent an excessive amount of time for calculating the prediction.

Meanwhile, as shown in FIG. 12B , unlike using a plurality of prediction models together, the cloud server 10 may update a prediction model by combining a plurality of prediction models with each other.

In this case, as shown in FIG. 12C , the cloud server 10 includes an existing model 1200 and at least one model derived from the learning process of the existing model 1200 (first derived model 1211, second derived model 1211). A derived model 1212) may be created as a model to be combined. That is, the first derived model 1211 and the second derived model 1212 may be the existing model 1200 in a state before learning is completed.

In this case, the cloud server 10 may overlap or combine at least a part of the existing model 1200, the first derived model 1211, and the second derived model 1212 with each other. That is, for example, nodes having equal weights in each hidden layer may be maintained, and nodes having different weights may be added to each hidden layer. In this case, as described above, the first derived model 1211 and the second derived model 1212 are the existing model 1200 in a state before learning is completed, and the number of hidden layers including exclusive layers or common layers is the same. It can be a predictive model.

The cloud server 10 may generate a combine model or an ensemble model (hereinafter, a combine model 1210) by overlapping and merging at least some of the plurality of prediction models. Further, learning may be performed again on the predicted model updated with the generated combination model 1210 . Accordingly, the cloud server 10 and the collection terminal 20 may repeatedly perform the process from step S506 to step S522 of FIG. 5 .

Meanwhile, as described above, when a prediction result is calculated for each biosignal received from the collection terminal 20 and the calculated prediction result is provided to the collection terminal 20, the prediction result is transmitted after the biometric information is transmitted. It may take a long time to reach. In addition, when there is a plurality of received biometric information, there is a problem that the computational load of the cloud server 10 may increase due to an increase in the amount of computation of the cloud server 10 .

Accordingly, the health prediction system according to an embodiment of the present invention, when the learning of the predictive model is completed, pre-calculates predictive degrees corresponding to bio-signals corresponding to a certain range, and sends the calculated predictive degrees to the collection terminal 20. By transmitting in advance, it is possible to reduce the amount of computation of the cloud server 10 and shorten the time required until the prediction result corresponding to the biosignal is transmitted.

13 is an operation process in which the cloud server 10 provides a caching table including diagnosis prediction results according to user's biometric information to the collection terminal 20 in the health prediction system according to the embodiment of the present invention. It is a flow chart showing 14 is an exemplary diagram illustrating an example in which a caching interval for extracting the caching table of FIG. 13 is determined according to user's biometric information and characteristics.

Referring to FIG. 13 , the collection terminal 20 may transmit information on the user's basic biometric characteristics (S1300). Here, the user's basic biometric characteristic information is information about the user's universal biological characteristics, such as the user's gender and age group, and is information stored in the collection terminal 20 or collected from user information registered in a measuring device. can

Then, the collection terminal 20 may set a universal prediction interval corresponding to the received information on the basic biometric feature (S1302). For example, if the user is a man with a common body type, a universal prediction range for weight may be set from 20 kg to 220 kg. In addition, if the user is a woman, since the weight is usually smaller than that of a man, a universal prediction range (eg, 20 kg to 150 kg) may be set that is smaller than the universal prediction range corresponding to the man.

Also, in this way, it is possible to set a range from 50 cm to 250 cm as a universal prediction interval for the user's height.

The universal prediction interval may be an arbitrary biometric information range that can be maximally obtained in a typical case with respect to the basic biometric characteristics of the received user. This universal prediction interval may be determined according to a number of experimental or statistical results related to the present invention. In the following description, for convenience of description, it is assumed that the biometric information detected from the user is weight and height, and an example in which a predictive value for diabetes is detected based on the weight and height will be described.

When a universal prediction interval is set from the user's basic biometric characteristics in step S1302, the cloud server 10 corresponds to each of the universal prediction intervals through a prediction model in which the likelihood of developing diabetes is learned according to the user's height and weight. Prediction results may be calculated in advance (S1304). For example, the cloud server 10 sets virtual keys corresponding to the key range of the universal prediction interval for each preset unit (eg 1 cm), and sets the weight of the universal prediction interval for each preset unit (eg 1 kg). Virtual body weights corresponding to the section may be set. And, by inputting each of the virtual weights corresponding to each of the virtual heights into the prediction model, prediction results corresponding to each of the biometric information (height for each section and weight for each section) constituting the universal prediction interval may be calculated. .

In this case, if the range of height set as the universal prediction interval is 50 cm to 250 cm, the range of weight set as the universal prediction interval is 20 kg to 220 kg, and if the basic unit of height is 1 cm and the basic unit of weight is 1 kg A total of 40000 prediction results can be calculated.

Meanwhile, when prediction results corresponding to each of the virtual biometric information constituting the universal prediction interval are calculated, the cloud server 10 may generate a prediction result table including the calculated prediction results (S1306). Therefore, a prediction result table including diabetic disease predictors matching each of a plurality of biometric information constituting the universal prediction interval can be generated as shown in [Table 2].

[Table 2]

Then, the cloud server 10 may set a part of the universal prediction interval as a caching interval based on the previously acquired biometric information of the user (S1308). For example, when the user's height according to the previously obtained user's biometric information is 160 cm, the cloud server 10 sets a certain range as a caching section based on the user's previously obtained height of 160 cm. can be set In this case, if the predetermined range is 20 cm, a caching section for keys from 140 cm to 180 cm may be set.

On the other hand, if the user's weight according to the user's previously obtained biometric information is 60 kg, the cloud server 10 may set a certain range as a caching section based on the user's previously obtained weight of 60 kg. . In this case, if the predetermined range is 20 kg, a caching section for weights ranging from 40 kg to 80 kg may be set.

Then, the cloud server 10 may extract a region corresponding to the caching section set in step S1308 as a caching table from the prediction result table (S1310). Therefore, part (140cm to 180cm) of the universal prediction interval of any one type of biometric information (height) constituting the prediction interval table and the universal prediction interval of the other type of biometric information (weight) constituting the prediction interval table A portion of the prediction interval table corresponding to a portion (40 kg to 80 kg) may be extracted as a caching table. In this case, the caching table may be a table including diabetic disease predictors corresponding to the biometric characteristics (height and weight) of the user in each of the extracted caching intervals, as shown in Table 3 below.

[Table 3]

Then, the cloud server 10 may transmit the caching table extracted in step S1310 to the collection terminal 20 (S1312). Then, the collection terminal 20 may store the received caching table (S1314).

Meanwhile, the caching period may be dynamically set based on a direction according to a user's biological characteristics. For example, when the age of the user is a teenager, the weight and height of the user may be on the trend of increasing. On the other hand, when the age of the user is middle-aged, the user's weight may be in an increasing trend, but the user's height may be in a stationary trend. Based on the user's direction for each age group, the cloud server 10 may dynamically set the caching interval.

14 illustrates an example in which a caching section is dynamically set according to a user's biological characteristics (eg, age group).

As described above, when the user's age group is adolescence, the user's weight and height may tend to increase. Accordingly, the cloud server 10 may set, as the caching period, a period increased by a predetermined value from the caching period set according to the previously acquired biometric information of the user.

Referring first to (a) of FIG. 14, (a) of FIG. 14 is an example in which a part of the prediction interval 1400 is set as the caching interval 1420 based on previously obtained biometric information (eg: 60 kg) of the user. is shown. Here, if the range of the caching range set by the reference value is 20 kg, the range from 40 kg, which is 20 kg less than the reference value, to 80 kg, which is 20 kg greater than the reference value, can be set as the caching interval for the body weight.

In this case, the cloud server 10 may infer the direction of the user's biological characteristics over time according to the user's biological characteristics. Here, the orientation of the biometric feature may be inferred based on statistical results related to the biometric feature.

Meanwhile, as described above, when the age of the user is a teenager, the weight of the user may tend to increase. Accordingly, the cloud server 10 may determine that the user's biological characteristics are in an increasing trend. Therefore, the cloud server 10 may set the caching interval in an increasing direction. Therefore, as shown in (b) of FIG. 14 , the minimum and maximum values of the caching section 1420 set according to the previously acquired biometric information of the user may be increased from the initially set values. That is, the entire set caching section 1420 may be moved in a direction in which the body weight increases.

15 is a flowchart illustrating an operation process of providing a diagnosis prediction result using a caching table in the health prediction system according to an embodiment of the present invention.

Referring to FIG. 15 , the collection terminal 20 of the health prediction system according to an embodiment of the present invention may first receive the user's biometric information from the measuring device (S1500). In addition, it may be detected whether there is a predictive value corresponding to the biometric information of the user received in step S1500 in the caching table received from the cloud server 10 (S1502).

For example, as described above with reference to FIG. 13, when the caching table stored in the collection terminal 20 is the predictive degree of onset of diabetes-related diseases based on the weight and height of the user, the collection terminal 20 performs the step S1500. can receive the user's height and weight from the measuring device. In addition, it is possible to determine whether or not predictions corresponding to the received height and weight are included in the caching table. For example, the collection terminal 20 predicts the biometric information of the user received in step S1500 in the caching table according to whether the biometric information received in step S1500 is included in each caching section constituting the caching table. You can judge whether there is a degree or not.

And, as a result of the determination in step S1502, if it is possible to detect the predictive value corresponding to the biometric information obtained in step S1500 through the caching table, the collection terminal 20 outputs a prediction result according to the predicted degree retrieved from the caching table. It can (S1504). In this case, the collection terminal 20 may apply a normally set threshold. Therefore, when the predictive value corresponding to the acquired biometric information can be detected through the caching table, there is no need to connect communication with the cloud server 10, so a disease prediction result corresponding to the acquired biometric information can be output very quickly. In addition, the computational load of the cloud server 10 may be reduced.

In this way, when prediction results are provided using the caching table, the collection terminal 20 may change the threshold according to the user's situation.

In this case, the collection terminal 20 may store information about change displacement of the threshold value corresponding to a preset situation. In addition, when the situation of the user calculated based on information collected from at least one peripheral device including the user corresponds to the preset situation, the threshold value may be changed based on the threshold change displacement corresponding to the preset situation. .

Also, the collection terminal 20 may detect a predictive value corresponding to the biometric information received from the measurement device based on the caching table. In addition, a prediction result determined based on the detected predictive value and the changed threshold may be output through the output unit 250 .

However, as a result of the determination in step S1502, if the predictive value corresponding to the biometric information obtained in step S1500 cannot be detected through the caching table, the collection terminal 20 transfers the biometric information obtained in step S1500 to the cloud server ( 10) can be transmitted (S1506). Then, the cloud server 10 may retrieve a prediction map corresponding to the biometric information transmitted from the collection terminal 20 from the prediction result table according to the universal prediction interval (S1508). Then, based on a predetermined threshold value, it is determined whether or not the disease is predicted for the retrieved predictive value, and the determined result may be transmitted to the collection terminal 20 as a predicted result (S1510). Then, the collection terminal 20 may output the received prediction result through the output unit 250 .

Meanwhile, the cloud server 10 may reset at least one caching interval based on the biometric information obtained in step S1506 when the predictive value corresponding to the user's biometric information is not detected from the transmitted caching table. (S512). And, a new caching table including prediction degrees corresponding to the reset caching interval may be extracted from the prediction result table (S1516). Then, the extracted new caching table may be transmitted to the collection terminal 20 (S1516). Then, the collection terminal 20 may update the caching table by replacing the previously stored caching table with the caching table received in step S1516 (S1518).

Meanwhile, according to the above description, the health prediction system according to an embodiment of the present invention is connected to a blockchain network, and an authorized third party can view the user's biometric information stored in the cloud server 10 through the blockchain network. I have mentioned that it can be done.

16 is a flowchart illustrating an operation process of allowing an authorized third party to view medical information of a user in the health prediction system according to an embodiment of the present invention.

Referring to FIG. 16 , first, the measurement device 30 may transmit measured biometric information (measurement information) of the user to the user terminal, that is, the collection terminal 20 (S1701). Then, the collection terminal 20 may transmit the collected measurement information to the cloud server 10 (S1702). In addition, the cloud server 10 may classify the received measurement information according to a classification criterion such as a time at which the measurement information was received or a type of biometric information, and may store the received measurement information in the biometric information DB 133 (S1703).

Meanwhile, when the user's biometric information is stored, the cloud server 10 may request authentication keys of at least one third party capable of accessing the user's biometric information from the blockchain network 1600 . Here, the third party who can access the biometric information of the user is a third party previously registered by the user, and may be a medical service worker, trainer, or the user's family.

In this case, the third party's authentication key request can be transmitted to the Registry Contract constituting the blockchain network 1600 (S1704). The registry contract 1601 is a component of a block chain that maps an ID and corresponding data, and can map a block chain address corresponding to an ID. Therefore, when the registry contract 1601 requests at least one third party authentication key pre-registered by the user from the cloud server 10, the requested at least one third party authentication is obtained from a data block distributed and stored in a blockchain method. Address information of the key is provided, and based on the provided address information of the third party authentication key, the cloud server 10 may receive at least one third party authentication key (S1705).

Then, the cloud server 10 may generate a token corresponding to at least one third party having received the authentication key (S1706). Here, the token represents the right for a third party to view the user's biometric information, and only a third party having the token can view the biometric information stored in the cloud server 10.

Here, the token may include a range that can be viewed by a third party among biometric information stored in the cloud server 10 (limitation of viewing range) and may include a valid period (limitation of viewing deadline). Therefore, even if the token is registered, it can be viewed only within the scope specified in the token (e.g., the time at which biometric information is stored or the type of biometric information), and when the above validity period elapses, the user's biometric information can be viewed through the token. this may be impossible

Meanwhile, when a token for a third party is generated in step S1706, the cloud server 10 may register the generated token in the blockchain network 1600 (S1707). In this case, the token may be transmitted to the access contract 1602 constituting the blockchain network 1600. Here, the access contract 1702 is a component that manages access rights to data blocks in the blockchain network 1600, and the token is access rights to biometric information stored in the cloud server 10, and the access contract It can be registered in (Access Contract) (1602).

On the other hand, in step S1707, if the third party's token is registered in the access contract 1602, a third party with legitimate authority to view the user's biometric information, that is, a third party designated in advance by the user (hereinafter, an authorized third party) ) may transmit an access request for user biometric information to the user terminal 20 through its own terminal 1603 (S1708). Then, the user terminal 20 may transmit the data access request received from the authorized third party to the cloud server 10 (S1709).

Then, the cloud server 10 may request a third party's token corresponding to the received data access request from the access contract 1602 in which the token is registered (S1710). In response to the token request, the access contract 1602 may transmit the third party's token corresponding to the data access request to the cloud server 10 (S1711). Then, the cloud server 10 checks the third party's viewing range and viewing deadline through the token received from the access contract 1602, and the user's biometric information according to the third party's request is allowed by the token. In the case of biometric information within the viewing range and viewing deadline, address information that allows access to an area in which biometric information corresponding to the third party's data access request is stored, that is, access address information, can be transmitted to the user terminal 20 (S1712).

Then, the user terminal 20 may perform additional authentication on the third party's terminal 1603 (S1703). For example, the user terminal 20 may determine whether the third party terminal 1603 is located within a preset distance from the user terminal 20 or within the same area. As a result of such region authentication, only when the third party terminal 1603 is located within a preset distance from the user terminal 20 or within the same area, the user terminal 20 receives the information received from the cloud server 10. The access address information may be transmitted to the third party terminal 1603 (S1714).

Then, based on the received access address information, the third party terminal 1603 may request to view some of the biometric information stored in the cloud server 10, which is allowed to be viewed by the token, among the biometric information of the user. Yes (S1715). Then, the cloud server 10 allows reading according to the reading request, so that some of the user's biometric information stored in the cloud server 10 may be displayed through the third party terminal 1603 ( S1716).

Meanwhile, even if the user's biometric information can be provided to the third party terminal 1603 through browsing of the access address transmitted in step S1716, this may be limited to browsing. That is, the third party terminal 1603 can only check the user's biometric information, and processing or storing the user's biometric information in the third party terminal 1603 may be restricted. This is because the user's biometric information provided from the cloud server 10 to the third-party terminal 1603 is merely permissible for viewing, and the biometric information is not actually provided.

The above-described present invention can be implemented as computer readable code on a medium on which a program is recorded. The computer-readable medium includes all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable media include Hard Disk Drive (HDD), Solid State Disk (SSD), Silicon Disk Drive (SDD), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. , and also includes those implemented in the form of a carrier wave (eg, transmission over the Internet). Also, the computer may be the server 10 of the health prediction system according to an embodiment of the present invention.

Therefore, the above detailed description should not be construed as limiting in all respects and should be considered as illustrative. The scope of the present invention should be determined by reasonable interpretation of the appended claims, and all changes within the equivalent scope of the present invention are included in the scope of the present invention.

Claims (20)

a communication unit that performs wireless communication with a collection terminal that collects biometric information of a user measured by at least one measuring device;
When characteristic information of the at least one measurement device is received through the communication unit, a predictive model for a specific disease is generated by reflecting the characteristic information of the received measurement device, and the predictive model is based on the collected biometric information of the user. An artificial intelligence unit that performs learning on; and,
When the prediction model is learned, the artificial intelligence unit controls the artificial intelligence unit to calculate a predictive value related to the specific disease in response to the biometric information collected in the collection terminal from the learned predictive model, and and a control unit for determining a prediction result for the specific disease based on a set threshold and transmitting the prediction result to the collection terminal so that the determined prediction result is displayed through the collection terminal. 's server. According to claim 1,
The collection terminal,
Obtaining information related to the user's situation from at least one peripheral device around the user, analyzing the user's situation based on the acquired information, and transmitting the analyzed situational information about the user's situation to the server;
The server,
A server of a health prediction system, characterized in that for changing a threshold value based on the situation information, and determining a prediction result for the specific disease based on the changed threshold value and the predictive value. The method of claim 2, wherein the at least one peripheral device,
A server of a health prediction system, characterized in that at least one mobile terminal located around a user or an access point (AP) installed in an area where the user is located is information. According to claim 1,
The predictive model,
At least one exclusive layer, which is a hidden layer including weights applied to input biometric information classified into a specific group, and at least one common layer, which is a hidden layer including weights applied to all input biometric information It is an artificial neural network including,
The biometric information,
It is input to either the exclusive layer or the common layer according to whether it is classified into the specific group,
The predictive value output as a result of the predictive model is,
A server of the health prediction system, characterized in that output from any one of the at least one common layer. The method of claim 4, wherein the control unit,
The server of the health prediction system, characterized in that the biometric information is input to the prediction model as the biometric information of the specific group according to the characteristic information of the measuring device in which the biometric information is measured. The method of claim 1, wherein the characteristic information of the measuring device,
At least one of the manufacturer of the measuring device, the type of biometric information measured by the measuring device, the type of the measuring device, a mounting method in which the user mounts the measuring device, and a measuring method in which the measuring device measures the user's biometric information. Server of the health prediction system comprising a. The method of claim 6, wherein the characteristic information of the measuring device,
The server of the health prediction system, characterized in that it includes quantified clinical accuracy quantifying the clinical accuracy of the measuring device. According to claim 1,
The control unit,
Updating the prediction model when a preset update condition is satisfied;
The preset update condition is,
A server of a health prediction system characterized in that it is satisfied when a preset update cycle expires or when a measuring device measuring user's biometric information is added, replaced, or any one measuring device is removed. The method of claim 8, wherein the control unit,
Change the weight assigned to at least one node among the hidden layers constituting the prediction model, remove at least one of the hidden layers, or update the prediction model to further include a new hidden layer including nodes to which a new weight is assigned and
The server of the health prediction system, characterized in that for determining the changed weight based on the device characteristic information of the added, replaced, or removed measuring device, or determining the hidden layer to be removed or the new hidden layer. The method of claim 8, wherein the control unit,
At least one new predictive model is generated according to the device characteristic information of the added, replaced, or removed measuring device, and the prediction calculated from the existing predictive model and the at least one new predictive model for the collected biometric information of the user. A server of a health prediction system, characterized in that for calculating statistical values of degrees as the predictive degree. The method of claim 8, wherein the control unit,
Based on the collected biometric information of the user, at least one predictive model derived in the process of learning the predictive model is combined with the learned predictive model to update the learned predictive model. server. The method of claim 11, wherein the control unit,
Combining the learned prediction model and the at least one derived model by comparing weights of hidden layers corresponding to each other from the learned prediction model and the at least one derived model, and adding nodes to which different weights have been assigned. The server of the characterized health prediction system. According to claim 1,
If the measuring device is a measuring device that measures the user's blood sugar, the predictive model is a model obtained by learning the user's blood sugar change amount that fluctuates according to each nutrient,
The control unit,
When the user inputs a virtual menu through the collection terminal, the nutrients of each food included in the input menu are analyzed, the user's blood sugar change according to each analyzed nutrient, and the value of each food input along with the virtual menu Based on the amount of intake, the server of the health prediction system, characterized in that for calculating the change in total blood sugar of the user corresponding to the virtual diet and transmitting it to the collection terminal as a response to the input of the virtual diet. The method of claim 2, wherein the control unit,
When the learning of the prediction model is completed, determining a range of minimum and maximum values of the user's biometric information that can be detected through the measurement device from the user's basic biometric characteristics;
Controls the artificial intelligence unit so that predictive degrees of the predictive model corresponding to each biometric information included in the range of the determined minimum and maximum values of biometric information are calculated, and some of the calculated predictive degrees are used as a caching table. A server of the health prediction system, characterized in that for transmitting to the collection terminal. The method of claim 14, wherein the collection terminal,
When the caching table is received from the server, a predictive value corresponding to the biosignal collected from the measuring device is detected from the caching table, and the predictive value detected from the detected caching table is based on the threshold value according to the situation information. The server of the health prediction system, characterized in that for determining the prediction result for the specific disease and outputting the determined prediction result. According to claim 1,
The server,
Further comprising a memory for storing the user's biometric information collected through the communication unit by dividing it according to the time at which the biometric information was measured and the type of the biometric information;
The control unit,
The server of the health prediction system, characterized in that for providing address information related to the storage area in which the divided biometric information is stored in the memory as a response to a user's request. The method of claim 16, wherein the control unit,
Registering a token representing the right of a third party to request biometric information of a user through a blockchain network, and providing the address information to the third party at the request of the user based on the registered token provide,
The server of the health prediction system, characterized in that, based on the token registered in the blockchain network, the range of the user's biometric information that can be viewed by the third party and the valid period during which the user's biometric information can be viewed are limited. A method for controlling a health prediction system that provides a prediction result related to a specific disease based on user's biometric information, comprising:
connecting, at the collection terminal of the health prediction system, communication with at least one measuring device to measure user's biometric information, and collecting device characteristic information of the at least one measuring device;
receiving, by a cloud server, the characteristic information of the measurement device collected from the collection terminal, and generating a predictive model for a specific disease in which the characteristic information of the measurement device is reflected;
receiving, by the collection terminal, biometric information of a user collected by the at least one measurement device;
receiving, by the cloud server, biometric information collected by the collection terminal, and learning the prediction model based on the received biometric information;
calculating, by the cloud server, a predictive value corresponding to the biometric information received from the collection terminal from the learned predictive model when learning of the predictive model is completed; and,
The cloud server or the collection terminal determines a prediction result for the specific disease based on a predetermined threshold value and the predictive value, and outputs the determined prediction result. control method. The method of claim 18, wherein the characteristic information of the measuring device,
At least one of the manufacturer of the measuring device, the type of biometric information measured by the measuring device, the type of the measuring device, a mounting method in which the user mounts the measuring device, and a measuring method in which the measuring device measures the user's biometric information. A control method of a health prediction system comprising a. According to claim 18,
The step of calculating the predictive degree,
determining, by the cloud server, a range of minimum and maximum values of the user's biometric information that can be detected through the measurement device, from basic biometric characteristics of the user;
calculating, by the cloud server, predictive degrees of the learned prediction model corresponding to each biometric information included in the range of the minimum and maximum values of the biometric information; and,
Transmitting, by the cloud server, some of the calculated predictive degrees to the collection terminal as a caching table;
The step of the cloud server or the collection terminal outputting the prediction result,
detecting, by the collecting terminal, any one predictive value corresponding to the biometric information received from the measuring device among predictive values included in the caching table; and,
and determining, by the collection terminal, a prediction result for the specific disease based on the detected predictive value and a predetermined threshold value, and outputting the determined prediction result.

Images