KR20200106691A - User terminal of providing disease prediction information through bid data analysis and artificial intelligence - Google Patents

Abstract

The present invention relates to a user terminal for providing disease prediction information through big data analysis and an artificial intelligence medical examination by interview to enable a user to easily and correctly check diseases related to a symptom before visiting a hospital. According to one embodiment of the present invention, the user terminal comprises: a user interface unit receiving a user′s input; an information storage unit storing at least one information table; a communication unit communicating with an external server storing a plurality of information tables; and an operation processing unit. The operation processing unit performs the following processes of (a) determining a first candidate symptom based on the user′s voice input received through the user interface unit; (b) determining a plurality of candidate diseases corresponding to the first candidate symptom; (c) providing a user with a medical question corresponding to each of the plurality of candidate diseases and assigning first weight to each of the plurality of candidate diseases based on the user′s answer input with respect to the medical question; (d) assigning second weight to each of the plurality of candidate diseases; (e) calculating a disease occurrence probability for each of the plurality of candidate diseases based on the first weight and the second weight; and (f) generating personalized disease prediction information based on the calculated disease occurrence probability and providing the information to the user.

Description

A user terminal that provides disease prediction information through big data analysis and artificial intelligence interviews {USER TERMINAL OF PROVIDING DISEASE PREDICTION INFORMATION THROUGH BID DATA ANALYSIS AND ARTIFICIAL INTELLIGENCE}

The present disclosure relates to a user terminal that provides disease prediction information through big data analysis and artificial intelligence interviews. More specifically, disease prediction information is generated and provided by analyzing user disease information generated based on user input. It relates to a user terminal capable of.

General people visit a hospital and see a doctor if their body has symptoms. Since most individuals lack specialized medical knowledge, they do not know which disease is based on their own symptoms. Therefore, individuals with symptoms (hereinafter referred to as "symptom holders") often visit a hospital that does not target the disease corresponding to their symptoms to receive medical examination. In such a case, there is a disadvantage that additional time and cost are consumed because the symptom holder must visit a hospital targeting specialized treatment for the disease corresponding to his or her symptoms and undergo a medical examination.

Symptom holders can also use the portal site to determine which disease their symptoms are related to. That is, the symptom holder can check which disease his symptom is related to based on the information retrieved by entering his symptom in the search item of the portal site. However, since the information searched through the portal site contains many advertising information and inaccurate information, it is difficult for symptom holders to accurately identify diseases related to their symptoms using only the information searched through the portal site. In addition, even when relatively accurate information is obtained through a search, since specialized medical knowledge is insufficient, it is difficult for a symptom holder to accurately identify a disease related to his or her symptoms. That is, since the same symptom can be commonly expressed in a number of diseases, it is difficult for a symptom holder to accurately identify a disease related to his or her symptoms simply by searching a portal site.

Also, symptom holders may think their symptoms are insignificant or may not visit the hospital because they are bothersome. In other words, time and money are consumed when visiting a hospital and receiving a medical examination, so symptom holders can neglect their symptoms. In that case, since the symptom holder visits a hospital and undergoes a medical examination when the disease worsens, there is a disadvantage of missing an opportunity to detect and treat the disease early.

Also, a symptom holder may visit a hospital because he/she is sensitive to his/her symptoms. In other words, even though there are no other diseases, they may mistake their symptoms as having a disease and visit a hospital.

Meanwhile, technology is being developed that allows symptom holders to diagnose diseases remotely without visiting a hospital. For example, Korean Patent No. 10-1616473 (Patent Document 1) entitled "Smartphone Remote Clinic", filed on July 16, 2015 by Lee Byung-hoon and Lee Jae-cheon and registered on April 22, 2016, is a stethoscope, Disclosed is a smartphone remote medical treatment device capable of diagnosing a user's disease name or health condition by linking a medical treatment device including a thermometer, a medical examination imager, and an emergency alarm generator with a smartphone. According to the configuration disclosed in Korean Patent Registration No. 10-1616473, there is an advantage in that the symptom holder can receive treatment even if he or she does not visit the hospital directly. However, according to the configuration disclosed in Korean Patent Registration No. 10-1616473, a medical treatment device including a stethoscope, a thermometer, a medical examination imager and an emergency alarm generator is required, or a configuration such as a temperature sensor, a stethoscope sound measurement unit, and a medical examination image observation unit There is a disadvantage that it requires a smart phone including. In addition, according to the configuration disclosed in Korean Patent Registration No. 10-1616473, a disease such as pneumonia can be diagnosed based on the measured data. However, the disadvantage is that the symptom holder cannot actively input his or her symptoms to check the disease.

Therefore, there is an increasing need for a method that can support symptom holders to more easily and accurately identify diseases related to their symptoms before visiting a hospital.

1. Korean Patent Registration No. 10-1616473

The present disclosure is intended to solve the above-described problems, and generates user disease information based on a plurality of information tables and user inputs calculated and stored in advance using deep learning algorithms and medical big data, and uses the generated user disease information. By generating and providing disease prediction information, before a symptom holder visits the hospital, it is easier and more accurate to check the disease related to his or her symptoms and recommends a hospital that can professionally examine the disease related to his or her symptoms. It aims to provide a user terminal that can be received.

A user terminal that provides disease prediction information through big data analysis and artificial intelligence interview according to an embodiment of the present disclosure to achieve the above object, includes a user interface unit receiving a user input, and storing at least one information table. An information storage unit, a first information table including a correspondence relationship between a voice input and a symptom, a second information table including a correspondence relationship between a symptom and a plurality of disease groups, and a third including a correspondence relationship between a disease and a plurality of interview groups Using the first information table based on the user's voice input received through the information table, a communication unit that transmits and receives from an external server that stores a fourth information table including a plurality of demographic information on diseases, and (a) a user interface unit. A process for determining a first candidate symptom, (b) a process for determining a plurality of candidate diseases corresponding to the first candidate symptom using a second information table, (c) a process for determining a plurality of candidate diseases using a third information table, respectively A process of providing the user with a questionnaire corresponding to the user's questionnaire, and assigning a first weight to each of the plurality of candidate diseases based on the user's answer input to the questionnaire, (d) each of the plurality of candidate diseases using the fourth information table. A process of assigning a second weight, (e) a process of calculating the occurrence probability of each of a plurality of candidate diseases based on the first and second weights, (f) generating personalized disease prediction information based on the calculated probability of occurrence And, it may include an operation processing unit that performs processing to provide the generated personalized disease prediction information to the user.

And the process (b) is, (b-1) a process of setting the initial value of i to 1, (b-2) determining an i-th candidate disease group corresponding to the i-th candidate symptom using the second information table. Treatment, (b-3) treatment of providing the user with symptoms other than the i-th candidate symptom group corresponding to the i-th candidate disease group, (b-4) the user-selected (i+1) candidate The treatment to determine the (i+1)th candidate disease group corresponding to both the symptom and the ith candidate symptom, (b-5) the treatment to determine whether a preset end condition is satisfied, (b-6) the end condition is not satisfied. If not, increase i by 1 and repeat treatments (b-2) to (b-5), and if the termination condition is satisfied, a disease belonging to the (i+1) candidate disease group is determined as a plurality of candidate diseases. May include processing to

In addition, the termination condition may be that i coincides with a preset number of processing j (j is a natural number of 2 or more).

In addition, the termination condition may be that the number of diseases belonging to the (i+1)th candidate disease group is less than or equal to a preset number.

In addition, processing (c) extracts the interview information corresponding to each of the plurality of candidate diseases using the third information table, configures different user interfaces according to the type of the extracted interview information, and provides the configured user interface to the user. can do.

In addition, the fourth information table includes a 4-1 information table including an incidence rate, an incidence ranking, and a weight value according to gender and a 4-2 information table including an incidence rate, an incidence ranking, and a weight value according to age. In (d), a second weight may be assigned to each of a plurality of candidate diseases through the fourth information table, using pre-input user age information and user gender information.

In addition, processing (f) provides personalized disease prediction information including a list of predicted diseases displayed in the order of calculated probability of onset to the user, and when the user selects a disease name included in the disease list, treatment related to the selected disease name You can provide a list of recommended hospitals that operate the subject.

In addition, the list of recommended hospitals may be provided in the order of hospitals located in close proximity based on the location of the user among hospitals operating treatment courses related to the selected disease name.

In addition, the recommended hospital list may include hospital name information, hospital location information, and hospital contact information, and may further include at least one of hospital evaluation grade information, medical cost information, and treatment period information.

According to various embodiments of the present disclosure as described above, before a symptom holder visits the hospital, it is possible to more easily and accurately check the disease related to his or her symptoms, and to receive a recommendation from a hospital that can examine his or her symptoms professionally. It works.

1 is a diagram showing the configuration of a disease prediction information providing system according to an embodiment of the present invention;
2 is a diagram showing the configuration of a user terminal according to an embodiment of the present invention;
3 is a diagram showing an exemplary configuration of a first information table;
4 is a diagram showing an exemplary configuration of a second information table;
5 is a diagram showing an exemplary configuration of a third information table;
6 to 8 are diagrams showing exemplary configurations of a fourth information table;
9 is a diagram illustrating an example of personalized disease prediction information
10 is a diagram showing an example of a graphic user interface for disease prediction information provided to a user;
11 is a diagram showing an exemplary configuration of a list of recommended hospitals;
12 is a diagram showing an exemplary configuration of a sixth information table, and,
13 is a diagram illustrating an exemplary configuration of a seventh information table.

In describing the present invention, when it is determined that a detailed description of a related known function or configuration may unnecessarily obscure the subject matter of the present invention, a detailed description thereof will be omitted. In addition, terms to be described later are terms defined in consideration of functions in the present invention and may vary according to a user, an operator, or a custom. Therefore, the definition should be made based on the contents throughout this specification.

Terms including ordinal numbers, such as first and second, may be used to describe various elements, but the elements are not limited by terms. The terms are only used for the purpose of distinguishing one component from another component. For example, without departing from the scope of the present invention, a first component may be referred to as a second component, and similarly, a second component may be referred to as a first component. The term and/or includes a combination of a plurality of related items or any one of a plurality of related items.

The terms used herein are used to describe embodiments, and are not intended to limit and/or limit the present invention. Singular expressions include plural expressions unless the context clearly indicates otherwise. In the present specification, terms such as include or have are intended to refer to the presence of features, numbers, actions, components, parts, or a combination thereof described in the specification, but one or more other features, numbers, actions, and components It is to be understood that the possibility of the presence or addition of, parts or combinations thereof is not preliminarily excluded.

Hereinafter, the present invention will be described in detail with reference to the accompanying drawings.

1 is a diagram showing the configuration of a disease prediction information providing system 1000 according to an embodiment of the present invention. According to the embodiment of FIG. 1, the disease prediction information providing system 1000 may include a user terminal 100, a medical information providing server 200, and an auxiliary server 300.

The user terminal 100 may be an electronic device capable of providing personalized disease prediction information to a user (symptom holder). For example, the user terminal 100 may be implemented with various devices such as a smart phone, a tablet, a notebook, and a PC. A detailed configuration of the user terminal 100 will be described later.

The medical information providing server 200 calculates the first to eighth information tables 480 and 410 to 480 using a deep learning algorithm and medical big data. In this specification, the deep learning algorithm is, for example, a deep neural network (DNN), a convolutional neural network (CNN), a recurrent neural network (RNN), a restricted Boltzmann machine (RBM), Algorithms such as Deep Belief Networks (DBNs) and Deep Q-Networks may be referred to, or other deep learning algorithms that are not illustrated may be referred to. In this specification, medical big data refers to patient treatment and prescription information of medical institutions such as hospitals, pharmacies and nursing institutions, information on the status of medical personnel and facilities or equipment possession of medical institutions, Ministry of Health and Welfare, Ministry of Food and Drug Safety, Public Health Center, National Health Insurance It may include information from the corporation, pharmaceutical companies, and supporting organizations such as the Health Insurance Review and Assessment Service.

The medical information providing server 200 receives medical big data from an institution that manages medical big data, such as the Health Insurance Review and Assessment Service, and uses a deep learning algorithm and big data analysis technology to provide first to eighth information tables 480 ) (410-480) can be calculated.

The user terminal 100 may communicate with the medical information providing server 200 to receive and process the first to eighth information tables 480 (410 to 480). The user terminal 100 may store and use at least one information table among the first to eighth information tables 480 and 410 to 480.

The first information table 410 includes a correspondence relationship between voice input and symptom groups including a plurality of symptom information. The second information table 420 includes a correspondence relationship between a symptom group including a plurality of symptom information and a disease group including a plurality of disease information. The third information table 430 includes a correspondence relationship between a disease group including a plurality of disease information and a medical examination group including a plurality of interview information.

The fourth information table 440 is a big data table including demographic information. The fourth information table 440 includes a 4-1 information table 441 including incidence rate information for each disease according to gender, and a 4-2 information table 442 including incidence rate information for each disease according to age. I can.

The fifth information table 450 includes information on hospitals, the sixth information table 460 includes information on hospital expenses, the seventh information table 470 includes health information, and an eighth information table 480 may contain information on hospital catalogs.

Mainly, the eighth information table 480 is stored in the user terminal 100 and may be used for fast provision of hospital search speed. The eighth information table 480 may be generated by extracting only some non-security information necessary for hospital search among hospital information in the fifth information table 450. This is because, if even security information is stored in the user terminal 100, a problem may occur in data security. For example, among the fifth information table 450, the eighth information table 480 is composed of the eighth information table 480, and the eighth information table 480 includes hospital location information and phone number information required when a user searches for a nearby hospital. May be stored in the user terminal 100.

The auxiliary server 300 may receive voice information from the user terminal 100 and perform a voice recognition process. For example, the auxiliary server 300 may include a speech recognition engine capable of performing a speech to text (STT) or text to speech (TTS) operation for converting speech into text. In the embodiment of FIG. 1, the user's voice input from the user terminal 100 is recognized through the auxiliary server 300, but in another embodiment, the user's voice is directly recognized by the user terminal 100 instead of the auxiliary server 300. You may.

2 is a diagram showing the configuration of a user terminal 100 according to an embodiment of the present invention. According to the embodiment of FIG. 2, the user terminal 100 may include a user interface unit 110, an information storage unit 130, an operation processing unit 150, and a communication unit 170.

The user interface unit 110 displays a user interface and receives a user input through the displayed user interface. The user interface and user input will be described later. However, the user input may include one of a voice input and a selection input through a graphic user interface.

The information storage unit 130 may store at least one of the first to eighth information tables 410 to 480 of the medical information providing server 200. In addition, the information storage unit 130 may store a temporary information table to store temporary information necessary to provide personalized disease prediction information. Information stored in the temporary information table may be transferred to the medical information providing server 200 after the process of the operation processing unit 150 is finished. The information table stored in the information storage unit 130 may be determined according to criteria such as data size and information processing speed. For example, the information storage unit 130 may include an eighth information table 480 including a hospital list database and a temporary information table.

The operation processing unit 150 may control the remaining components of the user terminal 100. The operation processing unit 150 may be implemented as a single processor (for example, one CPU or one AP) to perform a disease occurrence probability calculation operation, a personalized disease information analysis operation, etc., and may perform a plurality of processors and specific functions. It can also be implemented as an IP to perform.

The operation processing unit 150 may generate personalized disease prediction information using the information table and provide the generated disease prediction information to a user. The operation processing unit 150 may use an information table stored in the information storage unit 130 or may use an information table stored in the medical information providing server 200 connected through the communication unit 170. Hereinafter, for convenience of explanation, it is assumed that the first to eighth information tables 410 to 480 are stored in the medical information providing server 200. However, due to this assumption, embodiments in which some of the first to eighth information tables 410 to 480 are stored in the information storage unit 130 are not excluded.

The operation processing unit 150 may determine a first candidate symptom using the first information table 410 based on a user voice input received through the user interface unit 110. In addition, the operation processing unit 150 may determine what a plurality of candidate diseases corresponding to the first candidate symptom are using the second information table 420. In addition, the operation processing unit 150 may provide a user with a questionnaire corresponding to each of a plurality of candidate diseases by using the third information table 430. A first weight may be assigned to each of the plurality of candidate diseases based on the user's input of an answer to the questionnaire. Subsequently, the operation processing unit 150 may assign a second weight to each of the plurality of candidate diseases by using the fourth information table 440. The second weight is a value determined in consideration of the incidence rate and the incidence ranking based on age and gender information previously input by the user. The operation processing unit 150 may calculate the probability or risk of each of the plurality of candidate diseases based on the first and second weights. The operation processing unit 150 may generate personalized disease prediction information based on the calculated probability of occurrence and provide it to the user. Detailed operations of the operation processing unit 150 will be described later.

The communication unit 170 may communicate with the medical information providing server 200 and the auxiliary server 300 through a wired/wireless network to transmit and receive medical information data and voice signals. For example, the communication unit 170 may use a high definition multimedia interface (HDMI), a low voltage differential signaling (LVDS), a local area network (LAN), and the like as a wired communication method. As another example, the communication unit 170 may use various methods such as Bluetooth, Zigbee communication, WiFi, infrared (InfraRed, IR) communication, and Near Field Communication (NFC) as a wireless communication method.

3 illustrates an exemplary configuration of a first information table 410 including a correspondence relationship between a voice input and a symptom.

The first information table 410 stores voice data corresponding to each symptom learned through a natural language processing process. When a user's voice explaining a symptom is input through the user interface unit 110, the operation processing unit 150 may control the communication unit 170 to transmit the input voice to the auxiliary server 300. The auxiliary server 300 may recognize the received voice signal, convert it into text, and transmit the converted text to the user terminal 100. When the converted text is received, the operation processing unit 150 may determine what is the first candidate symptom corresponding to the converted text using the first information table 410. For example, when the user says "Sputum appears on the neck", the operation processing unit 150 may obtain the text "Sputum appears on the neck" through transmission/reception with the auxiliary server 300. Using the first information table 410, the operation processing unit 150 may determine that the first candidate symptom corresponding to the content “sputum appears on the neck” is “sputum” of symptom code S55. In the same manner, even when the user says “there is a foreign body in the neck”, the operation processing unit 150 may identify that the first candidate symptom is “phlegm”.

4 shows an exemplary configuration of a second information table 420 including a correspondence relationship between symptoms and diseases. The second information table 420 may include information on a plurality of disease groups corresponding to one symptom and information on a plurality of symptom groups corresponding to one disease.

When the first candidate symptom is determined, the operation processing unit 150 may determine a plurality of candidate diseases using the second information table 420. The operation processing unit 150 may determine a first candidate disease group corresponding to the first candidate symptom. For example, when the first candidate symptom is symptom code S55, the operation processing unit 150 corresponds to S55 {D21, D99, ... } May be determined as the first candidate disease group.

The operation processing unit 150 may search for symptoms corresponding to each candidate disease included in the first candidate disease group. In addition, the operation processing unit 150 may provide the user with symptoms other than the first candidate symptom among the searched symptoms. The user selects a symptom corresponding to his or her physical condition. For example, as the first candidate disease group {D21, D99,… } Is determined, the operation processing unit 150 selects S55, which is already determined as the first candidate symptom, from the combination of {S55, S56} corresponding to D21 and {S55, S57, S99, S102} corresponding to D99. Excluding the symptoms corresponding to {S56, S57, S99, S102} can be provided to the user. If the user selects S57, the operation processing unit 150 may determine {D21, D99} corresponding to both the first candidate symptom S55 and the second candidate symptom S57 as the second candidate disease group.

That is, the operation processing unit 150 may perform a repetitive operation using the second information table 420 in order to narrow the range of possible detailed diseases. By searching for a disease corresponding to a symptom and, conversely, searching for a symptom corresponding to the disease and repeating a selection process, the operation processing unit 150 can more accurately predict what the user's disease is.

The operation processing unit 150 may repeat the above-described process until a preset termination condition is satisfied. This can be explained by focusing on the process as follows. The operation processing unit 150 may set an initial value of i to 1. In addition, the operation processing unit 150 may determine the ith candidate disease group corresponding to the ith candidate symptom using the second information table 420. Subsequently, the operation processing unit 150 may provide the user with symptoms excluding the i-th candidate symptom from the i-th candidate symptom group corresponding to the i-th candidate disease group. The symptom selected by the user among the provided symptoms may be defined as the (i+1)th candidate symptom. The operation processor 150 may determine the (i+1)th candidate disease group corresponding to both the (i+1)th candidate symptom and the i-th candidate symptom selected by the user. Subsequently, the operation processing unit 150 may determine whether a preset end condition is satisfied. If the termination condition is not satisfied, the operation processing unit 150 may repeat the above-described process after increasing i by 1. Conversely, when the termination condition is satisfied, a disease belonging to the determined (i+1)th candidate disease group may be determined as a plurality of candidate diseases predicted to have a user.

The operation processing unit 150 may set various conditions such as the number of repetitions or the number of candidate diseases as termination conditions. For example, the arithmetic processing unit 150 may set the number of repetitions of 4 as the termination condition, or may set the number of diseases belonging to the candidate disease group less than 5 as the termination condition.

5 shows an exemplary configuration of a third information table 430 including a correspondence relationship between a disease and a plurality of questionnaire groups. As shown in the upper part of FIG. 5, the third information table 430 shows a group of questions corresponding to a disease. The operation processing unit 150 may provide a user with a questionnaire corresponding to each of a plurality of candidate diseases using the third information table 430, and may assign a first weight based on the user's input of an answer to the questionnaire. In order to increase the accuracy of disease prediction, various questionnaires that ask about the user's physical condition may be provided, and additional points or deductions may be given depending on the answer to the questionnaire.

For example, if it is assumed that disease codes D21, D22, and D99 are determined as a plurality of currently inventable candidate diseases, the operation processing unit 150 sequentially sets the corresponding questionnaire codes Q1, Q2, Q3, Q8, and Q10 to the user interface unit ( 110) can be provided to the user. According to the user's response input, the operation processing unit 150 may assign a predetermined first weight to each disease. For example, if the user selects Answer 2 as the answer to Q8, 15 points are added, and if the user selects Answer 1 as the answer to Q10, 5 points are deducted.

In FIG. 5, it is shown that only Answers 1 and 2 exist as if selecting one of the two answers of “Yes” and “No”, but it is also possible to receive an answer in the form of selecting a level among 5 choices or 1-10. It is possible. The operation processing unit 150 may configure a different questionnaire interface to be provided to a user according to an answer type of the determined questionnaire code, and provide the configured questionnaire interface through the user interface unit 110.

As another method for increasing the accuracy of disease prediction, the operation processing unit 150 may use the fourth information table 440 including a plurality of demographic information on the disease. The operation processor 150 may assign a second weight to each of the plurality of candidate diseases according to the age and gender previously input by the user.

The fourth information table 440 is a 4-1 information table 441 including an incidence rate according to sex, an onset order, and a weight value, and a 4-2 information table including an incidence rate according to age, an onset order, and a weight value. (442) may be included.

6 shows an exemplary configuration of the 4-1th information table 441. Referring to FIG. 6, the 4-1 information table 441 provides information on what the incidence rate of each disease is according to sex, and what is the ranking when the diseases are listed in the order of the most occurrence in the sex. Contains. In addition, the 4-1th information table 441 also includes weight value information previously calculated by the medical information providing server 200. This second weight can be said to be a weight derived using medical big data.

The operation processor 150 may determine a second weight for each of the plurality of candidate diseases by using the previously inputted user's gender information. For example, assuming that D21, D22, and D99 are determined as multiple candidate diseases, and the user is male, a weight of 2 points for D21 disease, 7 points for D22 disease, and 20 points for D99 disease. Is added.

7 shows an exemplary configuration of the 4-2 information table 442. In the example of FIG. 7, the age section is divided into 10-year-old units, and the incidence rate, incidence ranking, and weight value according to age are included in the 4-2 information table 442, but it is of course possible to configure the unit of the age section differently. .

The operation processor 150 may determine a second weight for each of the plurality of candidate diseases by using the pre-input user's age information. For example, assuming that D21, D22, and D99 were determined as multiple candidate diseases, and the user is 15 years old, a weight of 4 points for D21 disease and 15 points for D22 disease using the information table in the 10-19 year old section. A weight of 20 points is added to the D99 disease.

In the above-described example, an example in which the fourth information table 440 is composed of two sub-information tables has been described, but it goes without saying that the fourth information table 440 ′ may be configured as a single information table. 8 shows an embodiment in which both sex and age are configured as one information table.

The operation processing unit 150 may finally determine a risk ranking according to the calculation and weighting of the probability of occurrence of each of the plurality of candidate diseases. In addition, the operation processing unit 150 may provide the user with a list of diseases as many as a preset number according to the risk ranking. In addition, the calculation processing unit 150 may display the risk level to the user with a name represented by section, not a risk score. For example, by dividing the risk score section into three levels (high, low, lean), one of the three levels of risk can be provided to the user along with the disease name.

First, a method of determining a risk level by the operation processing unit 150 will be described in more detail, and then a method of displaying the risk level will be described.

For example, the operation processing unit 150 may calculate a disease occurrence probability by calculating a ratio of a symptom experienced by a user among symptom information for each of a plurality of candidate diseases. In addition, the operation processor 150 may calculate the risk by adding the first and second weights.

For example, a disease occurrence probability P(DD1) for the first candidate disease may be calculated as follows.

P(DD1) = N(1, user)/N(1, total)-(K11-N(1, total))/K11

N(1, total) is the total number of symptom information about the first candidate disease, N(1, user) is the number of symptom information the user is experiencing for the first candidate disease, and K11 is the number of symptom information about the first candidate disease. It is a designated value, preferably a value generated using medical big data in the medical information providing server 200. K11 may be changed in conjunction with the medical big data used in the medical information providing server 200 is updated.

The risk T(DD1) for the first candidate disease can be calculated as follows.

T(DD1) = P(DD1) X K12 + W1 + W2

K12 is a value previously designated for the first candidate disease, and is preferably a value generated by the medical information providing server 200 using medical big data. K12 may be changed in conjunction with the medical big data used in the medical information providing server 200 is updated. W1 is a first weight for a first candidate disease, and W2 is a second weight for a second candidate disease.

The above-described risk calculation formula is exemplary for constructing an initial learning model. Since the deep learning algorithm has the characteristic of constructing an algorithm by itself according to learning, it goes without saying that the above-described risk calculation formula may be changed in different forms depending on the amount of learning and the learning method.

After calculating the risk for each candidate disease, the operation processing unit 150 may classify and provide three stages according to the risk interval. For example, if the risk score is 100 to 70, it may be marked as'high', if it is 69 to 45, it may be marked as'low', and if it is less than 45, it may be marked as'lean'.

Referring to FIG. 9 as an example, as a result of calculating the risk for each of a plurality of candidate diseases including D21, D22, and D99, D21 received a'high' grade with 70 points, and D22 received a'low' grade with 48. , D99 can be assumed to be rated'lean' with 12 points. The operation processing unit 150 may generate personalized disease prediction information composed of a disease code of a plurality of candidate diseases, a treatment subject name, a treatment subject code, a risk level, and the like, in the order of risk score.

The operation processing unit 150 may process the generated personalized disease prediction information in various forms and provide it to a user. For example, the operation processing unit 150 provides the user with a list of expected diseases displayed in the order of the calculated probability of occurrence, and when the user selects a specific disease name included in the disease list, the operation processing unit 150 operates a treatment course related to the selected disease name. A list of recommended hospitals can be provided. 10 is an example of a graphic user interface 1010 provided to a user. In the embodiment of FIG. 10, probability, risk, and guidance phrases for each of two types of candidate diseases, bronchitis and tonsillitis, are provided. In addition, when the user clicks a separate button called'view hospital recommendation information', the screen is switched to a screen that provides a list of recommended hospitals. As another example, the operation processing unit 150 may provide hospital cost information corresponding to the selected disease name. An embodiment of providing hospital cost information will be described again below.

When the risk of all candidate diseases is lower than a preset value, for example, if all candidate diseases are included in the'lean' section, the operation processing unit 150 may not show the list of candidate diseases at all. In this case, the candidate disease list is displayed only when there is a separate user input such as requesting a candidate disease list from the user through the user interface unit 110.

The operation processing unit 150 may configure a recommended hospital list using at least one of the fifth information table 450 including hospital information and the eighth information table 480 including hospital list data. The first criterion for selecting a recommended hospital is that it must be a hospital that operates treatment subjects related to the selected disease name, and the second criterion is that the hospital must be located in a short distance based on the user's location. That is, the operation processing unit 150 may configure a list of recommended hospitals in the order of hospitals located in a short distance based on the location of the user among hospitals operating treatment subjects related to the selected disease name, and provide them to the user. The operation processing unit 150 may measure the current location of the user through components of the user terminal 100 such as GPS (not shown). However, the user's location used when recommending a hospital does not necessarily refer to the user's current location. Instead of the current location, the user's pre-entered home address and office address can also be used. This is to recommend a hospital that is easy to use according to the user's lifestyle.

The list of recommended hospitals may include hospital name information, hospital location information, and hospital contact information. In addition, the recommended hospital list may further include at least one of hospital evaluation grade information, treatment cost information, and treatment period information. 11 illustrates an exemplary configuration of a list of recommended hospitals provided when D21 disease is selected. In FIG. 11, since the user was in Seoul, hospitals that can be treated in order close to Seoul are displayed.

The operation processing unit 150 may provide hospital cost information using the sixth information table 460 including hospital cost information. Specifically, the calculation processing unit 150 may analyze and provide hospital expenses suitable for the user's age and gender. The hospital cost information included in the sixth information table 460 may be an average value obtained based on medical cost data for each injury and illness code for the last three years of the Health Insurance Review and Assessment Service. 12 illustrates an exemplary configuration of a sixth information table 460 including hospital cost information. For example, if the user is an 18-year-old female, the operation processing unit 150 may provide the user with information that the estimated medical cost for D21 disease is 9500 won.

12 is a diagram illustrating an exemplary configuration of a seventh information table 470. The seventh information table 470 includes detailed description information for each of a plurality of diseases, more specifically, a plurality of disease codes. For example, the 7th information table 470 indicates for the disease code D1, "Migraine is a type of headache that occurs periodically and seizure due to abnormal functions of blood vessels in the head, and it is often referred to as migraine because pain occurs only in one side of the head. Migraine. Is seen in about 60% of all migraine headaches and is more common in women Migraines occur at any age, but most often first in their teens, with 18% of women and 6% of men experiencing a migraine at least once in a lifetime. It can include detailed description information such as.” Detailed description information can be in various forms such as text, images, and videos.

Although the present invention as described above has been described by the limited embodiments and drawings, the present invention is not limited to the above embodiments, and various modifications and variations from these descriptions are those of ordinary skill in the field to which the present invention belongs. This is possible. Therefore, the scope of the present invention is limited to the described embodiments and should not be defined, and should not be defined by the claims to be described later, as well as the claims and equivalents.

According to the technology disclosed herein, user disease information is generated based on a plurality of pre-stored information tables and user inputs calculated using deep learning algorithms and medical big data, and personalized disease prediction using the generated user disease information By creating and providing information, a user who can more easily and accurately identify diseases related to his or her symptoms and receive recommendations for hospitals that can examine diseases related to his or her symptoms before a symptom holder visits the hospital. A terminal can be provided.

1000: disease prediction information providing system 100: user terminal
110: user interface unit 130: information storage unit
150: operation processing unit 170: communication unit
200: medical information providing server 300: secondary server

Claims (9)

In a user terminal that provides disease prediction information through big data analysis and artificial intelligence interview,
A user interface unit for receiving a user input;
An information storage unit for storing at least one information table;
A first information table including a correspondence relationship between a voice input and a symptom, a second information table including a correspondence relationship between a symptom and a plurality of disease groups, a third information table including a correspondence relationship between a disease and a plurality of interview groups, a disease A communication unit for transmitting and receiving to and from an external server storing a fourth information table including a plurality of demographic information about the; And
(a) a process of determining a first candidate symptom using the first information table based on a user voice input received through the user interface unit, (b) a process of determining a first candidate symptom using the second information table Processing of determining a plurality of corresponding candidate diseases, (c) providing a user with a questionnaire corresponding to each of the plurality of candidate diseases using the third information table, and based on the user's answer input to the questionnaire A process of assigning a first weight to each of a plurality of candidate diseases, (d) a process of assigning a second weight to each of the plurality of candidate diseases using the fourth information table, (e) the first and second weights Processing of calculating the probability of occurrence of each of the plurality of candidate diseases based on (f) generating personalized disease prediction information based on the calculated probability of occurrence, and providing the generated personalized disease prediction information to the user User terminal including; an operation processing unit that performs the processing. The method of claim 1,
The treatment (b),
(b-1) a process of setting the initial value of i to 1, (b-2) a process of determining an i-th candidate disease group corresponding to the i-th candidate symptom using the second information table, (b-3) ) A process of providing the user with symptoms other than the i-th candidate symptom among the i-th candidate symptom groups corresponding to the i-th candidate disease group, (b-4) the (i+1)th candidate symptom selected by the user, and A process of determining the (i+1)th candidate disease group corresponding to all of the i-th candidate symptoms, (b-5) a process of determining whether a preset termination condition is satisfied, (b-6) if the termination condition is not satisfied After increasing i by 1, the treatments (b-2) to (b-5) are repeated, and when the termination condition is satisfied, a disease belonging to the (i+1)th candidate disease group is selected as a plurality of candidate diseases. User terminal including processing to determine. The method of claim 2,
The ending condition is that i matches a preset number of processing j (j is a natural number of 2 or more). The method of claim 2,
The termination condition is a user terminal in which the number of diseases belonging to the (i+1)th candidate disease group is less than or equal to a preset number. The method of claim 1,
The treatment (c),
A user who extracts interview information corresponding to each of the plurality of candidate diseases using the third information table, configures different user interfaces according to the type of the extracted interview information, and provides the configured user interface to the user terminal. The method of claim 1,
The fourth information table includes a 4-1 information table including an incidence rate, an incidence ranking, and a weight value according to gender and a 4-2 information table including an incidence rate, an incidence ranking, and a weight value according to age,
The treatment (d),
A user terminal that assigns a second weight to each of the plurality of candidate diseases through the fourth information table using pre-input user age information and user gender information. The method of claim 1,
The treatment (f),
Provides the user with the personalized disease prediction information including a list of expected diseases displayed in the order of the calculated probability of occurrence, and when the user selects a disease name included in the disease list, a treatment subject related to the selected disease name is User terminal that provides a list of recommended hospitals operated. The method of claim 7,
The list of recommended hospitals is provided in the order of hospitals located in a short distance based on the location of the user among hospitals operating treatment courses related to the selected disease name. The method of claim 8,
The recommended hospital list includes hospital name information, hospital location information, and hospital contact information, and further includes at least one of hospital evaluation grade information, medical cost information, and treatment period information.

Images