KR20220165013A - Ai-based personal disease prediction device and method - Google Patents

Abstract

Provided is a device for predicting the personal disease based on artificial intelligence including an output unit, which may receive symptom information from a user, analyze the symptom information to extract feature information, learn multiples types of feature information to generate a symptom prediction model, generate one or more types of disease information matching the feature information based on the symptom prediction model, and output disease information.

Description

AI-based personal disease prediction device and method {AI-BASED PERSONAL DISEASE PREDICTION DEVICE AND METHOD}

The present invention relates to an apparatus and method for predicting a personal disease based on artificial intelligence, and more particularly, to an apparatus and method for predicting a disease of an individual using an artificial intelligence technique.

Conventionally, in order to identify a disease related to an individual's physical condition, symptoms are searched for on the Internet, or in most cases, people directly go to a hospital and receive medical treatment.

However, identifying diseases according to symptoms through Internet search can be regarded as lacking in accuracy due to lack of expertise. On the other hand, if you go directly to the hospital and receive treatment, most of them have excellent accuracy, but in the case of complex symptoms, there is confusion about exactly which hospital to visit, and depending on the size of the hospital and the ability of medical personnel, confusion arises. There are cases where it is misdiagnosed as some other disease. Therefore, some of them visit other hospitals again to receive treatment.

Therefore, there is a need for a method for identifying diseases that can be expressed through individual symptoms.

The technical problem to be solved by the present invention is to provide an artificial intelligence-based personal disease predicting device and method that receives a symptom of a disease from a user and provides a disease similar to the input symptom.

One aspect of the present invention is an input unit for receiving symptom information from a user; an extractor configured to extract feature information by analyzing the symptom information; a learning unit that learns a plurality of pre-prepared feature information and generates a symptom prediction model so that disease information is matched according to a combination of one or more feature information; One or more disease information matching the feature information is generated based on the symptom prediction model, but when the number of disease information matched to the feature information exceeds a threshold value, one or more additional symptom information is input to correspond to the feature information. Analysis unit requesting; and an output unit outputting the disease information.

The learning unit extracts personal information and disease information from the diagnosis information and generates the symptom prediction model so that the disease information is matched according to the personal information. However, the symptom prediction model may be generated by calculating a frequency at which the disease information is diagnosed according to the personal information based on diagnosis information of different users, and extracting disease information at which the calculated frequency exceeds a threshold value.

The learning unit extracts viewpoint information and disease information from the report information and generates the symptom prediction model so that the disease information is matched according to the viewpoint information. However, the symptom prediction model may be generated by calculating the frequency of disease information in which the time point information is within a preset period based on different news reports, and extracting disease information in which the calculated frequency exceeds a threshold value.

In addition, when the analysis unit analyzes the characteristic information and determines that there is an uninputted item among preset mandatory input items, it may control to receive additional symptom information corresponding to the uninputted item.

In addition, the analysis unit controls to receive additional symptom information indicating the elapsed period when an uninputted item among the mandatory input items is determined to be an elapsed period, receives a unit of the elapsed period first, and receives a period according to the unit can be input.

In addition, the analyzer controls to receive, as the additional symptom information, essential symptom items, pain-related items, and part-related items preset according to the characteristic information when the number of disease information matched with the characteristic information exceeds a threshold value. can

Another aspect of the present invention is a personal disease prediction method in an artificial intelligence-based personal disease prediction device, comprising: receiving symptom information from a user by an input unit; extracting characteristic information by an extractor analyzing the symptom information; generating a symptom prediction model such that disease information is matched according to a combination of one or more feature information by learning a plurality of pre-prepared feature information; An analysis unit generates one or more disease information matching the characteristic information based on the symptom prediction model, and when the number of disease information matched to the characteristic information exceeds a threshold value, one or more additional symptom information corresponding to the characteristic information requesting input of; and outputting the disease information by an output unit.

The learning unit extracts personal information and disease information from the diagnosis information and generates the symptom prediction model so that the disease information is matched according to the personal information. However, the symptom prediction model may be generated by calculating a frequency at which the disease information is diagnosed according to the personal information based on diagnosis information of different users, and extracting disease information at which the calculated frequency exceeds a threshold value.

The learning unit extracts viewpoint information and disease information from the report information and generates the symptom prediction model so that the disease information is matched according to the viewpoint information. However, the symptom prediction model may be generated by calculating the frequency of disease information in which the time point information is within a preset period based on different news reports, and extracting disease information in which the calculated frequency exceeds a threshold value.

In addition, when the analysis unit analyzes the characteristic information and determines that there is an uninputted item among preset mandatory input items, it may control to receive additional symptom information corresponding to the uninputted item.

In addition, the analysis unit controls to receive additional symptom information indicating the elapsed period when an uninputted item among the mandatory input items is determined to be an elapsed period, receives a unit of the elapsed period first, and receives a period according to the unit can be input.

In addition, the analyzer controls to receive, as the additional symptom information, essential symptom items, pain-related items, and part-related items preset according to the characteristic information when the number of disease information matched with the characteristic information exceeds a threshold value. can

According to one aspect of the present invention described above, by providing an artificial intelligence-based individual disease prediction device and method, it is possible to receive a disease symptom from a user and provide a disease similar to the input symptom.

1 is a schematic diagram of a personal disease prediction system including a personal disease prediction device according to an embodiment of the present invention.
2 is a control block diagram of an individual disease prediction device according to an embodiment of the present invention.
3 is a block diagram illustrating a process of extracting feature information by the extractor of FIG. 2 .
4 is a block diagram illustrating a process in which the communication unit of FIG. 2 receives information.
5 and 6 are block diagrams illustrating a process of generating disease information by the analysis unit of FIG. 2 .
7 is a block diagram illustrating a process of outputting disease information by the output unit of FIG. 2 .
8 is a flowchart of a method for predicting individual disease according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS The detailed description of the present invention which follows refers to the accompanying drawings which illustrate, by way of illustration, specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable one skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention are different from each other but are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be implemented in another embodiment without departing from the spirit and scope of the invention in connection with one embodiment. Additionally, it should be understood that the location or arrangement of individual components within each disclosed embodiment may be changed without departing from the spirit and scope of the invention. Accordingly, the detailed description set forth below is not to be taken in a limiting sense, and the scope of the present invention, if properly described, is limited only by the appended claims, along with all equivalents as claimed by those claims. Like reference numbers in the drawings indicate the same or similar function throughout the various aspects.

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

1 is a schematic diagram of a personal disease prediction system including a personal disease prediction device according to an embodiment of the present invention.

The personal disease prediction system 1 may be provided such that the personal disease prediction apparatus 100 receives symptom information from a user and outputs disease information corresponding to the symptom information.

Here, the symptom information may mean a symptom input by a user for a certain disease, and at this time, the personal disease prediction apparatus 100 may receive symptom information in the form of voice or text.

For example, the symptom information may be input as a voice saying 'I have a headache, a swollen neck, and diarrhea', and in this way, the symptom information may refer to information in which a user directly describes his/her symptoms.

Accordingly, the disease information may be information indicating a disease corresponding to the symptom information, and in this case, the disease information may further include information indicating a medical subject or medical institution 200 related to the disease according to the symptom information.

For example, the disease information may be information indicating a disease such as appendicitis or compound fracture, and the disease information may further include information indicating a medical department or medical institution 200 such as internal medicine or orthopedic surgery.

In this regard, the individual disease prediction apparatus 100 may collect diagnosis information from the medical server 210 or report information from the institutional server 310 .

Here, the medical server 210 may refer to a server provided in the medical institution 200 that provides medical treatment, treatment, hospitalization, and observation to users or patients. In this case, the medical institution 200 is a tertiary general hospital, It may include general hospitals, hospitals, nursing homes, clinics and health institutions.

Accordingly, the diagnosis information may refer to a user's disease or information indicating a disease diagnosed by a medical person, and for example, the diagnosis information may include information such as the user's personal information, prescription information, disease information, and the like. .

In other words, the diagnosis information may be information representing a patient's diagnosis process or diagnosis result input from a medical person.

In addition, the diagnosis information may be prepared to indicate information such as examination and treatment performed by a patient at a medical institution. For example, the diagnosis information may further include examination information and treatment information for a user's disease or disease. have.

Here, the examination information may be information indicating a user's disease or a test technique used to diagnose a disease, and the treatment information may be information indicating a user's disease or a treatment technique used to treat a disease.

Also, personal information may refer to information such as a user's name, phone number, resident registration number, e-mail address, and residence address.

Meanwhile, the institution server 310 may mean a server provided in the evaluation institution 300a or the press institution 300b.

Accordingly, the press information may be prepared to indicate the result of evaluation of the medical institution 200 by the evaluation institution 300a, or the press information may be prepared to indicate medical articles written by the media institution 300b. have.

For example, news information may include time information indicating a time point at which news information was written and disease information indicating a disease appearing from the news information.

In this regard, the evaluation institution 300a may refer to an institution that evaluates the level of the medical institution 200 such as services, medical expenses, and salaries provided by the medical institution 200. For example, the evaluation institution 300a ) may mean institutions such as the Health Insurance Review and Assessment Service and the National Health Insurance Corporation.

At this time, the evaluation institution 300a may perform evaluation such as treatment continuity evaluation, prescription evaluation, test evaluation, etc. for each medical institution 200, and the evaluation institution 300a may use different standards for different diseases or diseases. evaluation can be performed.

Here, the treatment continuity evaluation may be performed to indicate the level of the medical institution 200 according to the ratio of patients visiting each period, the number of prescription days, etc. It may be performed to indicate the level of the institution 200, and may be performed to indicate the level of the medical institution 200 according to whether a specific inspection technique is performed.

On the other hand, news information may be prepared to indicate a disease or a dissertation written about a disease.

Accordingly, the personal disease prediction apparatus 100 may generate disease information from symptom information based on diagnosis information or report information.

Hereinafter, an apparatus for predicting a personal disease according to an embodiment of the present invention will be described in more detail.

2 is a control block diagram of an individual disease prediction device according to an embodiment of the present invention.

The personal disease prediction device 100 may include an input unit 110, an extraction unit 120, a communication unit 130, a learning unit 140, an analysis unit 150, and an output unit 160.

In addition, the individual disease predicting apparatus 100 may be implemented with more components than those shown in FIG. 2 or fewer components than those shown in FIG. 2 . Alternatively, in the personal disease prediction apparatus 100, at least two components provided in the personal disease prediction apparatus 100 may be integrated into one component so that one component performs a complex function. Hereinafter, the above-described components will be described in detail.

The input unit 110 may receive symptom information from a user. To this end, the input unit 110 may be provided to enable voice input through a microphone or the like. In addition, the input unit 110 may be provided to enable text input such as a keyboard or touch recognition.

Accordingly, the input unit 110 may receive symptom information in the form of voice or text.

The extraction unit 120 may analyze symptom information and extract feature information. In this case, the characteristic information may mean any word or sentence extracted from voice or text. For example, the extractor 120 may refer to the case where the symptom information is 'headache, swollen throat, diarrhea' , Head, pain, neck, swelling, diarrhea, etc. can be generated as feature information.

In this regard, the extractor 120 may extract feature information from symptom information using a voice recognition technique.

Here, Speech Recognition (STT: Speech To Text) separates the voice signal into tens of milliseconds, extracts a feature vector that appears in 10 dimensions, and converts the voice signal into vowels or consonants through the extracted feature vector. It can be understood as a technique of segmenting, connecting the divided vowels or consonants to generate a random word, and analyzing the generated word to determine the meaning of a voice signal.

In addition, the extraction unit 120 may extract one or more morphemes from a series of sentences, and accordingly, the extraction unit 120 may determine the meaning of one or more morphemes extracted based on a morpheme analysis dictionary prepared in advance. can

Accordingly, the extractor 120 may generate feature information according to a relationship between adjacent morphemes or a frequency in which morphemes appear in a sentence.

To this end, the extraction unit 120 may use a conventionally known morpheme analysis technique, and the extraction unit 120 may use conventionally known syntax analysis and semantic analysis. have.

The communication unit 130 is connected to the medical server 210 through a wireless or wired network to receive diagnosis information, and the communication unit 130 is connected to the institutional server 310 through a wireless or wired network to transmit news information. can receive

The learning unit 140 may learn a plurality of pre-prepared feature information and generate a symptom prediction model so that disease information is matched according to a combination of one or more feature information.

To this end, the learning unit 140 may generate a symptom prediction model using artificial intelligence (AI) techniques such as machine learning.

Here, machine learning can be understood as a technique of generating a learning model to classify a plurality of information into one or more groups based on a plurality of pieces of information, and classifying arbitrary information based on the generated learning model. Similarly, machine learning is supervised learning that creates a learning model to classify arbitrary information according to a plurality of pieces of information classified by a manager, analyzes a plurality of pieces of information itself, or performs a clustering process to perform a learning model Unsupervised learning that generates a learning model, semi-supervised learning that creates a learning model by mixing supervised learning and unsupervised learning, and compensation generated in the process of performing an arbitrary operation on a plurality of information Reinforcement learning that creates a learning model according to this may be included.

Accordingly, the learning unit 140 may generate a symptom prediction model by receiving disease information for a plurality of characteristic information, or the learning unit 140 may predict symptoms by clustering different characteristic information for the disease information. You can also create models.

In one embodiment, the learning unit 140 may generate a symptom prediction model using a long short-term memory (LSTM) technique.

Here, LSTM learns information that changes according to a time series, and estimates current information by applying information estimated at a previous time to the current time, or applying information estimated at the current time to a future time. It may refer to an artificial intelligence technique prepared to predict future information.

Accordingly, the learning unit 140 may generate a symptom prediction model such that disease information is generated based on feature information according to initially input symptom information. At this time, the learning unit 140 may generate an additional symptom that is additionally input. A symptom prediction model may be generated so that previously generated disease information is corrected based on feature information according to the information.

In other words, the symptom prediction model is prepared so that disease information is generated based on feature information according to symptom information, but the symptom prediction model is a disease different from previously generated disease information when feature information according to additional symptom information is input. Arrangements may be made for information to be generated.

In this case, the symptom prediction model may be prepared so that one or more disease information is generated based on the feature information. In this case, the symptom prediction model, when the feature information based on the additional symptom information is input, the one or more previously generated disease information. At least one disease information among the disease information may be provided to be removed.

The analyzer 150 may generate one or more pieces of disease information matched to the feature information based on the symptom prediction model. At this time, the analyzer 150 may determine if the number of pieces of disease information matched to the feature information exceeds a threshold value. Input of one or more additional symptom information to correspond to the feature information may be requested.

Here, requesting the input of additional symptom information can be understood as the output unit 160 outputting text or graphics requesting input of the additional symptom information, and accordingly, the input unit 110 receives additional symptom information from the user. information can be entered.

At this time, the output unit 160 may output a plurality of preset items, and in this case, the input unit 110 may be set to receive selection of one or more items from the user.

Accordingly, the analyzer 150 may use one or more items selected by the user as additional symptom information.

The output unit 160 may output disease information. At this time, the output unit 160 may output disease information so that the user can identify it.

3 is a block diagram illustrating a process of extracting feature information by the extractor of FIG. 2 .

Referring to FIG. 3 , the input unit 110 may receive symptom information from a user. Accordingly, the extractor 120 may extract feature information by analyzing symptom information.

Through this, the learning unit 140 may learn a plurality of previously prepared feature information and generate a symptom prediction model so that disease information is matched according to a combination of one or more feature information.

4 is a block diagram illustrating a process in which the communication unit of FIG. 2 receives information.

The communication unit 130 may be connected to the medical server 210 through a wireless or wired network to receive diagnosis information.

Accordingly, the learning unit 140 may extract personal information and disease information from the diagnosis information and generate a symptom prediction model so that the disease information is matched according to the personal information.

At this time, the learning unit 140 may calculate the frequency at which disease information is diagnosed according to the personal information based on the diagnosis information of different users, and the learning unit 140 detects disease information for which the calculated frequency exceeds a threshold. By extracting, a symptom prediction model can be created.

For example, the learning unit 140 may calculate the frequency at which different disease information is diagnosed according to the age of different users. A high disease can be matched as disease information.

In addition, the learning unit 140 may calculate the frequency at which different disease information is diagnosed according to the gender of different users. can be matched as disease information.

In addition, the learning unit 140 may calculate the frequency at which different disease information is diagnosed according to the residences of different users. can be matched as disease information.

In addition, the learning unit 140 may calculate the frequency at which different disease information is diagnosed according to the time point at which the diagnosis information is generated. Diseases with a high probability can be matched as disease information. In this case, the learning unit 140 may classify the time point at which the diagnosis information is generated as a season.

In this way, the learning unit 140 may match a disease with a high probability of onset as disease information based on the user's personal information, where the personal information includes the user's gender, age, place of residence, diagnosis time, occupation, and the like. can include

In this regard, matching disease information by the learning unit 140 may be understood as first generating disease information having a high weight in the process of generating disease information according to symptom information by assigning a weight to the corresponding disease information.

Meanwhile, the communication unit 130 may be connected to the agency server 310 through a wireless or wired network to receive press information.

Accordingly, the learning unit 140 may extract viewpoint information and disease information from report information and generate a symptom prediction model so that the disease information is matched according to the viewpoint information.

In this case, the learning unit 140 may calculate the frequency of disease information whose viewpoint information is within a preset period based on different news information, and the learning unit 140 may calculate the disease information whose calculated frequency exceeds a threshold. By extracting, a symptom prediction model can be created.

For example, the learning unit 140 may calculate the frequency of disease information written within one month according to the time point information, and in this case, the learning unit 140 matches diseases that have been of great interest to the public for one month as disease information. can make it

In addition, the learning unit 140 may calculate the frequency of disease information written within 10 years according to the time point information, and in this case, the learning unit 140 may match diseases that have been consistently mentioned for 10 years as disease information. have.

5 and 6 are block diagrams illustrating a process of generating disease information by the analysis unit of FIG. 2 .

Referring to FIG. 5 , the input unit 110 may receive symptom information from a user. Accordingly, the extractor 120 may extract feature information by analyzing symptom information.

Accordingly, the analysis unit 150 may generate one or more disease information matching the feature information based on the symptom prediction model.

Meanwhile, referring to FIG. 6 , the analysis unit 150 may request input of one or more additional symptom information to correspond to the characteristic information when the number of disease information matched to the characteristic information exceeds a critical value.

For example, when the feature information is 'neck, pain', the analyzer 150 inputs the appearance of the neck, the degree of swelling, the presence of phlegm, the presence of voice, the presence of pain, the presence of fever, the presence of headache, etc. as additional symptom information. can request

In this regard, the analyzer 150 controls to receive, as additional symptom information, essential symptom items, pain-related items, and region-related items set in advance according to the characteristic information when the number of disease information matched with the characteristic information exceeds a threshold value. can do.

Here, the essential symptom item may refer to additional symptom information requested independently of the characteristic information, and for example, the essential symptom item may include fever, swelling, degree of swelling, bleeding, and the like.

In addition, the pain-related item may refer to additional symptom information for which a type of pain related to a disease region appearing from the characteristic information is requested. For example, the pain-related item includes internal pain, external pain, phlegm, It may include voice transformation and the like.

In addition, the site-related item may refer to additional symptom information for which other risky parts related to the disease occurrence region appearing from the feature information are requested. For example, the site-related item includes headache, heartbeat regularity, and hearing abnormality. , runny nose, etc.

As such, the additional symptom information may include essential symptom items, pain-related items, and site-related items, and such additional symptom information may be set differently according to the meaning of the characteristic information, or the additional symptom information may be characterized It may be set based on disease information generated according to the information.

Meanwhile, the analyzer 150 analyzes the characteristic information and when it is determined that there is an uninputted item among preset essential input items, it may control input of additional symptom information corresponding to the uninputted item.

Here, the required input items may refer to one or more items required to generate disease information, and for example, the required input items may include an elapsed period, a region where pain occurs, whether or not pain is aggravated, and the like.

Accordingly, in one embodiment, the analyzer 150 may control to receive additional symptom information indicating the elapsed period when an uninputted item among the required input items is determined to be an elapsed period, and in this case, the analyzer ( 150) may first receive an input of a unit of the elapsed period, and then, the analyzer 150 may receive an input of a period according to the unit.

7 is a block diagram illustrating a process of outputting disease information by the output unit of FIG. 2 .

Referring to FIG. 7 , the analyzer 150 may generate one or more disease information matching feature information based on the symptom prediction model. At this time, the analyzer 150 may determine the number of disease information matched with feature information. If ? exceeds the threshold, input of one or more additional symptom information may be requested to correspond to the characteristic information.

Accordingly, the output unit 160 may output disease information. At this time, the output unit 160 may output disease information so that the user can identify it.

8 is a flowchart of a method for predicting individual disease according to an embodiment of the present invention.

Since the personal disease prediction method according to an embodiment of the present invention proceeds on substantially the same configuration as the personal disease prediction apparatus 100 shown in FIG. 1, the same components as the personal disease prediction apparatus 100 of FIG. The same reference numerals are given, and repeated descriptions will be omitted.

The individual disease prediction method includes receiving symptom information from a user (800), analyzing symptom information and extracting feature information (810), and generating a symptom prediction model so that disease information is matched according to a combination of one or more feature information. It may include a step 820 of generating one or more disease information matching the characteristic information based on the symptom prediction model (830) and outputting the disease information (840).

Step 800 of receiving symptom information from the user may be a step of receiving symptom information from the user through the input unit 110 .

The step 810 of extracting feature information by analyzing symptom information may be a step in which the extractor 120 analyzes symptom information and extracts feature information.

In step 820 of generating a symptom prediction model so that disease information is matched according to a combination of one or more characteristic information, the learning unit 140 learns a plurality of previously prepared characteristic information, and disease information is obtained according to a combination of one or more characteristic information. It may be a step of generating a symptom prediction model to be matched.

In step 830 of generating one or more disease information matching feature information based on the symptom prediction model, the analysis unit 150 generates one or more disease information matching feature information based on the symptom prediction model, but the feature information It may be a step of requesting input of one or more additional symptom information to correspond to feature information when the number of disease information matched to exceeds a threshold.

The step 840 of outputting the disease information may be a step of outputting the disease information by the output unit 160 .

Such an individual disease prediction method may be implemented as an application or implemented in the form of program instructions that can be executed through various computer components and recorded on a computer-readable recording medium. The computer readable recording medium may include program instructions, data files, data structures, etc. alone or in combination.

Program instructions recorded on the computer-readable recording medium may be those specially designed and configured for the present invention, or those known and usable to those skilled in the art of computer software.

Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks and magnetic tapes, optical recording media such as CD-ROMs and DVDs, and magneto-optical media such as floptical disks. media), and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like.

Examples of program instructions include high-level language codes that can be executed by a computer using an interpreter or the like as well as machine language codes such as those produced by a compiler. The hardware device may be configured to act as one or more software modules to perform processing according to the present invention and vice versa.

Although the above has been described with reference to embodiments, it will be understood that those skilled in the art can variously modify and change the present invention without departing from the spirit and scope of the present invention described in the claims below. You will be able to.

1: Personal disease prediction system
100: personal disease prediction device
200: medical institution
210: medical server
300a: evaluation agency
300b: media organizations
310: institution server

Claims (12)

an input unit that receives symptom information from a user;
an extractor configured to extract feature information by analyzing the symptom information;
a learning unit that learns a plurality of pre-prepared feature information and generates a symptom prediction model so that disease information is matched according to a combination of one or more feature information;
One or more disease information matching the feature information is generated based on the symptom prediction model, but when the number of disease information matched to the feature information exceeds a threshold value, one or more additional symptom information is input to correspond to the feature information. Analysis unit requesting; and
An artificial intelligence-based personal disease prediction device comprising an output unit for outputting the disease information.
According to claim 1,
Further comprising a communication unit receiving diagnosis information from the medical server;
The learning unit,
Personal information and disease information are extracted from the diagnosis information, and the symptom prediction model is generated so that the disease information is matched according to the personal information, but the disease information is diagnosed according to the personal information based on the diagnosis information of different users. An artificial intelligence-based personal disease predicting device that calculates a frequency that occurs, and extracts disease information for which the calculated frequency exceeds a threshold value to generate the symptom prediction model.
According to claim 1,
It further includes; a communication unit that receives press information from the agency server;
The learning unit,
Time information and disease information are extracted from the news information, and the symptom prediction model is generated so that the disease information is matched according to the time information. An artificial intelligence-based personal disease prediction device that calculates a frequency and extracts disease information for which the calculated frequency exceeds a threshold value to generate the symptom prediction model.
The method of claim 1, wherein the analysis unit,
An artificial intelligence-based personal disease prediction device that analyzes the characteristic information and controls to receive additional symptom information corresponding to the uninputted item when it is determined that there is an uninputted item among the pre-set essential input items.
The method of claim 4, wherein the analysis unit,
If an uninputted item among the required input items is determined to be an elapsed period, control to receive additional symptom information representing the elapsed period, wherein the unit of the elapsed period is first input, and the period according to the unit is input, artificial intelligence Based personal disease prediction device.
The method of claim 1, wherein the analysis unit,
When the number of disease information matched to the characteristic information exceeds a threshold value, artificial intelligence-based personal disease control to receive pre-set essential symptom items, pain-related items, and site-related items as the additional symptom information according to the characteristic information. prediction device.
In the personal disease prediction method in an artificial intelligence-based personal disease prediction device,
receiving symptom information from a user through an input unit;
extracting characteristic information by an extractor analyzing the symptom information;
generating a symptom prediction model such that disease information is matched according to a combination of one or more feature information by learning a plurality of previously prepared feature information;
An analysis unit generates one or more disease information matching the characteristic information based on the symptom prediction model, and when the number of disease information matched to the characteristic information exceeds a threshold value, one or more additional symptom information corresponding to the characteristic information requesting input of; and
An individual disease prediction method comprising; outputting the disease information by an output unit.
According to claim 7,
Further comprising a communication unit receiving diagnosis information from the medical server;
The learning unit,
Personal information and disease information are extracted from the diagnosis information, and the symptom prediction model is generated so that the disease information is matched according to the personal information, but the disease information is diagnosed according to the personal information based on the diagnosis information of different users. A method for predicting individual diseases, which calculates a frequency of occurrence, and extracts disease information for which the calculated frequency exceeds a threshold value to generate the symptom prediction model.
According to claim 7,
It further includes; a communication unit that receives press information from the agency server;
The learning unit,
Time information and disease information are extracted from the news information, and the symptom prediction model is generated so that the disease information is matched according to the time information. A method for predicting individual diseases, wherein the symptom prediction model is generated by calculating a frequency and extracting disease information having a calculated frequency exceeding a threshold value.
The method of claim 7, wherein the analysis unit,
A personal disease prediction method for controlling input of additional symptom information corresponding to the uninputted item when it is determined that there is an uninputted item among pre-set essential input items by analyzing the characteristic information.
The method of claim 10, wherein the analysis unit,
When an uninputted item among the required input items is determined to be an elapsed period, control to receive additional symptom information representing the elapsed period, wherein a unit of the elapsed period is first input and a period according to the unit is input. prediction method.
The method of claim 7, wherein the analysis unit,
When the number of disease information matched to the characteristic information exceeds a threshold value, control to receive essential symptom items, pain-related items, and site-related items preset according to the characteristic information as the additional symptom information, personal disease prediction method.

Images