KR20250089712A - Apparatus and method for early detecting chronic diseases based on lifestyle pattern information - Google Patents

Abstract

The present invention relates to a method for predicting the possibility of developing a chronic disease or detecting its onset at an early stage by obtaining lifestyle pattern information through questionnaires and simple surveys without the user having to visit a hospital or specialized examination institution and analyzing the information using an artificial neural network.

Description

{APPARATUS AND METHOD FOR EARLY DETECTING CHRONIC DISEASES BASED ON LIFESTYLE PATTERN INFORMATION}

The present invention relates to a method for predicting the possibility of developing a chronic disease or detecting its onset at an early stage by obtaining lifestyle pattern information through questionnaires and simple surveys without the user having to visit a hospital or specialized examination institution and analyzing the information using an artificial neural network.

As the number of patients with hypertension, diabetes, and dyslipidemia, which are considered the three major chronic diseases, continues to increase, a systematic management strategy is required.

Diagnosing chronic diseases requires invasive blood and urine tests, and for busy modern people, it is inconvenient to take time out of their day to visit a hospital and undergo these tests regularly, which can delay the diagnosis of the disease.

According to a prior art for solving such a problem, a device for predicting the occurrence of three major chronic diseases receives data for predicting the occurrence of each chronic disease from a user and confirms whether or not each chronic disease has occurred from the input data.

That is, the user directly performed a physical examination and questionnaire, and then entered the results of the examination and questionnaire into the prediction device, and the prediction device confirmed the onset of the disease from the results of the examination and questionnaire using a prediction model.

However, since this predicts the occurrence of chronic diseases based only on the health status on the day of the examination and the results of the questionnaire answered by the user on that day, there is a problem in that the user's condition information cannot be accurately obtained because the user may not be aware of his or her long-term lifestyle patterns or may enter incorrect answers due to remembering them differently.

Therefore, there is a need for a method to accurately obtain lifestyle pattern information through online questionnaires and simple surveys without users having to visit hospitals or specialized examination institutions periodically, and analyze this information using an artificial neural network to predict the possibility of developing chronic diseases or detect their onset early.

The present invention has been created in consideration of the above circumstances, and the purpose of the present invention is to enable users to accurately obtain lifestyle pattern information through online questionnaires and simple surveys without visiting a hospital or specialized examination institution, and to analyze the information using an artificial neural network to predict the possibility of occurrence of a chronic disease before it occurs or to detect the onset of the disease at an early stage.

According to one embodiment of the present invention for achieving the above object, a method for early detection of chronic diseases based on an artificial neural network may include a database storage step for storing a questionnaire database including a list of queries for screening for chronic diseases and a personalized database for each user; a query output step for outputting a query selected from the questionnaire database; a response receiving step for receiving a response to the query; and a chronic disease prediction step for predicting a chronic disease of the user based on an artificial neural network model using the personalized database and the response to the query.

An early detection device for chronic diseases based on an artificial neural network according to one embodiment of the present invention may include a storage unit including a questionnaire database including a list of queries for screening for chronic diseases and a personalized database for each user, a query unit for outputting queries selected from the questionnaire database, a receiving unit for receiving responses to the queries, and a prediction unit for predicting a chronic disease of the user based on an artificial neural network model using the personalized database and the responses to the queries.

According to embodiments of the present invention, users can accurately obtain lifestyle pattern information through online questionnaires and simple surveys without visiting a hospital or specialized examination institution, and analyze the information using an artificial neural network to predict the possibility of developing a chronic disease before it occurs or detect the onset of the disease at an early stage.

According to embodiments of the present invention, by utilizing a user's daily life log data regarding exercise or food intake, etc., in a chronic disease prediction algorithm, specific data utilization is possible, and thus the probability of chronic disease prediction can be increased.

FIG. 1 illustrates a block diagram of a chronic disease prediction system (100) according to one embodiment of the present disclosure.
FIG. 2 illustrates a block diagram of a chronic disease prediction system according to another embodiment of the present disclosure.
FIG. 3 illustrates an artificial neural network model for predicting chronic diseases according to one embodiment of the present invention.
FIG. 4 is a block diagram showing a customized healthcare service system according to one embodiment of the present invention.
FIG. 5 is a flowchart of a chronic disease prediction method according to another embodiment of the present disclosure.

Various embodiments are now described with reference to the drawings. In this specification, various descriptions are set forth to provide an understanding of the present disclosure. However, it will be apparent that these embodiments may be practiced without these specific descriptions.

The terms "component," "module," "system," and the like, as used herein, refer to a computer-related entity, hardware, firmware, software, a combination of software and hardware, or an execution of software. For example, a component may be, but is not limited to, a procedure running on a processor, a processor, an object, a thread of execution, a program, and/or a computer. For example, an application running on a computing device and the computing device may both be components. One or more components may reside within a processor and/or a thread of execution. A component may be localized within a single computer. A component may be distributed between two or more computers. Furthermore, such components may execute from various computer-readable media having various data structures stored therein. The components may communicate via local and/or remote processes, for example, by a signal having one or more data packets (e.g., data from one component interacting with another component in a local system, a distributed system, and/or data transmitted via a network such as the Internet to another system via the signal).

Additionally, the term "or" is intended to mean an inclusive "or" rather than an exclusive "or." That is, unless otherwise specified or clear from the context, "X employs A or B" is intended to mean either of the natural inclusive permutations. That is, if X employs A; X employs B; or X employs both A and B, "X employs A or B" can apply to any of these cases. Furthermore, the term "and/or" as used herein should be understood to refer to and include all possible combinations of one or more of the associated items listed.

Also, the terms "comprises" and/or "comprising" should be understood to mean the presence of the features and/or components. However, it should be understood that the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other features, components, and/or groups thereof. Also, unless otherwise specified or clear from the context to refer to the singular form, the singular form as used in the specification and claims should generally be construed to mean "one or more."

And, the term "at least one of A or B" should be interpreted to mean "including only A", "including only B", or "combined in the composition of A and B".

Those skilled in the art should additionally recognize that the various illustrative logical blocks, configurations, modules, circuits, means, logics, and algorithm steps described in connection with the embodiments disclosed herein may be implemented as combinations of electronic hardware, computer software, or both. To clearly illustrate the interchangeability of hardware and software, various illustrative components, blocks, configurations, means, logics, modules, circuits, and steps have been described above generally in terms of their functionality. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the overall system. Skilled artisans may implement the described functionality in varying ways for each particular application. However, such implementation decisions should not be interpreted as causing a departure from the scope of the present disclosure.

The description of the disclosed embodiments is provided to enable a person skilled in the art to make or use the present invention. Various modifications to these embodiments will be apparent to a person skilled in the art. The general principles defined herein may be applied to other embodiments without departing from the scope of the present disclosure. Thus, the present invention is not limited to the embodiments disclosed herein. The present invention is to be accorded the widest scope consistent with the principles and novel features disclosed herein.

In the present disclosure, network function and artificial neural network and neural network can be used interchangeably.

FIG. 1 illustrates a block diagram of a chronic disease prediction system (100) according to one embodiment of the present disclosure.

Referring to FIG. 1, a chronic disease prediction system (100) according to one embodiment of the present disclosure includes a storage unit (110) that stores a questionnaire database including a list of questions for screening for chronic diseases and a personalized database for each user, a query unit (120) that outputs questions selected from the questionnaire database, a receiving unit (130) that receives responses to the questions, and a prediction unit (140) that predicts a chronic disease of the user using the personalized database and the responses to the questions.

Specifically, the storage unit (110) of the chronic disease prediction system (100) of the present disclosure stores a questionnaire database including a list of questions for screening for chronic diseases and a personalized database for each user.

Here, the questionnaire database may be a set of data including a list of questions consisting of questionnaire items used in hospitals, medical facilities, counseling centers (220 in FIG. 4), etc. to screen for chronic diseases, and the personalized database may be a set of data including information about individual users using the chronic disease prediction system of the present disclosure.

The query unit (120) of the chronic disease prediction system (100) of the present disclosure outputs a query selected from a questionnaire database. The query unit (120) can extract a query from a query list included in the questionnaire database to screen for a chronic disease and output the extracted query.

As an example, the query unit (120) may use a display to display a query selected from a survey database and output it visually to the user. As another example, the query unit (120) may use a speaker to output a query selected from a survey database audibly. The method by which the query unit (120) outputs a query is not limited thereto and may include other methods that may notify the user.

The receiving unit (130) of the chronic disease prediction system (100) of the present disclosure receives a response to a query. As an example, the receiving unit (130) may receive a response to a query output through a display. That is, a user may respond to a query by inputting a predetermined signal to the display or by touching it. As another example, the receiving unit (130) may receive a response to a query by receiving a user's voice.

Here, the query may include a first questionnaire for obtaining user information, a second questionnaire for obtaining lifestyle pattern information, and a third questionnaire for obtaining disease history information.

The first questionnaire may include questions about at least one of the following: nationality, age, gender, height, weight, occupation, education level, residential area, work area, and family members.

The second questionnaire may include questions about at least one of the following: total number of steps per day, water intake, number of times urinated, number of meals, total calories consumed, sugar intake, amount and frequency of smoking/drinking, total working hours, total sleeping hours, and total exercise hours.

The third questionnaire may include at least one question regarding past medical history and family history.

Here, the user's personal information, including the user's gender, age, and type of residence, can be registered when the chronic disease prediction system starts.

In another embodiment, the personalized database may include the user's hospital diagnosis information. The receiving unit (130) may receive the user's hospital diagnosis information from a server storing hospital or disease-related diagnosis information, or may receive hospital diagnosis information input by the user, and the storage unit (110) may store the received hospital diagnosis information in the personalized database.

In one embodiment, the questionnaire database stored in the storage unit (110) includes a list of questions including authorized chronic disease questionnaire items for measuring chronic diseases, and the query unit (120) can select a query from the questionnaire database based on the personalized database and output the selected query.

For example, if information about a user stored in a personalized database indicates that the user has diabetes and similar symptoms, or if the user's hospital diagnosis information stored in the personalized database includes diabetes, the query unit (120) can output a query belonging to a questionnaire item related to a diabetes diagnosis scale from the questionnaire database.

For example, the inquiry department (120) can ask questions in the form of voice or text through the device to obtain answers to one of the questionnaire items of the diabetes diagnosis scale, such as 'How many liters of water did you drink per day, how many times did you urinate, and how many kilograms did you lose weight?'

The prediction unit (140) of the chronic disease prediction system (100) of the present disclosure predicts the user's chronic disease based on an artificial neural network using a personalized database and responses to queries.

Chronic diseases predicted by the chronic disease prediction system (100) of the present disclosure include, but are not limited to, the three major chronic diseases of diabetes, hypertension, and chronic kidney disease, and may include various chronic diseases that can be examined at a hospital.

In one embodiment, the prediction unit (140) can predict a user's chronic disease by combining the user's prior information stored in a personalized database and responses to queries.

FIG. 2 illustrates a block diagram of a chronic disease prediction system according to another embodiment of the present disclosure. The chronic disease prediction system of FIG. 2 includes the configuration of the chronic disease prediction system (100) of FIG. 1, and additionally includes a processing unit (210). The configurations identical or corresponding to those in FIG. 1 also indicate the same or corresponding configurations in FIG. 2, and since the configurations are indicated by the same serial numbers, redundant descriptions are omitted for convenience of explanation, and the description focuses on the additional configurations.

Referring to FIG. 2, the receiving unit (130) of the chronic disease prediction system (100) according to another embodiment of the present disclosure can receive biosignals and physical activity information from the user's wearable device (10) for a predetermined period of time while simultaneously receiving a response to a query.

In addition, the chronic disease prediction system (100) of the present disclosure may further include a processing unit (210) that processes biosignals and physical activity information received from the user's wearable device (10) by the receiving unit (130) to generate the user's prior information.

The storage unit (110) can store the user's dictionary information generated by the processing unit (210) in a personalized database.

Specifically, the receiving unit (130) receives bio-signals and physical activity information from the user's wearable device (10) for a predetermined period of time, at the same time as or before the query unit (120) outputs a query selected from the questionnaire database and the receiving unit (130) receives a response to the query, and the chronic disease prediction system can collect the received real-time bio-signals and physical activity information of the user.

As an example, the receiving unit (130) receives at least one of a heart rate, an electrocardiogram, a body temperature, and a blood oxygen saturation level as a bio-signal through the user's wearable device (10), and receives at least one of the number of steps and the jogging time as physical activity information through the user's wearable device, and analyzes the bio-signal and physical activity information to generate user's prior information, and the storage unit (110) can store the user's prior information generated by the processing unit (210) in a personalized database.

Here, the wearable device (10) may be a smartwatch, smartphone, smart glasses, HMD (head mount display), band, etc. that the user wears or carries, and any device that can be worn or carried by the user and obtain user bio-signals and physical activity information may be employed.

The chronic disease prediction system of the present disclosure utilizes biosignals and physical activity information collected over a certain period of time for more accurate diagnosis.

In one embodiment, the query unit (120) can output a query selected from the questionnaire database periodically or when a predetermined condition is satisfied. The chronic disease prediction system (100) of the present disclosure can periodically query the user in daily life. For example, the query unit (120) can output a selected query at a specific time.

In one embodiment, the query unit (120) can output a selected query when a pre-configured application starts operating. For example, if the pre-configured application is a portal site app, the query unit (120) can output a selected query when a user runs the portal site app.

And the receiving unit (130) receives the user's response to the query and input data including bio-signal and physical activity information, and the prediction unit (140) can predict the user's chronic disease based on an artificial neural network using the input data.

In one embodiment, the prediction unit (140) can predict a user's chronic disease through a fully connected neural network based on input data including question and answer, bio-signals, and physical activity information. The prediction process of the prediction unit is described in detail below based on FIG. 3.

The prediction unit (140) predicts the user's chronic disease based on an artificial neural network.

FIG. 3 illustrates an artificial neural network model for predicting chronic diseases according to one embodiment of the present invention. Here, the artificial neural network model according to the present invention may be a combined form of a convolutional neural network (CNN) model and a recurrent neural network (RNN) model.

In the case of the convolutional neural network (CNN) model, classification is performed using the feature values included in the current input data without considering past information, so it may not be suitable for predicting aspects that change over time. However, the convolutional neural network (CNN) model is suitable for learning hierarchical and abstract representations for inputs related to performing specific tasks.

In addition, the recurrent neural network (RNN) model is suitable for learning time-series correlations because it includes a recurrent connection structure to reflect past output values to current input data operations.

Accordingly, according to the present invention, input data acquired through a questionnaire and a wearable device can be expressed as a vector, and an artificial neural network model combining a CNN model and an RNN model can be trained so that the input vector and the chronic disease prediction value match each other, and the error can be minimized by calculating the result value using the classified values through the trained artificial neural network model.

The artificial neural network model learned using learning data according to the above-described process can predict the user's chronic disease by receiving response data to the user's query, bio-signals obtained through a wearable device, and physical activity information. This will be explained later with reference to the drawings.

Referring to FIG. 3, a predetermined number of data can be extracted as input data from responses to a query, bio-signals, and physical activity information.

Responses to the questions may include responses to each of the first questionnaire question for obtaining user information, the second questionnaire question for obtaining lifestyle pattern information, and the third questionnaire question for obtaining disease history information, as well as biosignals and physical activity information obtained through a wearable device.

Responses to the first questionnaire may include responses to at least one of the following questions: nationality, age, gender, height, weight, occupation, education level, residential area, work area, and family members.

Responses to the second questionnaire may include responses to at least one of the following: total number of steps per day, water intake, number of meals, total calories consumed, sugar intake, amount and frequency of smoking/drinking, total working hours, total sleeping hours, and total exercise hours.

Responses to the third questionnaire may include responses to at least one question regarding past medical history and family history.

The biosignal may include at least one of heart rate, electrocardiogram, body temperature, and blood oxygen saturation, and the physical activity information may include at least one of the number of steps and jogging time.

The artificial neural network model inputs input data into a convolutional neural network model (501). Specifically, the convolutional neural network model (501) may include a convolutional neural network corresponding to the number of input data extracted from a response to a query, a biosignal, and physical activity information, and a plurality of input data may be input into each convolutional neural network (501-1, 501-2, . . . 501-N). Here, the input data may be composed of text or numbers.

That is, the first convolutional neural network (501-1) can receive the first input text (t0), the second convolutional neural network (501-2) can receive the second input text (t1), and the Nth convolutional neural network (501-N) can receive the Nth input text (tN).

And, each of the plurality of convolutional neural networks (501-1, .. 501-N) can analyze the input text and extract features of the input text. That is, each of the plurality of convolutional neural networks (501-1, .. 501-N) can include a plurality of layers, and each layer can receive the input text and process the input data of the corresponding layer to generate output data.

Specifically, each of the plurality of convolutional neural networks (501-1, . . . 501-N) may include a convolutional layer that outputs a feature map as output data through a convolutional operation on the input text. These convolutional early layers may be operated to extract low-level features from the input, and may be operated to extract more complex features, such as spatial features of conductive signals associated with chronic diseases, as they go to higher levels of measurement.

In addition, the convolutional neural network (501-1, ..501-N) may include a pooling layer in which a pooling operation is performed in addition to a convolutional layer in which a convolution operation is performed. Here, the pooling layer may perform a pooling operation to reduce the spatial size of data in the corresponding region, such as max pooling that selects the maximum value in the corresponding region or average pooling that selects the average value in the corresponding region.

Additionally, the convolutional neural network (501-1, ..501-N) may include a fully-connected layer that converts each layer into a one-dimensional vector and connects the layers converted into one-dimensional vectors into one vector. Here, the fully-connected layer may determine a classification.

The convolutional neural network (501-1, .. 501-N) according to the present invention can extract chronic disease spatial features related to chronic diseases of input text through a convolutional layer, a pooling layer, and a fully connected layer.

An artificial neural network model according to the present invention can learn a correlation between a response to a first questionnaire question, a response to a second questionnaire question, a response to a third questionnaire question, a biosignal and physical activity information value, and a chronic disease.

As an additional example, the output of the convolutional neural network model (501) may be input to the recurrent neural network model (502). The recurrent neural network model (502) may include a recurrent neural network corresponding to the number of the plurality of convolutional neural networks (501-1, . . . 501-N), and each of the outputs of the plurality of convolutional neural networks (501-1, . . . 501-N) may be input to each of the recurrent neural networks (502-1, . . . 502-N). That is, the first recurrent neural network (502-1) may receive the output of the first convolutional neural network (501-1), the second recurrent neural network (502-2) may receive the output of the second convolutional neural network (501-2), and the Nth recurrent neural network (502-N) may receive the output of the Nth convolutional neural network (501-N).

Here, a recurrent neural network (RNN) is an artificial neural network technique that simultaneously considers current data and past data, and can simultaneously consider current data and past data by forming a directed cycle between the connections between units that make up the artificial neural network as well as the output of the convolutional neural network (501).

This recurrent neural network model (502) can extract blood sugar features according to time series changes in chronic disease spatial features by learning chronic disease spatial features of each of multiple input texts extracted through multiple convolutional neural networks (501-1,...501-N) in time order.

Specifically, each of the plurality of recurrent neural networks (502-1, . . . 502-N) can extract a chronic disease feature vector representing the time-series change of the chronic disease spatial features of each of the plurality of input texts extracted through the plurality of convolutional neural networks (501-1, . . . 501-N). In addition, the recurrent neural network (502-N) can extract the chronic disease features produced by synthesizing the extracted chronic disease features.

And, when chronic disease features are extracted from each of the multiple input texts, the recurrent neural network model (502) can predict chronic disease onset based on the chronic disease features (S250). Here, the predicted chronic disease onset information can be a value corresponding to the probability of chronic disease onset and the severity of the chronic disease.

Here, different artificial neural networks are trained according to the disease corresponding to the chronic disease, and the probability of contracting each disease or the severity of the disease can be obtained as a result using the trained artificial neural networks.

Meanwhile, according to the example described above, the recurrent neural network model (502) was described as an example in which RNN was used, but according to another implementation example of the present invention, it may be implemented with LSTM (Long Short-Term Memory).

According to the present invention, by presenting an artificial neural network model learned to predict a user's chronic disease based on extracted features, the onset of a user's chronic disease can be predicted more easily and quickly.

That is, the prediction unit (140) can predict the user's chronic disease based on the probability of the chronic disease returned through the artificial neural network.

In another embodiment, the prediction unit (140) may generate an input vector representing the user's health status based on responses to each questionnaire, and may also diagnose the user's chronic disease at an early stage based on the similarity between the generated input vector and the disease vector.

For example, the prediction unit (140) may generate an input vector for each questionnaire item based on the distribution of scores for the first questionnaire item to the third questionnaire item, and may generate an input vector by synthesizing the generated vectors based on the weights for each questionnaire item.

Here, the weights may vary depending on the demographic characteristics and disease history of the subject.

Meanwhile, the chronic disease prediction system (100) may further include a notification unit that notifies the user of the results of the prediction unit's prediction of the user's chronic disease.

For example, if the probability of a chronic disease returned from the prediction unit (140) exceeds a predetermined value or is determined to be higher than a predetermined ratio compared to the probability of a chronic disease in the past, and thus it is predicted that a chronic disease exists or is at risk, the notification unit can inform the user of this conclusion.

That is, if an abnormality is detected in the user, the user can be guided to receive professional treatment. For example, a notification is given to the user, such as, 'There are signs of chronic kidney disease' or 'This may be an early symptom of diabetes.' The notification unit can display the results of the prediction unit (140) to the user through a display, or can notify the user by outputting them through voice.

FIG. 4 is a block diagram showing a customized healthcare service system according to one embodiment of the present invention.

Referring to FIG. 4, the customized healthcare service system may include a wearable device (10), a user device (100), a chronic disease prediction server (210), and a medical institution server (220). Here, the chronic disease prediction server (210) corresponds to the chronic disease prediction system described above.

The wearable device (10) can be worn or attached to the user's body to obtain the user's bio-signals and physical activity information. Here, the wearable device (10) can be a watch type as shown in FIG. 4. However, it is not limited thereto, and the wearable device (10) can have various forms such as a bracelet, a band, a watch, a ring, a necklace, an anklet, a belt, an HMD, etc.

The wearable device (10) can acquire bio-signals and physical activity information, and input the acquired bio-signals and physical activity information into an artificial neural network model. The specific operations of the chronic disease prediction server (210) for learning the artificial neural network model and predicting the onset of chronic diseases using the learned model have been described above, so a detailed description thereof will be omitted.

FIG. 5 is a flowchart of a chronic disease prediction method according to another embodiment of the present disclosure.

Referring to FIG. 5, a chronic disease prediction method according to an embodiment of the present disclosure includes a database storage step (S910) for storing a questionnaire database including a list of queries for screening for a chronic disease and a personalized database for each user, a query output step (S920) for outputting a query selected from the questionnaire database, a response receiving step (S930) for receiving a response to the query and bio-signal and physical activity information through a wearable device, and a chronic disease prediction step (S940) for predicting a chronic disease of the user using the personalized database and the response to the query.

Here, the chronic disease prediction method may be a method performed by various types of electronic devices, such as smartphones, Bluetooth speakers, wearable devices, etc., which can store a database, output queries to users, and receive responses from users.

In the database storage step (S910), a questionnaire database including a list of questions for screening for chronic diseases and a personalized database for each user are stored.

Here, the questionnaire database may be a set of data including a list of questions consisting of questionnaire items used in hospitals, medical facilities, counseling centers (220 in FIG. 4), etc. to screen for chronic diseases, and the personalized database may be a set of data including information about individual users using the chronic disease prediction system of the present disclosure.

In the query output step (S920), the query selected from the questionnaire database can be displayed using a display, so as to be visually output to the user. As another example, the query selected from the questionnaire database can be output audibly using a speaker. The method of outputting the query is not limited thereto, and may include other methods that can notify the user.

The response receiving step (S930) receives a response to a query. As an example, a response to a query output through a display can be received. That is, a user can respond to a query by inputting a predetermined signal to the display or by touching it. As another example, a response to a query can be received by inputting a user's voice.

Here, the query may include a first questionnaire for obtaining user information, a second questionnaire for obtaining lifestyle pattern information, and a third questionnaire for obtaining disease history information.

The first questionnaire may include questions about at least one of the following: nationality, age, gender, height, weight, occupation, education level, residential area, work area, and family members.

The second questionnaire may include questions about at least one of the following: total number of steps per day, water intake, number of times urinated, number of meals, total calories consumed, sugar intake, amount and frequency of smoking/drinking, total working hours, total sleeping hours, and total exercise hours.

The third questionnaire may include at least one question regarding past medical history and family history.

Here, the user's personal information, including the user's gender, age, and type of residence, can be registered when the chronic disease prediction system starts.

In another embodiment, the personalized database may include the user's hospital diagnosis information. The user's hospital diagnosis information may be received from a server that stores hospital or disease-related diagnosis information, or the hospital diagnosis information may be received as input by the user.

The chronic disease prediction step (S940) predicts the user's chronic disease based on an artificial neural network using a personalized database and responses to queries.

Chronic diseases predicted by the chronic disease prediction system of the present disclosure include, but are not limited to, the three major chronic diseases of diabetes, hypertension, and chronic kidney disease, and may include various chronic diseases that can be examined at a hospital.

In one embodiment, the user's chronic disease can be predicted based on an artificial neural network model by combining the user's prior information stored in a personalized database and responses to queries.

Here, the artificial neural network model can be a combination of a convolutional neural network (CNN) model and a recurrent neural network (RNN) model.

In the case of the convolutional neural network (CNN) model, classification is performed using the feature values included in the current input data without considering past information, so it may not be suitable for predicting aspects that change over time. However, the convolutional neural network (CNN) model is suitable for learning hierarchical and abstract representations for inputs related to performing specific tasks.

In addition, the recurrent neural network (RNN) model is suitable for learning time-series correlations because it includes a recurrent connection structure to reflect past output values to current input data operations.

Accordingly, according to the present invention, input data acquired through a questionnaire and a wearable device can be expressed as a vector, and an artificial neural network model combining a CNN model and an RNN model can be trained so that the input vector and the chronic disease prediction value match each other, and the error can be minimized by calculating the result value using the classified values through the trained artificial neural network model.

An artificial neural network model trained using learning data according to the above-described process can predict a user's chronic disease by receiving response data to a user's query, biosignals obtained through a wearable device, and physical activity information.

According to embodiments of the present invention, users can accurately obtain lifestyle pattern information through online questionnaires and simple surveys without visiting a hospital or specialized examination institution, and analyze the information using an artificial neural network to predict the possibility of developing a chronic disease before it occurs or detect the onset of the disease at an early stage.

Meanwhile, the functional operations and subject matter implementations described in this specification may be implemented as digital electronic circuits, or may be implemented as computer software, firmware or hardware including the structures disclosed in this specification and their structural equivalents, or may be implemented as a combination of one or more of these. The subject matter implementations described in this specification may be implemented as one or more computer program products, that is, one or more modules regarding computer program instructions encoded on a tangible program storage medium for processing the operations of a processing system or for execution by the same.

The computer-readable medium may be a machine-readable storage device, a machine-readable storage substrate, a memory device, or a combination of one or more of these.

The term "system" or "device" in this specification encompasses any apparatus, device or machine for processing data, including, for example, a programmable processor, a computer or a multiprocessor or a computer. In addition to hardware, a processing system may include code that forms an execution environment for a computer program upon request, such as code constituting processor firmware, a protocol stack, a database management system, an operating system or a combination of one or more of these.

A computer program (also known as a program, software, software application, script, or code) may be written in any form of a programming language, including compiled or interpreted languages, a priori or procedural languages, and may be deployed in any form, including as a stand-alone program, a module, a component, a subroutine, or other unit suitable for use in a computing environment. A computer program does not necessarily correspond to a file in a file system. A program may be stored in a single file provided to a requested program, in multiple interacting files (e.g., a file storing one or more modules, subprograms, or portions of code), or in part of a file containing other programs or data (e.g., one or more scripts stored in a markup language document). A computer program may be deployed to be executed on a single computer or on multiple computers located at a single site or distributed across multiple sites and interconnected by a communications network.

Meanwhile, a computer-readable medium suitable for storing computer program instructions and data may include all forms of non-volatile memory, media and memory devices, including semiconductor memory devices such as EPROM, EEPROM and flash memory devices, magnetic disks such as internal hard disks or external disks, magneto-optical disks and CD-ROM and DVD-ROM disks. The processor and memory may be supplemented by, or incorporated into, special purpose logic circuitry.

Implementations of the subject matter described herein may be implemented in a computing system that includes a back-end component, such as a data server, a middleware component, such as an application server, a front-end component, such as a client computer having a web browser or a graphical user interface via which a user can interact with an implementation of the subject matter described herein, or any combination of one or more of such back-end, middleware or front-end components. The components of the system may be interconnected by any form or medium of digital data communication, such as a communications network.

While this specification contains details of a number of specific implementations, these should not be construed as limitations on the scope of any invention or what may be claimed, but rather as descriptions of features that may be unique to particular embodiments of a particular invention. Likewise, particular features described herein in the context of separate embodiments may also be implemented in combination in a single embodiment. Conversely, various features described in the context of a single embodiment may also be implemented in multiple embodiments, either individually or in any suitable subcombination. Furthermore, although features may operate in a particular combination and may initially be described as being claimed as such, one or more features from a claimed combination may in some cases be excluded from that combination, and the claimed combination may be modified into a subcombination or variation of a subcombination.

Also, although this specification depicts operations in the drawings in a particular order, this should not be understood to imply that such operations must be performed in the particular order depicted or in any sequential order to achieve desirable results, or that all depicted operations must be performed. In certain instances, multitasking and parallel processing may be advantageous. Also, the separation of the various system components of the embodiments described above should not be understood to require such separation in all embodiments, and it should be understood that the program components and systems described may generally be integrated together in a single software product or packaged into multiple software products.

As such, this specification is not intended to limit the invention to the specific terms presented. Accordingly, although the invention has been described in detail with reference to the above-described examples, those skilled in the art can make modifications, changes, and variations to the examples without departing from the scope of the invention. The scope of the invention is indicated by the claims described below rather than the above-described detailed description, and all changes or modifications derived from the meaning and scope of the claims and their equivalents should be interpreted as being included in the scope of the invention.

Claims (10)

A database storage step for storing a questionnaire database including a list of questions for screening for chronic diseases and a personalized database for each user;
A query output step for outputting queries selected from the above questionnaire database;
A response receiving step for receiving a response to the above query; and
An artificial neural network-based early detection method for chronic diseases, comprising a chronic disease prediction step of predicting a user's chronic disease based on an artificial neural network model using the personalized database and responses to the query. In paragraph 1,
The above chronic disease prediction steps are:
A step of receiving bio-signals and physical activity information from the user's wearable device for a predetermined period of time while receiving a response to the above query; and
Further comprising a pre-information generation step of generating pre-information of the user by analyzing the user's bio-signals and physical activity information,
The above database storage step is,
An artificial neural network-based method for early detection of chronic diseases, which stores the user's prior information in the personalized database. In the first paragraph,
The above survey database is,
Includes a list of questions containing validated chronic disease questionnaire items to measure chronic diseases;
The above query output step is,
An artificial neural network-based early detection method for chronic diseases, which selects a query from the questionnaire database based on the personalized database and outputs the selected query. In the first paragraph,
The above query output step is,
An artificial neural network-based method for early detection of chronic diseases, which outputs a query selected from the above-mentioned questionnaire database periodically or when a predetermined condition is satisfied. In paragraph 1,
The above question is,
It includes a first questionnaire to obtain user information, a second questionnaire to obtain lifestyle pattern information, and a third questionnaire to obtain disease history information.
The first questionnaire above is,
Includes a questionnaire on at least one of the following: nationality, age, gender, height, weight, occupation, education level, residential area, work area, and family members.
The second questionnaire above is,
Includes a questionnaire on at least one of the following: total number of steps per day, water intake, number of meals, total calories consumed, sugar intake, amount and frequency of smoking/drinking, total working hours, total sleeping hours, and total exercise hours.
The third questionnaire above is,
An artificial neural network-based method for early detection of chronic diseases, comprising at least one questionnaire regarding past medical history and family history. In paragraph 1,
Further comprising a step of training the artificial neural network,
The above artificial neural network expresses input data obtained through the above questionnaire and the wearable device as a vector, and is an artificial neural network-based chronic disease early detection method in which a CNN model and an RNN model are combined so that the input vector and the chronic disease prediction value are matched with each other. In paragraph 1,
The above chronic disease prediction steps are:
A step of generating an input vector representing the health status of the user based on the response to each of the above questionnaires; and
A method for early detection of chronic diseases based on lifestyle pattern information, comprising a step of early diagnosing a chronic disease of the user based on the similarity between the generated input vector and the disease vector. In paragraph 7,
The steps for generating the above input vector are:
A step of generating a vector for each questionnaire item based on the distribution of scores of the first questionnaire item to the third questionnaire item; and
Including a step of generating the state vector by synthesizing the generated vector based on the weight for each of the questionnaire items,
The above weights are based on lifestyle pattern information, and are an early detection method for chronic diseases, which varies depending on the demographic characteristics and disease history of the diagnosed person. A storage unit including a questionnaire database containing a list of questions for screening for chronic diseases and a personalized database for each user;
A query section for outputting queries selected from the above questionnaire database;
A receiving unit for receiving a response to the above query; and
An artificial neural network-based chronic disease early detection device, comprising a prediction unit that predicts a user's chronic disease based on an artificial neural network model using the personalized database and responses to the query. In Article 9,
The above prediction part,
An artificial neural network-based chronic disease early detection device that receives a response to the above query, simultaneously receives bio-signals and physical activity information from the user's wearable device for a predetermined period of time, analyzes the user's bio-signals and physical activity information to generate prior information about the user, and stores the user's prior information in the personalized database.