Device, method, and computer program for determining state of user on basis of 6-lead electrocardiogram data

Abstract

The present disclosure provides a device, method, and computer program for determining the state of a user on the basis of 6-lead electrocardiogram data. The method according to one embodiment of the present disclosure comprises the steps of: acquiring 6-lead electrocardiogram data of a user; and determining the state of the user on the basis of the acquired 6-lead electrocardiogram data through a trained neural network model.

Description

Device, method and computer program for determining a user's condition based on 6-induced electrocardiogram data

[0001] The present disclosure relates to deep learning technology in the medical field, and more particularly, to a device, method, and program for determining a user's condition based on 6-lead electrocardiogram data.

[0002] With recent advancements in information and communication technology and deep learning algorithms, the medical field is actively researching technologies that use artificial intelligence to automatically analyze and predict diseases based on a patient's biosignals. In particular, survival rates and prognoses for fatal heart diseases such as myocardial infarction significantly depend on early diagnosis, leading to a growing demand for technologies capable of monitoring and assessing cardiac status in real time. The electrocardiogram (ECG), a signal reflecting the heart's electrophysiological activity, is widely used for assessing heart disease due to its relatively low cost and noninvasive measurement capabilities.

[0003] In particular, the 6-lead ECG, compared to the 12-lead ECG, offers the portability and convenience of a measuring device while also allowing for sufficient analysis of abnormalities in key cardiac areas. This makes it increasingly useful not only in clinical practice but also in routine healthcare and remote medical care. Accordingly, there is a growing need for technological solutions that can more accurately analyze 6-lead ECG data, which can vary depending on the user's diverse biological characteristics and measurement environment, and thereby enable precise diagnosis of heart disease.

[0004] The present disclosure, conceived in response to the aforementioned background technology, aims to provide a device, method, and computer program for determining a user's condition based on 6-lead electrocardiogram data. However, the problems to be solved by the present disclosure are not limited to the problems mentioned above, and other problems not mentioned will be clearly understood based on the description below.

[0005] A method for determining a user's condition based on 6-lead electrocardiogram data, which is performed by a computing device including at least one processor for solving a task as described above, includes a step of acquiring 6-lead electrocardiogram data of the user and a step of determining the user's condition based on the acquired 6-lead electrocardiogram data through a pre-trained neural network model.

[0006] Alternatively, the pre-trained neural network model includes a first neural network model trained to produce a score corresponding to a disease possibility based on the acquired 6-lead electrocardiogram data, and the step of inputting the acquired 6-lead electrocardiogram data into the pre-trained neural network model to determine the user's condition includes the step of inputting the acquired 6-lead electrocardiogram data into the first neural network model to obtain a score corresponding to a disease possibility of the user, and the step of determining the user's condition with respect to the disease based on the obtained score.

[0007] Alternatively, the 6-lead electrocardiogram data includes a plurality of electrocardiogram data each acquired according to a standard 6-lead measurement method, and the step of inputting the acquired 6-lead electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease includes the step of arranging the plurality of electrocardiogram data in a preset order to obtain integrated electrocardiogram data in which the plurality of electrocardiogram data are integrated, and inputting the acquired integrated electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease.

[0008] Alternatively, the pre-learned neural network model includes a 1-1 neural network model corresponding to a first arrangement order of the plurality of electrocardiogram data and a 1-2 neural network model corresponding to a second arrangement order of the plurality of electrocardiogram data, and the step of inputting the acquired integrated electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease includes a step of inputting first integrated electrocardiogram data obtained by arranging the plurality of electrocardiogram data according to the first arrangement order into the 1-1 neural network model to obtain a first score corresponding to the possibility of the user's disease, and a step of inputting second integrated electrocardiogram data obtained by arranging the plurality of electrocardiogram data according to the second arrangement order into the 1-2 neural network model to obtain a second score corresponding to the possibility of the user's disease, and the step of determining the user's condition with respect to the disease based on the acquired score includes a step of determining the user's condition with respect to the disease based on the acquired first score and the second score.

[0009] Alternatively, the first neural network model includes a plurality of sub-neural network models corresponding to the types of the plurality of electrocardiogram data included in the standard 6-lead measurement method, and the step of inputting the acquired 6-lead electrocardiogram data into the first neural network model to determine the user's condition includes a step of inputting the plurality of electrocardiogram data into sub-neural network models corresponding to each type, and obtaining a plurality of scores corresponding to the possibility of the user's disease from the plurality of sub-neural network models, and the step of determining the user's condition with respect to the disease based on the obtained scores includes a step of determining the user's condition with respect to the disease based on the obtained plurality of scores.

[0010] Alternatively, the disease includes left ventricular systolic dysfunction, and the learned neural network model includes a second neural network model learned to calculate left ventricular ejection fraction based on the acquired 6-lead electrocardiogram data, and the step of inputting the acquired 6-lead electrocardiogram data into the second neural network model to acquire the left ventricular ejection fraction of the user, and the step of verifying the result of determining the user's condition regarding left ventricular systolic dysfunction according to the acquired score based on the left ventricular ejection fraction.

[0011] Alternatively, the step of verifying the result of the determination of the user's condition includes a step of determining whether to verify the result of the determination of the user's condition regarding left ventricular systolic dysfunction according to the obtained score based on the reliability corresponding to the obtained score.

[0012] Alternatively, the pre-trained neural network model comprises a plurality of neural network models classified based on at least one of a first factor regarding an underlying disease of the user, a second factor regarding a biological characteristic of the user, a third factor regarding a type of biosignal measuring device that measured the 6-lead electrocardiogram data, and a fourth factor regarding the posture of the user when measuring the 6-lead electrocardiogram data.

[0013] Alternatively, the pre-learned neural network model includes a plurality of neural network models classified according to a first element related to the underlying disease of the user, and the step of determining the user's condition based on the acquired 6-lead electrocardiogram data through the pre-learned neural network model includes a step of identifying a first element corresponding to the user among the plurality of neural network models, a step of identifying a third neural network model corresponding to the identified first element, and a step of inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the user's condition.

[0014] Alternatively, the pre-learned neural network model includes a plurality of neural network models classified according to a second element regarding the biological characteristics of the user, and the step of determining the user's condition based on the acquired 6-lead electrocardiogram data through the pre-learned neural network model includes the steps of identifying a second element corresponding to the user among the plurality of neural network models, identifying a third neural network model corresponding to the identified second element, and inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the user's condition.

[0015] Alternatively, the learned neural network model includes a plurality of neural network models classified according to a third element related to a biosignal measuring device that measured the 6-lead electrocardiogram data, and the step of determining the user's condition based on the acquired 6-lead electrocardiogram data through the learned neural network model includes a step of identifying a third element corresponding to the user among the plurality of neural network models, a step of identifying a third neural network model corresponding to the identified third element, and a step of inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the user's condition.

[0016] Alternatively, the pre-learned neural network model includes a plurality of neural network models classified according to a fourth element regarding the posture of the user when measuring the 6-lead electrocardiogram data, and the step of determining the state of the user based on the acquired 6-lead electrocardiogram data through the pre-learned neural network model includes a step of identifying a fourth element corresponding to the user among the plurality of neural network models, a step of identifying a third neural network model corresponding to the identified fourth element, and a step of inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the state of the user.

[0017] Alternatively, the step of determining the user's condition based on the acquired 6-lead electrocardiogram data through the pre-learned neural network model includes the step of identifying, when a plurality of elements corresponding to the user are identified among the first element, the second element, the third element, and the fourth element, a step of identifying a third neural network model corresponding to an element having the highest priority according to a preset priority among the identified plurality of elements, and a step of inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the user's condition.

[0018] A computing device for determining a user's condition based on 6-lead electrocardiogram data for solving the aforementioned problem includes a processor including at least one core and a memory including program codes executable by the processor, wherein the processor acquires the user's 6-lead electrocardiogram data and determines the user's condition based on the acquired 6-lead electrocardiogram data through a pre-trained neural network model.

[0019] A computer program stored in a computer-readable storage medium for solving the aforementioned problem, wherein the computer program, when executed on one or more processors, performs an operation for determining a user's condition based on 6-lead electrocardiogram data, the operation including an operation for obtaining the user's 6-lead electrocardiogram data and an operation for determining the user's condition based on the obtained 6-lead electrocardiogram data through a pre-learned neural network model.

[0020] According to the present disclosure, a method for determining a user's condition based on 6-lead electrocardiogram data utilizes a neural network model based on 6-lead electrocardiogram data to quantitatively assess whether the user has a heart condition, thereby enabling more accurate and reliable diagnosis. Furthermore, personalized diagnosis is possible by selecting a neural network model based on user characteristics, and the possibility of misdiagnosis can be reduced through reliability-based verification that reflects trends in score changes and discrepancies in judgment between models.

[0021] FIG. 1 is a configuration diagram of a computing device according to an embodiment of the present disclosure.

[0022] FIG. 2 is a flowchart of a method for determining a user's condition based on 6-lead electrocardiogram data according to an embodiment of the present disclosure.

[0023] FIG. 3 is an exemplary diagram of a method for determining a user's condition using a neural network model trained to produce a score corresponding to the possibility of heart disease according to one embodiment of the present disclosure.

[0024] FIG. 4 is an exemplary diagram of a method for determining a user's condition using a plurality of sub-neural network models trained to produce a score corresponding to the possibility of heart disease according to one embodiment of the present disclosure.

[0025] FIG. 5 is an exemplary diagram of a method for determining a user's condition using a neural network model trained to produce a score corresponding to the possibility of heart disease and a neural network model trained to produce a left ventricular ejection fraction according to an embodiment of the present disclosure.

[0026] FIG. 6 is an exemplary diagram of multiple objects classified according to a first element related to a user's underlying disease according to one embodiment of the present disclosure.

[0027] FIG. 7 is a block diagram of a computing device according to another embodiment of the present disclosure.

[0028] Below, embodiments of the present disclosure are described in detail with reference to the attached drawings so that those skilled in the art can easily implement the present disclosure. The embodiments presented in this disclosure are provided to enable those skilled in the art to utilize or implement the contents of the present disclosure. Accordingly, various modifications to the embodiments of the present disclosure will be apparent to those skilled in the art. That is, the present disclosure may be implemented in various different forms and is not limited to the embodiments described below.

[0029] Throughout the specification of this disclosure, identical or similar drawing numbers refer to identical or similar components. Furthermore, for clarity in the description of this disclosure, drawing numbers for parts unrelated to the description of this disclosure may be omitted in the drawings.

[0030] The term "or" as used herein is intended to mean an inclusive "or" rather than an exclusive "or." That is, unless otherwise specified herein or clear from context, "X employs A or B" should be understood to mean either of its natural inclusive permutations. For example, unless otherwise specified herein or clear from context, "X employs A or B" can be interpreted to mean either X employs A, X employs B, or X employs both A and B.

[0031] The term "and / or" as used herein should be understood to refer to and include all possible combinations of one or more of the related concepts listed.

[0032] The terms "comprises" and / or "comprising" as used herein should be understood to mean the presence of certain 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, other components, and / or combinations thereof.

[0033] Unless otherwise specified in this disclosure or unless the context makes it clear that the singular form is being referred to, the singular should generally be construed to include “one or more.”

[0034] The term "Nth (N is a natural number)" used in this disclosure can be understood as an expression used to mutually distinguish components of this disclosure based on a predetermined standard such as a functional perspective, a structural perspective, or convenience of explanation. For example, components performing different functional roles in this disclosure can be distinguished as a first component or a second component. However, components that are substantially the same within the technical spirit of this disclosure but must be distinguished for convenience of explanation may also be distinguished as a first component or a second component.

[0035] The term "acquisition" as used in this disclosure may be understood to mean not only receiving data through a wired or wireless communication network with an external device or system, but also generating data in an on-device form.

[0036] Meanwhile, the term "module" or "unit" used in the present disclosure can be understood as a term referring to an independent functional unit that processes computing resources, such as a computer-related entity, firmware, software or a part thereof, hardware or a part thereof, or a combination of software and hardware. At this time, the "module" or "unit" may be a unit composed of a single element, or a unit expressed as a combination or set of multiple elements. For example, as a narrow concept, a "module" or "unit" may refer to a hardware element of a computing device or a set thereof, an application program that performs a specific function of software, a processing process implemented through software execution, or a set of instructions for program execution, etc. In addition, as a broad concept, a "module" or "unit" may refer to the computing device itself that constitutes the system, or an application running on the computing device, etc. However, since the above-described concept is only an example, the concept of “module” or “part” may be defined in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.

[0037] The term "model" as used herein may be understood as a system implemented using mathematical concepts and language to solve a specific problem, a set of software units to solve a specific problem, or an abstract model of a processing process to solve a specific problem. For example, a neural network "model" may refer to the entire system implemented as a neural network that has problem-solving capabilities through learning. In this case, the neural network can have problem-solving capabilities by optimizing the parameters connecting nodes or neurons through learning. A neural network "model" may include a single neural network or a set of neural networks that are a combination of multiple neural networks.

[0038] The term "data" used in this disclosure may include "images," signals, and the like. The term "image" used in this disclosure may refer to multidimensional data composed of discrete image elements. In other words, "image" may be understood as a term referring to a digital representation of an object visible to the human eye. For example, "image" may refer to multidimensional data composed of elements corresponding to pixels in a two-dimensional image. "Image" may refer to multidimensional data composed of elements corresponding to voxels in a three-dimensional image.

[0039] The explanation of the above terms is intended to aid understanding of the present disclosure. Therefore, unless explicitly stated as limiting the contents of the present disclosure, it should be noted that the above terms are not intended to limit the technical ideas of the present disclosure.

[0040] FIG. 1 is a configuration diagram of a computing device according to an embodiment of the present disclosure.

[0041] Referring to FIG. 1, a computing device (100) according to an embodiment of the present disclosure may include a processor (110), a memory (120), and a communication interface (130). However, FIG. 1 is merely an example, and the computing device (100) may include other components for implementing a computing environment. In addition, only some of the disclosed components may be included in the computing device (100).

[0042] A processor (110) according to an embodiment of the present disclosure may be understood as a configuration unit including hardware and / or software for performing computing operations. For example, the processor (110) may read a computer program to perform data processing for machine learning. The processor (110) may process computational processes such as processing input data for machine learning, feature extraction for machine learning, and error calculation based on backpropagation. The processor (110) for performing such data processing may include a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), a tensor processing unit (TPU), an application specific integrated circuit (ASIC), or a field programmable gate array (FPGA). The above-described type of processor (110) is only one example, and thus, the type of processor (110) may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.

[0043] The processor (110) is connected to other components of the computing device (100) (i.e., memory (120) and communication interface (130)) and controls the overall operation of the computing device (100).

[0044] The memory (120) according to one embodiment of the present disclosure may be understood as a configuration unit including hardware and / or software for storing and managing data processed in the computing device (100). That is, the memory (120) may store any type of data generated or determined by the processor (110) and any type of data received by the communication interface (130). For example, the memory (120) may include at least one type of storage medium among a flash memory type, a hard disk type, a multimedia card micro type, a card type memory, a random access memory (RAM), a static random access memory (SRAM), a read-only memory (ROM), an electrically erasable programmable read-only memory (EEPROM), a programmable read-only memory (PROM), a magnetic memory, a magnetic disk, and an optical disk. In addition, the memory (120) may also include a database system that controls and manages data in a predetermined system. The type of memory (120) described above is only one example, and thus the type of memory (120) can be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.

[0045] The memory (120) can structure and organize and manage data, combinations of data, and program codes executable by the processor (110) required for the processor (110) to perform operations. For example, the memory (120) can store a neural network model learned to identify the user's condition based on electrocardiogram data (20) received through a communication interface (130) to be described later, and can store program codes that operate to perform learning of the neural network model, program codes that operate the neural network model to receive electrocardiogram data (20) and perform inference in accordance with the purpose of use of the computing device (100) (i.e., the purpose of calculating risk), and output or processed data generated as the program code is executed. In addition, the memory (120) can also store learning data for training the neural network model.

[0046] A communication interface (130) according to an embodiment of the present disclosure may be understood as a component that transmits and receives data through any known wired or wireless communication system. For example, the communication interface (130) may perform data transmission and reception using a wired or wireless communication system such as a local area network (LAN), wideband code division multiple access (WCDMA), long term evolution (LTE), wireless broadband internet (WiBro), fifth generation mobile communication (5G), ultrawide-band, ZigBee, radio frequency (RF) communication, wireless LAN, wireless fidelity, near field communication (NFC), or Bluetooth. Since the above-described communication systems are only examples, the wired and wireless communication system for data transmission and reception of the communication interface (130) may be applied in various ways other than the above-described examples.

[0047] The communication interface (130) can receive data necessary for the processor (110) to perform calculations through wired or wireless communication with any system or any client, etc. In addition, the communication interface (130) can transmit data generated through calculations of the processor (110) through wired or wireless communication with any system or any client, etc. For example, the communication interface (130) can receive the user's electrocardiogram data (20) through communication with a biosignal measuring device (200). The communication interface (130) can transmit output data of the neural network model, intermediate data derived from the calculation process of the processor (110), processed data, etc. through communication with the aforementioned database, server, or computing device, etc.

[0048] FIG. 2 is a flowchart of a method for determining a user's condition based on 6-lead electrocardiogram data (20) according to one embodiment of the present disclosure.

[0049] According to one embodiment of the present disclosure, a processor (110) obtains 6-lead electrocardiogram data (20) of a user (S210). Specifically, the processor (110) may obtain 6-lead electrocardiogram data (20) for the user from a biosignal measuring device connected through a communication interface. At this time, the 6-lead electrocardiogram data (20) includes standard limb lead (Lead I, Lead II, and Lead III) data and augmented limb lead (aVR, aVL, and aVF) data. Meanwhile, a plurality of electrocardiogram data (20) included in the 6-lead electrocardiogram data (20) may be data obtained by sampling each electrocardiogram signal measured through the biosignal measuring device at 500 points per second (Sampling rate = 500 Hz) and dividing the data into 8-second-long sections.

[0050] The bio-signal measuring device may be a device that measures the user's electrocardiogram according to a standard 6-lead measurement method to obtain 6-lead electrocardiogram data (20). The bio-signal measuring device may include a plurality of electrodes, and may obtain 6-lead electrocardiogram data (20) by attaching the plurality of electrodes to the user's body. For example, the bio-signal measuring device may include three electrodes, and may be a device that attaches the three electrodes to the user's body at the same time, thereby obtaining electrocardiogram data (20) of Lead ° and Lead ± based on a potential difference between the three electrodes, and then calculates electrocardiogram data (20) of the remaining Leads using the obtained electrocardiogram data (20) (i.e., electrocardiogram data (20) of Lead ° and Lead ±). Alternatively, the biosignal measuring device may be a device (e.g., a smartwatch, etc.) that includes two electrodes and acquires electrocardiogram data (20) corresponding to the attachment location included in the 6-lead electrocardiogram data (20) by attaching the two electrodes to specific locations on the user's body.

[0051] However, the present invention is not limited thereto, and the processor (110) may directly obtain 6-lead electrocardiogram data (20) for the user by using a sensing unit including a plurality of electrodes included in the computing device. The above-described description is equally applicable to the method of directly obtaining 6-lead electrocardiogram data (20), so a detailed description thereof will be omitted.

[0052] Meanwhile, the processor (110) can determine the user's condition based on the 6-lead electrocardiogram data (20) acquired through the pre-trained neural network model (S220). Here, the user's condition may include the user's disease potential (e.g., heart disease potential), stress index, or potential health abnormality symptoms.

[0053] The processor (110) can input 6-lead electrocardiogram data (20) into a pre-trained neural network model to obtain an indicator for determining the user's condition. To this end, the neural network model can be trained based on training data in which the 6-lead electrocardiogram data (20) is input data and an indicator for determining the user's condition is set as a label. In particular, the label may relate to the presence or absence of a disease in the user. For example, the disease may include left ventricular systolic dysfunction (LVSD), acute myocardial infarction, cardiac conduction disturbance, or other cardiovascular abnormalities.

[0054] The processor (110) can input a plurality of 6-lead electrocardiogram data (20) included in the input data into a neural network model, and calculate a loss function based on the difference between the output value of each 6-lead electrocardiogram data (20) of the neural network model and the label data. The loss function can be defined as a cross entropy loss or an objective function that optimizes the balance between precision and recall. Based on the calculated loss function, the processor (110) can adjust the weights of the model through backpropagation. By repeating this process, the processor (110) (110) can improve the classification performance of the neural network model for heart disease, and finally, can obtain a neural network model trained to determine the possibility of a user having a heart disease based on the electrocardiogram data (20). At this time, the neural network model can be implemented with a multi-layer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network, or residual neural network (ResNet) structure.

[0055] The processor (110) inputs the acquired 6-lead electrocardiogram data (20) into a pre-trained neural network model to obtain output information about the user's condition, and can determine the user's condition based on the acquired output information.

[0056] FIG. 3 is an exemplary diagram of a method for determining a user's condition using a neural network model trained to produce a score corresponding to the possibility of heart disease according to one embodiment of the present disclosure.

[0057] According to one embodiment of the present disclosure, a pre-trained neural network model may include a neural network model (hereinafter, a first neural network model (310)) trained to produce a score corresponding to the possibility of heart disease based on feature information of acquired 6-lead electrocardiogram data (20). Here, the score may be a numerical value indicating the possibility of a user's heart disease determined based on feature information extracted from the 6-lead electrocardiogram data (20) (e.g., feature points (Q wave, R wave, etc.) in the electrocardiogram signal, intervals between feature points (RR interval, etc.), waveform shape, amplitude, etc.). The score may be a probability value output from the first neural network model (310) (e.g., a softmax layer of the first neural network model (310)), or may be a value produced by applying a preset weight to the probability value. In addition, the description of the above-described neural network model may be equally applicable to the first neural network model (310).

[0058] The processor (110) can input the acquired 6-lead electrocardiogram data (20) into the first neural network model (310) to obtain a score corresponding to the user's likelihood of having a heart disease. Furthermore, the processor (110) can determine the user's condition regarding heart disease based on the acquired score. In this case, a higher score can be determined to indicate a higher likelihood of the user having a heart disease.

[0059] In particular, the processor (110) can compare the acquired score with a preset reference value to determine the user's condition regarding the possibility of the user's heart disease by classifying it into a preset grade. Here, the preset grade may be a classification of the severity of the user's heart disease. For example, in the case of left ventricular systolic dysfunction, if the acquired score is lower than a first reference value, the user's condition may be determined as normal, if the acquired score is higher than or equal to the first reference value and lower than a second reference value, the user's condition may be determined as low-risk, and if the acquired score is higher than or equal to the second reference value, the user's condition may be determined as high-risk.

[0060] According to one embodiment of the present disclosure, the processor (110) can arrange a plurality of electrocardiogram data (20), i.e., six electrocardiogram data (20), included in the 6-lead electrocardiogram data (20), in a preset order to obtain integrated electrocardiogram data (20) in which the plurality of electrocardiogram data (20) are integrated. Referring to FIG. 3, the processor (110) can arrange six electrocardiogram data (20) included in the 6-lead electrocardiogram data (20) in the order of Lead I, Lead II, Lead III, aVR, aVL, and aVF to obtain integrated electrocardiogram data (20) in a matrix form. In addition, the processor (110) can input the obtained integrated electrocardiogram data (20) into a first neural network model (310) to obtain a score corresponding to the possibility of a user having a heart disease. In this way, the first neural network model (310) can more precisely determine the possibility of a user's heart disease by integrating and analyzing the characteristic information of the six electrocardiogram data (20) included in the six-lead electrocardiogram data (20). To this end, the input data included in the learning data used to train the first neural network model (310) may also be in the form of integrated electrocardiogram data (20) in which the six electrocardiogram data (20) are arranged in a preset order.

[0061] Meanwhile, the first neural network model (310) may include a plurality of neural network models according to the arrangement order of the plurality of electrocardiogram data (20) included in the 6-lead electrocardiogram data (20). That is, the plurality of neural network models included in the first neural network model (310) may learn or perform inference using the plurality of electrocardiogram data (20) arranged in different orders as input data. For example, the first neural network model (310) may include a 1-1 neural network model corresponding to the first arrangement order (the order of Lead I, Lead II, Lead III, aVR, aVL, and aVF) of the plurality of electrocardiogram data (20) and a 1-2 neural network model corresponding to the second arrangement order (the order of Lead I, Lead II, Lead III, aVL, aVR, and aVF) of the plurality of electrocardiogram data (20). The 1-1 neural network model and the 1-2 neural network model can be trained using integrated data having different orders of aVR and aVL data as input data, respectively. The processor (110) inputs the first integrated electrocardiogram data (20) obtained by arranging a plurality of electrocardiogram data (20) according to the first arrangement order (the order of Lead I, Lead II, Lead III, aVR, aVL, and aVF) into the 1-1 neural network model to obtain a first score corresponding to the possibility of the user's heart disease, and inputs the second integrated electrocardiogram data (20) obtained by arranging a plurality of electrocardiogram data (20) according to the second arrangement order (the order of Lead I, Lead II, Lead III, aVL, aVR, and aVF) into the 1-2 neural network model to obtain a second score corresponding to the possibility of the user's heart disease.

[0062] At this time, the processor (110) can determine the user's condition regarding heart disease based on the acquired first score and second score. Specifically, the processor (110) can determine the user's condition by comparing the first score and the second score with a preset reference value. The preset reference value may be set differently for the first score and the second score, and this may be determined according to the sensitivity characteristics of the 1-1 neural network model and the 1-2 neural network model during the learning process or the distribution characteristics of the output scores. If the judgment results regarding the user's condition do not match, the processor (110) can determine the judgment result according to the higher value score as the final judgment result. Alternatively, the processor (110) can calculate the average value of the first score and the second score, and compare the calculated average value with the preset reference value to determine the user's condition regarding the possibility of heart disease.

[0063] FIG. 4 is an exemplary diagram of a method for determining a user's condition using a plurality of sub-neural network models (311 to 316) trained to produce a score corresponding to the possibility of heart disease according to one embodiment of the present disclosure.

[0064] Meanwhile, according to one embodiment of the present disclosure, the first neural network model (310) may include a plurality of sub-neural network models (311 to 316) corresponding to the types of the plurality of electrocardiogram data (20) included in the standard 6-lead measurement method. At this time, the processor (110) inputs six electrocardiogram data (20) included in the 6-lead electrocardiogram data (20), rather than the integrated data of the 6-lead electrocardiogram data (20), into the sub-neural network models corresponding to each type, thereby obtaining a plurality of scores corresponding to the possibility of the user's heart disease from the plurality of sub-neural network models (311 to 316). Specifically, each sub-neural network model may be trained with a plurality of electrocardiogram data (20) corresponding to different types as input data. For example, referring to FIG. 4, the first sub-neural network model can be trained with a plurality of electrocardiogram data (20) corresponding to Lead I as input data, and the second sub-neural network model can be trained with a plurality of electrocardiogram data (20) corresponding to Lead II as input data. The remaining sub-neural network models can also be trained with different types of electrocardiogram data (20) included in the 6-lead electrocardiogram data (20).

[0065] The processor (110) can input six pieces of electrocardiogram data (20) into each sub-neural network model, thereby obtaining multiple scores for the possibility of heart disease from each sub-neural network model. Furthermore, the processor (110) can determine the user's condition regarding heart disease based on the obtained multiple scores.

[0066] For example, the processor (110) may calculate an average value of multiple scores and compare the calculated average value with a preset value to determine the user's heart disease status. At this time, the processor (110) may calculate a standard deviation based on the difference between the multiple scores, remove scores exceeding a preset multiple of the standard deviation, and then determine the average value of the remaining scores as the final score.

[0067] Alternatively, the processor (110) may obtain a judgment result regarding the user's heart disease status by comparing a plurality of scores with scores set corresponding to each sub-neural network model, and may finally determine the user's status by applying a majority of the judgment results. For example, referring again to FIG. 4, if the user's status is determined to be a high-risk group for left ventricular systolic dysfunction based on the scores of the first to third sub-neural network models and the sixth sub-neural network model, and the user's status is determined to be a low-risk group for left ventricular systolic dysfunction based on the scores of the fourth and fifth sub-neural network models, the processor (110) may finally determine the user's status to be a high-risk group for left ventricular systolic dysfunction based on a majority of the judgment results. Meanwhile, the preset reference value may be set differently for the first to sixth sub-neural network models. This may be determined according to the sensitivity characteristics of the first to sixth sub-neural network models during the learning process or the distribution characteristics of the output scores.

[0068] Meanwhile, the processor (110) can obtain multiple scores regarding the possibility of heart disease by using the above-described 1-1 neural network model, the 1-2 neural network model, and multiple sub-neural network models (311 to 316) (1st to 6th sub-neural network models), and can determine the user's condition regarding heart disease by using the obtained multiple scores. Specifically, the processor (110) inputs first integrated electrocardiogram data (20) in which six electrocardiogram data (20) are arranged in a first arrangement order into a first-first neural network model to obtain a first score, inputs second integrated electrocardiogram data (20) in which six electrocardiogram data (20) are arranged in a second arrangement order into a first-second neural network model to obtain a second score, and inputs six electrocardiogram data (20) into a plurality of sub-neural network models (311 to 316) (first to sixth sub-neural network models) to determine the user's heart disease status based on a plurality of scores (for example, third to eighth scores).

[0069] FIG. 5 is an exemplary diagram of a method for determining a user's condition using a neural network model trained to produce a score corresponding to the possibility of heart disease and a neural network model trained to produce a left ventricular ejection fraction according to an embodiment of the present disclosure.

[0070] According to one embodiment of the present disclosure, when the heart disease is left ventricular systolic dysfunction, the pre-trained neural network model may include a neural network model (hereinafter, a 12th neural network model) trained to calculate left ventricular ejection fraction based on feature information of acquired 6-lead electrocardiogram data (20). The processor (110) may input the 6-lead electrocardiogram data (20) into the pre-trained second neural network model to obtain an index regarding left ventricular ejection fraction. To this end, the neural network model may be trained based on training data in which the 6-lead electrocardiogram data (20) is set as input data and the user's left ventricular ejection fraction is set as a label. In particular, the label may be a value of left ventricular ejection fraction measured from the user at the time of acquiring the electrocardiogram data (20) of the input data.

[0071] The processor (110) can input a plurality of 6-lead electrocardiogram data (20) included in the input data into a neural network model, and calculate a loss function based on the difference between the output value of each 6-lead electrocardiogram data (20) of the neural network model and the label data. The loss function can be defined as a cross entropy loss or an objective function that optimizes the balance between precision and recall. Based on the calculated loss function, the processor (110) can adjust the weights of the model through backpropagation. By repeating this process, the processor (110) (110) can improve the left ventricular ejection fraction inference function of the neural network model, and finally, can obtain a neural network model trained to calculate the user's left ventricular ejection fraction based on the electrocardiogram data (20). At this time, the neural network model can be implemented with a multi-layer perceptron (MLP), convolutional neural network (CNN), recurrent neural network (RNN), generative adversarial network, or residual neural network (ResNet) structure.

[0072] The processor (110) can input the acquired 6-lead electrocardiogram data (20) into a second neural network model to acquire the user's left ventricular ejection fraction, and verify the result of determining the user's condition regarding left ventricular systolic dysfunction according to the score acquired based on the left ventricular ejection fraction. In particular, the processor (110) can determine that the user's condition is normal if the left ventricular ejection fraction is less than a preset third value, determine that the user's condition is low risk for left ventricular systolic dysfunction if the left ventricular ejection fraction is equal to or greater than the preset third value and less than the fourth value, and determine that the user's condition is high risk for left ventricular systolic dysfunction if the left ventricular ejection fraction is equal to or greater than the preset fourth value.

[0073] Meanwhile, the second neural network model, like the first neural network model (310), may be input in the form of integrated data in which six pieces of electrocardiogram data (20) included in the six-lead electrocardiogram data (20) are arranged in a preset order, and a plurality of neural network models (i.e., the 2-1st neural network model and the 2-2nd neural network model) may be included depending on the arrangement order. In addition, a plurality of sub-neural network models (311 to 316) (i.e., the 7th to 12th sub-neural network models) into which six pieces of electrocardiogram data (20) included in the 6-lead electrocardiogram data (20) are each input may be included. In this regard, the above-described description may be equally applied.

[0074] The processor (110) may determine whether to verify the result of the determination of the user's condition regarding left ventricular systolic dysfunction according to the obtained score based on the reliability corresponding to the obtained score. The reliability may be evaluated based on the change trend of the obtained score. For example, if the score of the first neural network model (310) is maintained below a preset first value, but the last obtained score is equal to or higher than a second value and is within a preset time from the time when the previous score was calculated, the reliability of the last obtained score may be evaluated as low. The magnitude of the reliability may be calculated based on the amount of change in the last obtained score and the interval between the time when the previous score was calculated and the time when the last score was calculated, and if the reliability is less than the preset value, the processor (110) may determine to verify the result of the determination of the user's condition.

[0075] In addition, when a score includes multiple scores, the reliability can be evaluated based on whether the judgment results based on the multiple scores match. For example, if the judgment results based on the first and second scores obtained from the 1-1 neural network model and the 1-2 neural network model described above do not match, the reliability can be evaluated as low. In particular, when using multiple sub-neural network models (311 to 316), the reliability can be calculated based on the number of scores for which the judgment results do not match or the difference between the scores. In this case, if the reliability is less than a preset value, the processor (110) can determine to verify the judgment result of the user's status.

[0076] In addition, according to one embodiment of the present disclosure, the first neural network model (310) may be configured to simultaneously calculate a probability value (score) corresponding to the possibility of heart disease, along with a confidence level for the probability value. The confidence level may be a value that quantifies the degree of uncertainty or confidence level for the predicted result indicated by the score. For example, the first neural network model (310) may additionally output a confidence value that reflects how internally consistent the predicted result is derived, based on the distribution characteristics of the score calculated based on the softmax function, the variance value of dropout-based inference, etc. The processor (110) may determine to additionally perform verification on the judgment result if the confidence level is less than a preset reference value.

[0077] FIG. 6 is an exemplary diagram of multiple objects classified according to a first element regarding a user's underlying disease according to one embodiment of the present disclosure.

[0078] According to one embodiment of the present disclosure, the pre-trained neural network model may include a plurality of neural network models classified based on at least one of a first factor regarding an underlying disease of the user, a second factor regarding a biological characteristic of the user, a third factor regarding a type of biosignal measuring device that measured 6-lead electrocardiogram data (20), and a fourth factor regarding a posture of the user when measuring 6-lead electrocardiogram data (20).

[0079] Referring to FIG. 6, the processor (110) can identify a plurality of clusters (41, 42, and 43) including a plurality of objects by clustering the objects (and the electrocardiogram data (20) of the objects) according to the type of underlying disease (specifically, the 1-1 element, the 1-2 element, and the 1-3 element) of the plurality of objects within a preset coordinate space (300). Here, the plurality of objects may be measurement targets of the electrocardiogram data constituting the learning data of the neural network model. Similarly, the processor (110) can classify the objects (and the electrocardiogram data (20) of the objects) according to a biological characteristic (e.g., age, sex, weight, height, etc.) (i.e., the second element), the type of the biosignal measuring device that measured the electrocardiogram data (20) (i.e., the third element), and the user's posture (i.e., the fourth element). At this time, clustering according to the first element, the second element, the third element, and the fourth element can be performed within the same coordinate space (300).

[0080] In particular, the processor (110) can identify the type of an element (i.e., at least one of the first to fourth elements) corresponding to a plurality of objects, classify the plurality of objects based on the identified elements, and cluster them within a preset coordinate space. For example, a plurality of objects without an underlying disease of the user can be identified as a first cluster, a plurality of objects whose underlying disease is diabetes (elements 1-2) can be identified as a second cluster, and a plurality of objects whose underlying disease is hypertension (elements 1-2) can be identified as a third cluster. In other words, a plurality of objects corresponding to the same element can be identified within the same cluster.

[0081] Meanwhile, although the first to fourth elements have been described as being applied independently, the present invention is not limited thereto, and the processor (110) may identify clusters for multiple objects by combining at least two of the first to fourth elements. That is, if the underlying disease of a specific object is diabetes (first element), the object is a man in his 60s (second element), and electrocardiogram data is acquired using a smart watch (third element), the object may be classified into the same cluster as another object corresponding to a man in his 60s who measures an electrocardiogram using a smart watch and has diabetes. When multiple elements are combined, the processor (110) may calculate a vector corresponding to the combination of the multiple elements, and perform clustering based on the similarity (and distance) between the calculated vector and the vector of another object.

[0082] The user's posture (fourth factor) can have a direct impact on the electrocardiogram waveform (and the electrocardiogram data (20)). For example, the influence of gravity, changes in the chest structure, the attachment location and contact pressure of the electrodes, etc. may differ between the supine position and the sitting position, which may result in differences in the amplitude of the electrocardiogram signal, the shape of the waveform, or key indicators such as the RR interval and the QT interval. Therefore, the processor (110) classifies the electrocardiogram data (20) based on posture information, and trains a neural network model specialized for each posture, thereby deriving more accurate and reliable heart disease diagnosis results.

[0083] The processor (110) can train each neural network model using electrocardiogram data (20) acquired from the objects included in each cluster. Accordingly, the processor (110) can obtain multiple neural network models trained corresponding to each element (and multiple element combinations).

[0084] For example, referring back to FIG. 6, the processor (110) can obtain a learned neural network model using electrocardiogram data (20) acquired from a plurality of subjects without an underlying disease, can obtain a learned neural network model using electrocardiogram data (20) acquired from a plurality of subjects with diabetes as an underlying disease, and can obtain a learned neural network model using electrocardiogram data (20) acquired from a plurality of subjects with hypertension as an underlying disease.

[0085] Alternatively, the processor (110) may obtain a neural network model learned using electrocardiogram data (20) acquired from a plurality of subjects corresponding to men in their 60s, and the processor (110) may obtain a neural network model acquired using electrocardiogram data (20) acquired from a plurality of subjects corresponding to obese men in their 40s.

[0086] Alternatively, the processor (110) may obtain a learned neural network model using electrocardiogram data (20) obtained from a subject whose electrocardiogram was measured in a lying position, or may obtain a learned neural network model using electrocardiogram data (20) obtained from a subject whose electrocardiogram was measured through a smart watch.

[0087] At this time, according to one embodiment of the present disclosure, the processor (110) may obtain a plurality of neural network models corresponding to each cluster based on the first neural network model (310) that has been previously learned. In this regard, the processor (110) may perform additional learning (e.g., fine-tuning) on ​​the first neural network model (310) that has been previously learned using learning data corresponding to each cluster (specifically, learning data provided as electrocardiogram data obtained from a plurality of subjects included in each cluster), thereby obtaining a plurality of neural network models corresponding to each cluster.

[0088] Specifically, the processor (110) trains the first neural network model (310) using electrocardiogram data collected from all subjects, and then additionally trains the first neural network model (310) for each cluster using cluster-specific training data including electrocardiogram data acquired from multiple subjects included in each cluster. At this time, the additional training may be performed using a fine-tuning method. Alternatively, in addition to a method of readjusting the parameters of the neural network of the first neural network model (310), a method of selectively updating only some layers or fixing the existing model and additionally training only a new classification layer may be included. Through this, the processor can generate a neural network model optimized for the characteristics of each cluster and each element.

[0089] The processor (110) can identify at least one of the user's underlying disease (first element), biological characteristics (second element), a biosignal measuring device used by the user (third element), and the user's posture (fourth element), and select a neural network model corresponding to the at least one identified element from among a plurality of neural network models. The processor (110) can identify a cluster in which the user is included by identifying at least one of the user's first element, second element, third element, and fourth element, and can identify a neural network model corresponding to the identified cluster.

[0090] The processor (110) can obtain related information through a communication interface or an input interface in the case of underlying diseases and biological characteristics, information about a biosignal measuring device can be obtained through metadata of electrocardiogram data (20), and the user's posture can be estimated through the type of the biosignal measuring device or determined based on a sensing value obtained through a sensing unit (e.g., a gyro sensor) of the computing device (100). At this time, if the neural network model includes the first neural network model (310) described above, the processor (110) can input the selected neural network model to obtain a score regarding a heart disease (e.g., possibility of left ventricular systolic dysfunction), and determine the user's condition based on the obtained score. Meanwhile, the reference value set for the neural network model corresponding to each cluster can be set differently, and this can be determined according to the sensitivity characteristics of the neural network model or the distribution characteristics of the output score in the learning process performed based on the electrocardiogram data included in the cluster.

[0091] Meanwhile, if the processor (110) identifies that there are multiple elements corresponding to the user among the first element, the second element, the third element, and the fourth element, the processor (110) may identify a neural network model corresponding to an element with the highest priority according to a preset priority among the identified multiple elements. Specifically, if there are multiple elements identified by the user, the processor (110) may determine the priorities of the multiple elements, and determine a neural network model corresponding to an element with the highest priority according to the priorities. At this time, the priorities may include each element (i.e., the first to fourth elements) independently, or may include a combination of at least two of the multiple elements (i.e., the first to fourth elements). For example, the priorities may be set in the order of the first and second elements, the first element, and the second element.

[0092] FIG. 7 is a block diagram of a computing device (700) according to another embodiment of the present disclosure.

[0093] Referring to FIG. 7, a computing device (700) according to an embodiment of the present disclosure includes a processor (710), a memory (720), a communication interface (730), a sensing unit (740), a display (750), a user interface (760), a camera (770), and a speaker (780). Among the configurations illustrated in FIG. 7, the processor (710), the memory (720), and the communication interface (730) correspond to the configurations of the processor (110) and the memory (120) of the computing device (100) illustrated in FIG. 1, and thus a detailed description thereof will be omitted.

[0094] The sensing unit (740) can obtain electrocardiogram data (20) of the user. For example, the sensing unit (740) can include a plurality of electrodes. In this case, the processor (110) can obtain the user's electrocardiogram data (20) as electrocardiogram data (20) through at least one electrode. Alternatively, the processor can obtain sensing values ​​related to the user's posture by using a gyro sensor, IMU sensor, or the like included in the sensing unit (740).

[0095] The display (750) can display various images. Here, the images include both still images and moving images. The display (750) can output the results of determining the user's condition or output information related to an electrocardiogram (such as an electrocardiogram graph). The display (750) can be implemented as a display in various forms, such as an LCD (Liquid Crystal Display Panel), an OLED (Organic Light Emitting Diodes), an LCoS (Liquid Crystal on Silicon), a DLP (Digital Light Processing), etc. In addition, the display (750) can also include a driving circuit, a backlight unit, etc., which can be implemented in a form, such as an a-TFT, an LTPS (low temperature poly silicon) TFT, an OTFT (organic TFT), etc.

[0096] Meanwhile, the display (750) may be implemented as a touch screen by being combined with a touch panel, and in this case, the display (750) may perform the function of not only an output interface that outputs an image through the touch screen, but also an input interface that receives a user's touch input.

[0097] The user interface (760) is a component used by the computing device (700) to perform interaction with the user, and may include, but is not limited to, at least one of a touch sensor, a motion sensor, a button, a jog dial, and a switch. The processor (710) may receive input of the user's underlying disease and biological information through the user interface (760).

[0098] The camera (770) captures images of objects surrounding the computing device (700). Specifically, the camera (770) can capture images of the user. At this time, the processor (710) can identify the user's underlying disease based on the captured images. To this end, the camera (770) may be implemented with an imaging device such as a CMOS image sensor (CIS) having a CMOS structure, a charge coupled device (CCD) having a CCD structure, etc. However, the present invention is not limited thereto, and the camera (770) may be implemented with camera modules of various resolutions capable of capturing subjects.

[0099] Meanwhile, the camera (770) may be implemented as a depth camera (e.g., an IR depth camera), a stereo camera, or an RGB camera.

[0100] The speaker (780) is a component that outputs various audio data that have undergone various processing operations, such as decoding, amplification, and noise filtering, by an audio processing unit (not shown). The speaker (780) can output various notification sounds or voice messages. According to one embodiment of the present disclosure, the processor (710) can convert an electrical signal received from an external device into a user's voice and output it through the speaker (780). For example, the speaker (780) can output a judgment result regarding the user's status in the form of a voice message.

[0101] The various embodiments of the present disclosure described above can be combined with additional embodiments and modified within the scope understood by those skilled in the art in light of the detailed description above. It should be understood that the embodiments of the present disclosure are illustrative in all respects and not restrictive. For example, each component described as a single component may be implemented in a distributed manner, and likewise, components described as distributed may be implemented in a combined manner. Accordingly, all changes or modifications derived from the meaning, scope, and equivalent concepts of the claims of the present disclosure should be construed as being included within the scope of the present disclosure.

Claims

1. A method for determining a user's condition based on 6-lead electrocardiogram data performed by a computing device including at least one processor, Step of acquiring the user's 6-lead electrocardiogram data; and A step of determining the user's condition based on the acquired 6-induced electrocardiogram data through a pre-learned neural network model; including; method.

2. In paragraph 1, The above-mentioned pre-trained neural network model is, A first neural network model learned to produce a score corresponding to the possibility of a disease based on the above-mentioned acquired 6-lead electrocardiogram data, The step of inputting the above-obtained 6-lead electrocardiogram data into a pre-trained neural network model to determine the user's condition is as follows: A step of inputting the above-obtained 6-lead electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease; and A step of determining the user's condition regarding the disease based on the obtained score; including; method.

3. In paragraph 2, The above 6-lead electrocardiogram data is, Contains multiple electrocardiogram data each acquired according to the standard 6-lead measurement method, The step of inputting the above-obtained 6-lead electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease is as follows: A step of arranging the plurality of electrocardiogram data in a preset order to obtain integrated electrocardiogram data in which the plurality of electrocardiogram data are integrated, and inputting the obtained integrated electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease; including; method.

4. In paragraph 3, The above-mentioned pre-trained neural network model is, It includes a 1-1 neural network model corresponding to the first arrangement order of the plurality of electrocardiogram data and a 1-2 neural network model corresponding to the second arrangement order of the plurality of electrocardiogram data, The step of inputting the acquired integrated electrocardiogram data into the first neural network model to obtain a score corresponding to the possibility of the user's disease is as follows: A step of inputting first integrated electrocardiogram data obtained by arranging the plurality of electrocardiogram data according to the first arrangement order into the 1-1 neural network model to obtain a first score corresponding to the possibility of the user's disease, and a step of inputting second integrated electrocardiogram data obtained by arranging the plurality of electrocardiogram data according to the second arrangement order into the 1-2 neural network model to obtain a second score corresponding to the possibility of the user's disease; including, The step of determining the user's condition regarding the disease based on the obtained score is as follows: A step of determining the user's condition regarding the disease based on the first score and the second score obtained above; comprising, method 5. In paragraph 2, The above first neural network model is, Includes a plurality of sub-neural network models corresponding to the types of electrocardiogram data included in the above standard 6-induced measurement method, The step of inputting the above-obtained 6-lead electrocardiogram data into the first neural network model to determine the user's condition is as follows: A step of inputting the plurality of electrocardiogram data into a sub-neural network model corresponding to each type, and obtaining a plurality of scores corresponding to the possibility of the user's disease from the plurality of sub-neural network models; The step of determining the user's condition regarding the disease based on the obtained score is as follows: A step of determining the user's condition regarding the disease based on the acquired plurality of scores; method.

6. In paragraph 2, The above disease is, Including left ventricular systolic dysfunction, The above-mentioned pre-trained neural network model is, A second neural network model trained to calculate left ventricular ejection fraction based on the above-mentioned acquired 6-lead electrocardiogram data is included. A step of inputting the above-obtained 6-lead electrocardiogram data into the second neural network model to obtain the user's left ventricular ejection fraction; and A step of verifying the result of the judgment of the user's condition regarding the left ventricular systolic dysfunction according to the obtained score based on the left ventricular ejection fraction; including; method.

7. In paragraph 6, The step of verifying the result of the judgment of the above user's status is as follows: A step of determining whether to verify the result of the judgment of the user's condition regarding the left ventricular systolic dysfunction according to the obtained score based on the reliability corresponding to the obtained score; including; method.

8. In paragraph 1, The above-mentioned pre-trained neural network model is, A plurality of neural network models classified based on at least one of a first factor regarding the user's underlying disease, a second factor regarding the user's biological characteristics, a third factor regarding the type of biosignal measuring device that measured the 6-lead electrocardiogram data, and a fourth factor regarding the user's posture when measuring the 6-lead electrocardiogram data. method.

9. In paragraph 1, The above-mentioned pre-trained neural network model is, Comprising a plurality of neural network models classified according to a first factor regarding the user's underlying disease, The step of judging the user's condition based on the 6-induced electrocardiogram data obtained above through the learned neural network model is as follows. A step of identifying a first element corresponding to the user among the plurality of neural network models, and identifying a third neural network model corresponding to the identified first element; and A step of inputting the obtained 6-induced electrocardiogram data into the third neural network model to determine the user's condition; method.

10. In paragraph 1, The above-mentioned pre-trained neural network model is, comprising a plurality of neural network models classified according to a second factor regarding the biological characteristics of the user; The step of judging the user's condition based on the 6-induced electrocardiogram data obtained above through the learned neural network model is as follows. A step of identifying a second element corresponding to the user among the plurality of neural network models, and identifying a third neural network model corresponding to the identified second element; and A step of inputting the acquired 6-lead electrocardiogram data into the third neural network model to determine the user's condition; method.

11. In paragraph 1, The above-mentioned pre-trained neural network model is, A plurality of neural network models classified according to a third element of a biosignal measuring device that measures the above 6-induced electrocardiogram data, The step of judging the user's condition based on the 6-induced electrocardiogram data obtained above through the learned neural network model is as follows. A step of identifying a third element corresponding to the user among the plurality of neural network models, and identifying a third neural network model corresponding to the identified third element; and A step of inputting the obtained 6-induced electrocardiogram data into the third neural network model to determine the user's condition; method.

12. In paragraph 1, The above-mentioned pre-trained neural network model is, Including multiple neural network models classified according to a fourth factor regarding the user's posture when measuring the above 6-induced electrocardiogram data, The step of judging the user's condition based on the 6-induced electrocardiogram data obtained above through the learned neural network model is as follows. A step of identifying a fourth element corresponding to the user among the plurality of neural network models, and identifying a third neural network model corresponding to the identified fourth element; and A step of inputting the obtained 6-induced electrocardiogram data into the third neural network model to determine the user's condition; method.

13. In paragraph 8, The step of judging the user's condition based on the 6-induced electrocardiogram data obtained above through the learned neural network model is as follows. When a plurality of elements corresponding to the user are identified among the first element, the second element, the third element, and the fourth element, a step of identifying a third neural network model corresponding to an element with the highest priority according to a preset priority among the identified plurality of elements; and A step of inputting the obtained 6-induced electrocardiogram data into the third neural network model to determine the user's condition; method. 14.6 In a computing device that determines the user's condition based on 6-Lead electrocardiogram data, a processor comprising at least one core; and a memory including program codes executable by the processor; The above processor, Obtaining the user's 6-lead electrocardiogram data, and determining the user's condition based on the obtained 6-lead electrocardiogram data through a pre-trained neural network model. Computing device.

15. A computer program stored in a computer-readable storage medium, wherein the computer program, when executed on one or more processors, performs an operation of determining the user's condition based on 6-Lead electrocardiogram data. The above action is, The operation of acquiring the user's 6-lead electrocardiogram data; and An operation of determining a user's condition based on the acquired 6-induced electrocardiogram data through a pre-learned neural network model; including; Computer program.

Images