WO2024101954A1 - 생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 - Google Patents
생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 Download PDFInfo
- Publication number
- WO2024101954A1 WO2024101954A1 PCT/KR2023/018074 KR2023018074W WO2024101954A1 WO 2024101954 A1 WO2024101954 A1 WO 2024101954A1 KR 2023018074 W KR2023018074 W KR 2023018074W WO 2024101954 A1 WO2024101954 A1 WO 2024101954A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- disease
- user
- information
- neural network
- type
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Ceased
Links
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/358—Detecting ST segments
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/24—Detecting, measuring or recording bioelectric or biomagnetic signals of the body or parts thereof
- A61B5/316—Modalities, i.e. specific diagnostic methods
- A61B5/318—Heart-related electrical modalities, e.g. electrocardiography [ECG]
- A61B5/346—Analysis of electrocardiograms
- A61B5/349—Detecting specific parameters of the electrocardiograph cycle
- A61B5/364—Detecting abnormal ECG interval, e.g. extrasystoles, ectopic heartbeats
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
- G06N20/20—Ensemble learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/047—Probabilistic or stochastic networks
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/09—Supervised learning
-
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N7/00—Computing arrangements based on specific mathematical models
- G06N7/01—Probabilistic graphical models, e.g. probabilistic networks
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/20—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for computer-aided diagnosis, e.g. based on medical expert systems
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/70—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for mining of medical data, e.g. analysing previous cases of other patients
-
- G—PHYSICS
- G16—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR SPECIFIC APPLICATION FIELDS
- G16H—HEALTHCARE INFORMATICS, i.e. INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR THE HANDLING OR PROCESSING OF MEDICAL OR HEALTHCARE DATA
- G16H50/00—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics
- G16H50/30—ICT specially adapted for medical diagnosis, medical simulation or medical data mining; ICT specially adapted for detecting, monitoring or modelling epidemics or pandemics for calculating health indices; for individual health risk assessment
Definitions
- the present disclosure relates to an electronic device that diagnoses a user's disease based on biological signals and a method of controlling the same. More specifically, it relates to an electronic device that diagnoses a user's disease using a previously learned neural network model and a control method thereof.
- Myocardial infarction and ischemic heart disease are the most common causes of sudden death, and recently, the number of patients with myocardial infarction and ischemic heart disease is increasing every year. In the case of myocardial infarction and ischemic heart disease, early detection and treatment can prevent death and disability, so many studies are being conducted on early diagnosis and advance prediction of myocardial infarction and ischemic heart disease.
- the present disclosure was created in response to the above-mentioned background technology, and the problem to be solved by the present disclosure is to provide an electronic device that diagnoses a user's disease based on a pre-learned neural network model and the user's biological signals and a method of controlling the same. will be.
- a method performed by an electronic device for realizing the above-described problem is disclosed.
- the method includes obtaining biometric data of a user, inputting the acquired biometric data into a previously learned first neural network model to generate first information about the first disease of the user, and generating first information about the first disease of the user.
- the step of generating the diagnosis result may include adjusting a first probability value included in the first information based on a second probability value included in the second information, and adjusting the first probability value included in the first information based on the first probability value. This may include generating a diagnosis result of the user regarding the first disease.
- generating the diagnostic result may include, if the second probability value is greater than or equal to a reference value, adjusting the first probability value by applying a weight corresponding to the second probability value to the first probability value. and if the second probability value is less than the reference value, maintaining the first probability value.
- generating the diagnostic result may include generating a first diagnostic result corresponding to the first disease if the first probability value is greater than or equal to the first value, and the first probability value is greater than or equal to the first value. If the first probability value is less than the second value, generate a second diagnosis result corresponding to the first disease, and if the first probability value is less than the second value, generate a third diagnosis result corresponding to the first disease. It may include steps to:
- the first information may be information regarding the presence or absence of the first disease
- the second information may be information regarding the first type of the first disease.
- the method may include extracting first and second neural network models from among a plurality of neural network models, based on the first disease and the diagnosis result type.
- the step of generating information about the presence or absence of the first disease and the step of generating information about the first type may be performed in parallel.
- the first and second neural network models are pre-trained based on learning data to which a plurality of labels are respectively set to different standards for the same ECG signal, and the plurality of labels are determined by determining whether the first disease exists or not. It may include a first type of label corresponding to and a second type of label corresponding to the first type of the first disease.
- the biometric data includes an electrocardiogram signal
- the first disease includes any one of myocardial infarction and ischemic heart disease
- the first type is ST segment elevation myocardial infarction (STEMI: ST segment elevation myocardial infarction).
- An electronic device is disclosed according to an embodiment of the present disclosure for realizing the above-described problem.
- the electronic device acquires a communication interface, a memory storing previously learned first and second neural network models, and biometric data of the user, and inputs the obtained biometric data into the previously learned first neural network model to create the user's first neural network model.
- Generate first information about the disease input the acquired biometric data into a pre-trained second neural network model to generate second information about the user's first disease, and generate second information about the user's first disease based on the first and second information.
- it includes one or more processors that generate the user's diagnosis result regarding the first type of the first disease.
- a non-transitory computer-readable recording medium that stores computer instructions that, when executed by a processor of an electronic device, cause an operation of the electronic device to be performed, wherein the operation is performed on the user's biological device.
- Acquiring data inputting the acquired biometric data into a pre-trained first neural network model to generate first information about the user's first disease, applying the acquired biometric data to a pre-trained second neural network Generating second information about the first disease of the user by inputting it into a model, and generating a diagnosis result of the user about the first type of the first disease based on the first and second information.
- an electronic device can learn a plurality of neural network models to generate information about the presence or absence of a disease and information about the type of a disease using the same learning data, and thus the learning data is secured and pre-processed. The cost and time required to do this can be reduced.
- FIG. 1 is an exemplary diagram of an author device according to an embodiment of the present disclosure.
- Figure 2 is a block diagram of an electronic device according to an embodiment of the present disclosure.
- FIG. 3 is a flowchart schematically showing a method of controlling an electronic device according to an embodiment of the present disclosure.
- FIG. 4 is an exemplary diagram showing previously learned first and second neural network models stored in an electronic device according to an embodiment of the present disclosure.
- Figure 5a is an exemplary diagram showing a method of learning the first and second neural network models (20-1 and 20-2) according to an embodiment of the present disclosure
- Figure 5b is a diagram showing conventional information and specific information on the presence or absence of a specific disease.
- This is an example diagram showing a method of learning multiple neural network models to generate information about a specific type of disease.
- Figure 6 is an exemplary diagram illustrating a method of diagnosing a user's disease using a first and second neural network model according to an embodiment of the present disclosure.
- FIG. 7 is a detailed configuration diagram of an electronic device according to an alternative embodiment of the present disclosure.
- the term “or” is intended to mean an inclusive “or” and not an exclusive “or.” That is, unless otherwise specified in the present disclosure or the meaning is not clear from the context, “X uses A or B” should be understood to mean one of natural implicit substitutions. For example, unless otherwise specified in the present disclosure or the meaning is not clear from the context, “X uses A or B” means that It can be interpreted as one of the cases where all B is used.
- the term “at least one of A or B” should be interpreted to refer to all of A, B, and a combination of A and B.
- N is a natural number
- N is a natural number
- components performing different functional roles may be distinguished as first components or second components.
- components that are substantially the same within the technical spirit of the present disclosure but must be distinguished for convenience of explanation may also be distinguished as first components or second components.
- module refers to a computer-related entity, firmware, software or part thereof, hardware or part thereof.
- the “module” or “unit” can be understood as a term referring to an independent functional unit that processes computing resources, such as a combination of software and hardware.
- the “module” or “unit” may be a unit composed of a single element, or may be a unit expressed as a combination or set of multiple elements.
- a “module” or “part” in the narrow sense is a hardware element or set of elements of an electronic device, an application program that performs a specific function of software, a process implemented through software execution, or a program execution. It may refer to a set of instructions for .
- module or “unit” may refer to the electronic device itself constituting the system, or an application running on the electronic device.
- module or “unit” may be defined in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- model refers to a system implemented using mathematical concepts and language to solve a specific problem, a set of software units to solve a specific problem, or a process to solve a specific problem. It can be understood as an abstract model of a process.
- a neural network “model” may refer to an overall system implemented as a neural network that has problem-solving capabilities through learning. At this time, the neural network can have problem-solving capabilities by optimizing parameters connecting nodes or neurons through learning.
- a neural network “model” may include a single neural network or a neural network set in which multiple neural networks are combined.
- FIG. 1 is an exemplary diagram of an electronic device according to an embodiment of the present disclosure.
- the electronic device 100 acquires the user's biosignal 10 from an external electronic device 200 that cooperates with the electronic device 100.
- external electronic devices 200-1 and 200-2, hereinafter 200
- the electronic device 100 can be implemented with various devices that perform the function of measuring the biological signals 10 (10-1 and 10-2, hereinafter referred to as 10).
- the external electronic device 200 may be implemented as an electrocardiogram measurement device, smart watch, display device, etc.
- the external electronic device 200 provides information about the external electronic device 200 (or information about the organization where the external electronic device 200 is located) in advance to the electronic device 100. You can also register at .
- the electronic device 100 may generate a user diagnosis result based on a user signal obtained from the external electronic device 200. Specifically, the electronic device 100 may analyze the user signal obtained from the external electronic device 200 to determine the user's health status or identify the presence or absence of the user's disease and the type of disease. In particular, in order to generate a user's diagnosis result, the electronic device 100 may use a previously learned neural network model 20. As an example, the electronic device 100 inputs the acquired biosignal 10 into the previously learned neural network model 20, such as user's health status information or user's disease information (e.g., presence or absence of disease or type of disease). Information including, etc.) can be obtained as the output value of the previously learned neural network model 20. To this end, the pre-trained neural network model 20 may be trained in advance to output the user's health status information or the user's disease information based on various user's biosignals 10.
- the pre-trained neural network model 20 may be trained in advance to output the user's health status information or the user's disease information
- the electronic device 100 may transmit the generated diagnosis result to the external electronic device 200 or to a user terminal device.
- the user can determine health status information simply by measuring biosignals 10 without the help of a specialized facility or expert.
- FIG. 2 is a block diagram of an electronic device 100 according to an embodiment of the present disclosure.
- the electronic device 100 may be a hardware device or part of a hardware device that performs comprehensive processing and calculation of data, or may be a software-based computing environment connected to a communication network.
- the electronic device 100 may be a server (e.g., a platform server, etc.) that performs an intensive data processing function and shares resources, or a client (e.g., a platform server, etc.) that shares resources through interaction with the server. It may be a client).
- the electronic device 100 may be a cloud system that allows a plurality of servers and clients to interact and comprehensively process data.
- the type of the electronic device 100 may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- the electronic device 100 may be implemented as an electronic device 100 that measures a user's biosignals (eg, electrocardiogram signals, etc.).
- a user's biosignals eg, electrocardiogram signals, etc.
- the electronic device 100 of the present disclosure will be described assuming that it is a server.
- the electronic device 100 may include a memory 110, a communication interface (network unit) 120, and one or more processors 130.
- the electronic device 100 may include other components for implementing a computing environment. Additionally, only some of the disclosed configurations may be included in the electronic device 100.
- the memory 110 may be understood as a structural unit including hardware and/or software for storing and managing data processed in the electronic device 100. That is, the memory 110 can store any type of data generated or determined by the processor 130 and any type of data received by the processor 130 through the communication interface 120.
- the memory 110 may store a plurality of previously learned neural network models. Each neural network model may be matched with at least one other neural network model according to the type of disease being diagnosed and the type of diagnostic result being generated and stored in the memory 110. At this time, a plurality of matched neural network models may be trained to output different types of information based on the same input (eg, biosignal, etc.).
- the first and second neural network models may be stored in the memory 110.
- the first and second neural network models may be learned in advance based on the same learning data and then stored in the memory 110, and the previously learned first and second neural network models may be used to analyze the user's biological signals and Each can be learned to generate different types of information from signals.
- training data used to train the first and second neural network models may be stored in the memory 110.
- the memory 110 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory, or a random access RAM (RAM).
- memory SRAM (static random access memory), ROM (read-only memory), EEPROM (electrically erasable programmable read-only memory), PROM (programmable read-only memory), magnetic It may include at least one type of storage medium among memory, magnetic disk, or optical disk.
- the memory 110 may include a database system that controls and manages data in a predetermined system. Since the type of memory 110 described above is only an example, the type of memory 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.
- the communication interface 120 may be understood as a structural unit that transmits and receives data through any type of known wired or wireless communication system.
- the electronic device 100 can transmit and receive various information to and from the external electronic device 100 through the communication interface 120.
- the electronic device 100 may receive the user's biosignal measured by the external electronic device 100 through the communication interface 120, or the electronic device 100 may generate One user's diagnosis result information may be transmitted to the external electronic device 100.
- the communication interface 120 is a local area network (LAN), wideband code division multiple access (WCDMA), long term evolution (LTE), and wireless broadband (WiBro). internet), 5th generation mobile communication (5G), ultrawide-band wireless communication, ZigBee, radio frequency (RF) communication, wireless LAN, wireless fidelity, Data transmission and reception can be performed using a wired or wireless communication system such as near field communication (NFC) or Bluetooth. Since the above-described communication systems are only examples, the wired and wireless communication systems for data transmission and reception of the communication interface 120 may be applied in various ways other than the above-described examples.
- LAN local area network
- WCDMA wideband code division multiple access
- LTE long term evolution
- WiBro wireless broadband
- 5G 5th generation mobile communication
- ultrawide-band wireless communication ZigBee
- RF radio frequency
- NFC near field communication
- Bluetooth Bluetooth
- the processor 130 is electrically connected to the memory 110 and the communication interface 120 and can control the overall operation of the electronic device 100.
- the processor 130 may be understood as a structural unit that includes hardware and/or software for performing computing operations.
- the processor 130 may read a computer program and perform data processing for machine learning.
- the processor 130 can process computational processes such as processing input data for machine learning, extracting features for machine learning, and calculating errors based on backpropagation.
- the processor 130 for performing such data processing includes a central processing unit (CPU), a general purpose graphics processing unit (GPGPU), a tensor processing unit (TPU), and a custom processing unit. It may include a semiconductor (ASIC: application specific integrated circuit), or a field programmable gate array (FPGA: field programmable gate array). Since the type of processor 130 described above is only an example, the type of processor 130 may be configured in various ways within a range understandable to those skilled in the art based on the contents of the present disclosure.
- FIG. 3 is a flowchart schematically showing a method of controlling the electronic device 100 according to an embodiment of the present disclosure.
- the processor 130 obtains biometric data of the user 1 (S310). Specifically, the processor 130 obtains biometric data of the user 1 from the external electronic device 200 registered in the electronic device 100 through the communication interface 120. At this time, information about the external electronic device 200 may be stored in the memory 110 of the electronic device 100.
- the biometric data of the user 1 may be an electrocardiogram signal.
- the external electronic device 200 may be an electrocardiogram device including a plurality of electrodes (or a plurality of patches including each electrode) attached to different parts of the body of the user 1.
- the external electronic device 200 may generate an ECG signal (or ECG data) corresponding to the user 1 based on the ECG of the user 1 measured from each electrode and transmit it to the electronic device 100.
- the processor 130 may acquire the ECG signal generated by the external electronic device 200 through the communication interface 120.
- biometric data will be described assuming that it is an electrocardiogram signal.
- the processor 130 may generate first information about the disease of the user 1 by inputting the acquired biometric data into a previously learned first neural network model (S320).
- the first information may be information regarding the presence or absence of a disease.
- the first neural network model may be a model trained in advance to generate information on the presence or absence of the user 1's disease. Specifically, when the electrocardiogram signal of the user 1 is input, the first neural network model determines whether a disease exists in the user 1 based on the electrocardiogram signal and generates information about the presence or absence of the disease in the user 1. It may be a model learned on .
- the first neural network model may be trained in advance to determine only the presence or absence of a specific disease.
- a specific disease will be referred to as a first disease.
- the first neural network model determines whether the user 1 has a first disease based on the electrocardiogram signal and generates information about the presence or absence of the first disease as an output value. can do.
- the information regarding the presence or absence of the first disease may include a probability value (or score) that the first disease exists.
- the processor 130 detects the first disease through the Sigmoid Layer or SoftMax Layer disposed at the output stage among the plurality of layers included in the first neural network model. The probability value of this existence can be obtained as an output value.
- the processor 130 may generate second information about the disease of the user 1 by inputting the acquired biometric data into a pre-trained second neural network model (S330).
- the second information may be information about the type of disease.
- the second neural network model may be a model trained in advance to generate information about the type of disease of the user 1. Specifically, when the electrocardiogram signal of the user 1 is input, the second neural network model identifies the type of disease of the user 1 based on the electrocardiogram signal and generates information about the type of disease of the user 1. It may be a pre-trained model.
- the second neural network model may be trained in advance to generate information about the type of a specific disease, where the specific disease is the same disease as the disease for which the first neural network model identifies the presence or absence (i.e., the first disease). It can be.
- the second neural network model may be trained in advance to generate information about a specific type of specific disease. That is, if there are various types related to a specific disease, the second neural network model may be trained in advance to identify whether the specific disease of the user 1 corresponds to a specific type among the various types related to the specific disease.
- the second neural network model can be learned to analyze bio-signals, identify the characteristics of the bio-signals, and then identify a specific type of specific disease corresponding to the identified characteristics.
- a specific type will be referred to as a first type.
- the second neural network model determines whether the first disease of the user 1 corresponds to the first type based on the electrocardiogram signal, and determines whether the first disease of the user 1 corresponds to the first type of the first disease. Related information can be generated as output value.
- the information about the first type of the first disease may include a probability value (or score) that the disease of the user 1 corresponds to the first type of the first disease.
- the processor 130 operates on the user 1 through the Sigmoid Layer or SoftMax Layer disposed at the output stage among the plurality of layers included in the second neural network model.
- the probability that the disease corresponds to the first type of the first disease can be obtained as an output value.
- the first and second neural network models may be implemented as a Convolution Neural Network (CNN) model, a Recurrent Neural Network (RNN) model, a Relu model, etc.
- CNN Convolution Neural Network
- RNN Recurrent Neural Network
- the processor 130 performs a pre-processing process on the acquired biosignal 10 (e.g., electrocardiogram signal) before inputting it into the first and second neural network models. It may be possible.
- the acquired biosignal 10 e.g., electrocardiogram signal
- a plurality of neural network models may be stored in the memory 110.
- multiple neural network models may be matched with other neural network models depending on the type of disease and type of diagnosis result. Therefore, the processor 130 identifies the type of disease to be diagnosed and the type of diagnosis result (or the type of disease and diagnosis result set by the user), and the type of disease and diagnosis result identified among the plurality of neural network models. A combination of multiple neural network models corresponding to the type can be determined. Additionally, the processor 130 may extract a plurality of neural network models according to the determined combination and then generate a diagnosis result of a specific disease for the user 1 using the extracted plurality of neural network models. Meanwhile, combined information of a plurality of neural network models corresponding to the type of disease and the type of diagnosis result may be stored in the memory 110 in the form of a table.
- the processor 130 when it is determined that the processor 130 generates a first type of diagnosis result regarding a first disease, the processor 130 extracts a first neural network model and a second neural network model from the plurality of neural network models stored in the memory 110. You can. Additionally, the processor 130 may generate a first type of diagnosis result regarding the first disease for the user 1 by inputting the acquired biosignals into the extracted first and second neural network models, respectively.
- FIG. 4 is an example diagram showing previously learned first and second neural network models 20-1 and 20-2 stored in the electronic device 100 according to an embodiment of the present disclosure.
- the processor 130 may input the obtained ECG signal into the first and second neural network models 20-1 and 20-2, respectively. At this time, the processor 130 may obtain a plurality of probability values through each neural network model (first and second neural network models 20-1 and 20-2). Specifically, the processor 130 obtains a probability value (hereinafter, first probability value) that the first disease exists in the user 1 through the first neural network model 20-1, and uses the second neural network model 20-1. Through -2), a probability value (hereinafter referred to as a second probability value) that the disease of the user 1 corresponds to the first type of the first disease can be obtained.
- first probability value a probability value that the first disease exists in the user 1 through the first neural network model 20-1
- a probability value hereinafter referred to as a second probability value
- the processor 130 may perform the steps of generating information about the presence or absence of a first disease and the steps of generating information about the first type of the first disease in parallel. That is, the processor 130 inputs the acquired electrocardiogram signal into the first and second neural network models 20-1 and 20-2 arranged in parallel, respectively, to form the first and second neural network models 20-1 and 20-2. By obtaining the output values (i.e., the first and second probability values) of 20-2) respectively, the time required to generate information about the presence or absence of the first disease and information about the first type of the first disease can be shortened. You can.
- Figure 5a is an exemplary diagram showing a method of learning the first and second neural network models (20-1 and 20-2) according to an embodiment of the present disclosure
- Figure 5b is a diagram showing conventional information and specific information on the presence or absence of a specific disease.
- This is an example diagram showing a method of learning multiple neural network models to generate information about a specific type of disease.
- the first and second neural network models 20-1 and 20-2 are to be previously learned based on training data to which a plurality of labels are respectively set to different standards for the same ECG signal. You can.
- the first and second neural network models 20-1 and 20-2 may be previously trained with the same training data.
- the learning data may include a plurality of electrocardiogram signals and may also be referred to as a learning data group.
- the first and second neural network models 20-1 and 20-2 may each be learned based on the same learning data including a plurality of ECG signals.
- different labels may be assigned to a plurality of ECG signals included in the learning data to generate different types of information (information about the presence or absence of the first disease and information about the first type of the first disease). there is.
- the different labels are labels set according to different standards, and may include a first type of label corresponding to the presence or absence of the first disease and a second type of label corresponding to the first type of the first disease. That is, referring to FIG. 5A, the learning data according to an embodiment of the present disclosure includes a first type of label corresponding to the presence or absence of the first disease and a second type of label corresponding to the first type of the first disease. may be granted.
- the first neural network model 20-1 is learned based on the first type of label of the learning data to generate information about the presence or absence of the first disease
- the second neural network model 20-2 is 1
- training data may be learned based on a label of a second type.
- a plurality of labels set with different standards are used.
- the training data can be used repeatedly to train multiple neural network models (i.e., the first and second neural network models (20-1 and 20-2)), thereby allowing the training data to It is possible to reduce the cost and time required for pre-processing and secure sufficient learning data to train each neural network model.
- the first and second neural network models 20-1 and 20-2 may be learned using different learning data.
- the first neural network model 20-1 may be trained to identify the presence or absence of myocardial infarction based on first learning data including electrocardiogram data assigned a first label (a label for distinguishing myocardial infarction)
- the second neural network model 20-2 determines whether the ST segment elevation myocardial infarction is based on the second learning data including ECG data to which a second label (a label that distinguishes ST segment elevation myocardial infarction) is assigned. can be learned to identify.
- the processor 130 provides a diagnosis result of the user 1 regarding the first disease of the first type based on the information regarding the presence or absence of the first disease and the information regarding the first type of the first disease. can be created (S340).
- the processor 130 may generate a diagnosis result regarding the first disease of the user 1 by combining information about the presence or absence of the first disease and information about the first type.
- the processor 130 may generate a diagnosis result regarding the first type of the first disease by combining information regarding the presence or absence of the first disease and information regarding the first type of the first disease. That is, when the processor 130 uses the first neural network model 20-1, it can only determine the presence or absence of the first disease. However, according to an embodiment of the present disclosure, the processor 130 uses the first neural network model 20-1.
- the processor 130 may transmit the generated diagnosis result regarding the first disease to the external electronic device 200 through the communication interface 120 and provide it to the user 1.
- the first disease includes any one of myocardial infarction and ischemic heart disease
- the first type of the first disease may be a high risk group for myocardial infarction. That is, the processor 130 generates information about the presence or absence of myocardial infarction or ischemic heart disease of the user 1 based on the electrocardiogram signal through the first neural network model 20-1, and the second neural network model 20-1 Through -2), information about the high-risk group myocardial infarction of the user (1) can be generated.
- the first disease is myocardial infarction.
- the processor 130 may detect ST segment elevation myocardial infarction (STEMI) or non-ST elevation myocardial infarction (NSTEMI) based on the electrocardiogram signal of the user 1. If any one of these is identified, it may be determined that the user 1's myocardial infarction falls into a high-risk group.
- STMI ST segment elevation myocardial infarction
- NSTEMI non-ST elevation myocardial infarction
- the second neural network model 20-2 can be trained to calculate as an output value whether the user 1 is suffering from ST-segment elevation myocardial infarction or non-ST-segment elevation myocardial infarction based on the input ECG signal.
- the second probability value may be a probability value that the acute myocardial infarction of the user 1 corresponds to an ST segment elevation myocardial infarction or a probability value that corresponds to a non-ST segment elevation myocardial infarction. That is, the processor 130 determines that the disease of the user 1 is either ST segment elevation myocardial infarction or non-ST segment elevation myocardial infarction, based on the second probability value obtained through the second neural network model 20-2.
- information output through the second neural network model 20-2 may also be referred to as information about the first type of the first type of the first disease. That is, the second neural network model 20-2 may be trained in advance to output information about the first type of the first type of the first disease.
- the first type of the first disease will be described assuming ST segment elevation myocardial infarction.
- the first disease of the first type may be acute myocardial infarction. Accordingly, the processor 130 determines that the user 1 has a myocardial infarction based on the first probability value, and ST segment elevation myocardial infarction of the user 1 is identified based on the second probability value. If determined, a diagnosis result regarding acute myocardial infarction of the user 1 can be generated.
- FIG. 6 is a flowchart showing a method of diagnosing a disease of a user 1 using the first and second neural network models 20-1 and 20-2 according to an embodiment of the present disclosure. Steps S610 to S630 shown in FIG. 6 may respectively correspond to steps S310 to S330 shown in FIG. 3, and detailed description thereof will be omitted.
- the processor 130 adjusts the first probability value included in the first information based on the second probability value included in the second information, and adjusts the first probability value based on the first probability value.
- a diagnosis result of the user 1 regarding a first disease of one type may be generated.
- the processor 130 determines the presence or absence of the first disease based on the second probability value.
- the first probability value regarding can be adjusted.
- identifying the first type of a first disease through a second neural network model may mean that there is a high probability that the first type of the first disease exists in the user. Accordingly, the processor 130 may adjust the first probability value regarding the presence or absence of the first disease based on the second probability value regarding the presence or absence of the first type of the first disease.
- the processor 130 may adjust the first probability value by applying a weight corresponding to the second probability value to the first probability value. Additionally, the processor 130 may maintain the first probability value if the second probability value is less than a reference value.
- the processor 130 may determine whether the second probability value is greater than or equal to the reference value.
- the reference value may be set differently for each user (1) depending on the user's (1) age, body type, medical history, etc.
- the processor 130 may determine that the second probability value is greater than or equal to the reference value and that the user 1 has a first disease of a first type. Specifically, the processor 130 may determine that the second probability value is greater than or equal to the reference value and that ST-segment elevation myocardial infarction is identified based on the ECG signal of the user 1. Additionally, if it is determined that ST segment elevation myocardial infarction is identified, the processor 130 may adjust the first probability value by applying a weight corresponding to the second probability value to the first probability value. On the other hand, the processor 130 may determine that the second probability value is less than the reference value and that ST-segment elevation myocardial infarction is not identified based on the ECG signal of the user 1. Additionally, if it is determined that ST segment elevation myocardial infarction is not identified, the processor 130 may maintain the first probability value without applying the weight corresponding to the second probability value to the first probability value.
- the processor 130 may maintain the obtained first probability value 40 as is.
- the processor 130 may adjust the first probability value based on the second probability value in various ways. For example, if the second probability value is greater than or equal to the reference value, the first probability value may be changed to the second probability value.
- the processor 130 may apply different weights depending on the difference between the first probability value and the second probability value. Specifically, if the second probability value is greater than or equal to the reference value and the difference between the first and second probability values is greater than or equal to a preset difference value, the first weight is applied, the second probability value is greater than or equal to the reference value, and the second probability value is greater than or equal to the reference value. If the difference between the first probability value and the second probability value is less than a preset difference value, a second weight smaller than the first weight may be applied.
- the processor 130 may apply different weights depending on the section containing the first probability value among a plurality of preset sections. Specifically, if the second probability value is greater than or equal to the reference value and the first probability value is included in the first interval (0 to less than the first reference value), a third weight is applied, and the second probability value is greater than or equal to the reference value. , If the first probability value is included in the second interval (more than the first reference value and less than the second reference value), a fourth weight smaller than the third weight is applied, and the first probability value is applied to the third interval (the second reference value). or less than the third standard value), a fifth weight smaller than the fourth weight can be applied.
- the processor 130 may generate a diagnosis result based on the first probability value.
- the first probability value may include a first probability value adjusted based on the second probability value or a first probability value output from the first neural network model (i.e., a maintained first probability value).
- the processor 130 generates a first diagnosis result corresponding to a first disease of a first type if the first probability value is greater than or equal to the first value, and if the first probability value is less than the first value, a second diagnosis result is generated. If the value is greater than or equal to the value, generate a second diagnosis result corresponding to the first disease of the first type, and if the first probability value is less than the second value, generate a third diagnosis result corresponding to the first disease of the first type. can do.
- the processor 130 determines that the first probability value (adjusted first probability value or maintained first probability value) is set to the first value (3 ), the user can generate a diagnosis result that is normal (or a diagnosis result that the possibility of acute myocardial infarction is low).
- the processor 130 provides a diagnosis result that the user belongs to a medium risk group for acute myocardial infarction (or, acute myocardial infarction). A diagnostic result indicating that a corresponding possibility exists) can be generated.
- the processor 130 generates a diagnosis result that the user belongs to a high risk group for myocardial infarction (or a diagnosis result that the user is likely to have an acute myocardial infarction). can do.
- the processor 130 compares the first probability value and the second probability value with the first value and the second value, and if the first probability value is identified as being less than the first value, the user 1 does not have the disease of myocardial infarction. If it is determined that the second probability value is greater than or equal to the reference value, it may be determined that the user 1's myocardial infarction falls into the high risk group.
- the processor 130 identifies the ECG signal of the user 1 as abnormal, generates information requesting to measure the ECG signal for the user 1 again, and transmits it to an external electronic device through the communication interface 120. It can be transmitted to (200) or output through an output interface (e.g., speaker, display, etc.).
- the electronic device 100 may be implemented as an electrocardiogram measurement device.
- an electronic device implemented as an electrocardiogram measurement device according to an embodiment of the present disclosure will be described.
- FIG. 7 is a detailed configuration diagram of an electronic device according to an alternative embodiment of the present disclosure.
- the electronic device 100' includes a memory 110', a communication interface 120', a measuring unit 140', a display 150', a user interface 160', and a processor 130'.
- the description of the memory 110, communication interface 120, and processor 130 shown in FIG. 2 is the same as for the configuration memory 110', communication interface 120', and processor 130' shown in FIG. 7. Since it can be applied in many ways, detailed description is omitted.
- the measuring unit 140' includes a plurality of leads attached to different parts of the user's body, and is based on the voltage difference between the user's body parts measured through the plurality of leads.
- the electrocardiogram signal can be generated.
- the measurement unit 140' may include limb leads and chest leads.
- the limb lead may include four electrodes (hereinafter referred to as limb electrodes) attached to the limb.
- the chest lead may include six electrodes (hereinafter referred to as chest electrodes) attached to the chest.
- the limb electrodes may include a right arm electrode (RA), a left arm electrode (LA), a right leg electrode (RL), and a left leg electrode (LL).
- the right leg electrode RL may be a common electrode or a ground electrode.
- the limb electrodes may be attached to positions corresponding to the right arm, left arm, right leg, and left leg, respectively.
- the chest electrode (or pre-chest electrode) includes the first chest electrode (V1), the second chest electrode (V2), the third chest electrode (V3), the fourth chest electrode (V4), and the fifth chest electrode (V5). And it may include a sixth chest electrode (V6).
- the display 150’ can output various image information.
- the processor 130 may output the diagnosis result of the user 1 through the display 150'.
- the processor 130 may output warning information of the first type of the first disease to the user 1 through the display 150'.
- the display 150' may be implemented as a display including a self-luminous element or a display including a non-luminous element and a backlight.
- a display including a self-luminous element or a display including a non-luminous element and a backlight.
- LCD Liquid Crystal Display
- OLED Organic Light Emitting Diodes
- LED Light Emitting Diodes
- micro LED micro LED
- Mini LED Plasma Display Panel
- QD Quantum dot
- QLED Quantum dot light-emitting diodes
- the display 150' may also include a driving circuit and a backlight unit that can be implemented in the form of a-si TFT, LTPS (low temperature poly silicon) TFT, OTFT (organic TFT), etc. Meanwhile, the display 150' includes a touch screen combined with a touch sensor, a flexible display, a rollable display, a 3D display, and a display in which a plurality of display modules are physically connected. It can be implemented.
- the display 150' may form a touch screen together with a touch panel.
- the user interface 160' is a configuration used to perform interaction between the electronic device 100 and the user 1, and the processor 130 can output diagnostic results through the user interface 160'.
- the user interface 160' may include at least one of a touch sensor, a motion sensor, a button, a jog dial, a switch, and a microphone, but is not limited thereto.
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Life Sciences & Earth Sciences (AREA)
- Data Mining & Analysis (AREA)
- Biomedical Technology (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Mathematical Physics (AREA)
- General Physics & Mathematics (AREA)
- Evolutionary Computation (AREA)
- General Engineering & Computer Science (AREA)
- Biophysics (AREA)
- Computing Systems (AREA)
- Molecular Biology (AREA)
- Medical Informatics (AREA)
- Public Health (AREA)
- Computational Linguistics (AREA)
- Cardiology (AREA)
- Pathology (AREA)
- Veterinary Medicine (AREA)
- Animal Behavior & Ethology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Surgery (AREA)
- Heart & Thoracic Surgery (AREA)
- Primary Health Care (AREA)
- Probability & Statistics with Applications (AREA)
- Databases & Information Systems (AREA)
- Epidemiology (AREA)
- Physiology (AREA)
- Fuzzy Systems (AREA)
- Psychiatry (AREA)
- Signal Processing (AREA)
- Algebra (AREA)
- Computational Mathematics (AREA)
- Mathematical Analysis (AREA)
- Mathematical Optimization (AREA)
- Pure & Applied Mathematics (AREA)
Abstract
Description
Claims (11)
- 적어도 하나의 프로세서를 포함하는 전자 장치에 의해 수행되는, 방법으로서,사용자의 생체 데이터를 획득하는 단계;상기 획득된 생체 데이터를 기 학습된 제1 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제1 정보를 생성하는 단계;상기 획득된 생체 데이터를 기 학습된 제2 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제2 정보를 생성하는 단계; 및상기 제1 및 제2 정보에 기초하여, 상기 제1 질병의 제1 유형에 관한 상기 사용자의 진단 결과를 생성하는 단계;를 포함하는, 방법.
- 제1항에 있어서,상기 진단 결과를 생성하는 단계는,상기 제2 정보에 포함된 제2 확률 값에 기초하여, 상기 제1 정보에 포함된 제1 확률 값을 조정하고, 상기 제1 확률 값에 기초하여 상기 제1 유형의 제1 질병에 관한 상기 사용자의 진단 결과를 생성하는 단계;를 포함하는, 방법.
- 제2항에 있어서,상기 진단 결과를 생성하는 단계는,상기 제2 확률 값이 기준 값 이상이면, 상기 제2 확률 값에 대응하는 가중치를 상기 제1 확률 값에 적용하여 상기 제1 확률 값을 조정하는 단계; 및상기 제2 확률 값이 상기 기준 값 미만이면, 상기 제1 확률 값을 유지하는 단계;를 포함하는 방법.
- 제2항에 있어서,상기 진단 결과를 생성하는 단계는,상기 제1 확률 값이 제1 값 이상이면, 상기 제1 유형의 제1 질병에 대응하는 제1 진단 결과를 생성하고, 상기 제1 확률 값이 상기 제1 값 미만이고, 제2 값 이상이면, 상기 제1 유형의 제1 질병에 대응하는 제2 진단 결과를 생성하고, 상기 제1 확률 값이 상기 제2 값 미만이면, 상기 제1 유형의 제1 질병에 대응하는 제3 진단 결과를 생성하는 단계;를 포함하는, 방법.
- 제1항에 있어서,상기 제1 정보는 상기 제1 질병의 유무에 관한 정보이고, 상기 제2 정보는 상기 제1 질병의 제1 종류에 관한 정보인, 방법.
- 제1항에 있어서,복수의 신경망 모델 중, 상기 제1 질병 및 상기 진단 결과 유형에 기초하여, 제1 및 제2 신경망 모델을 추출하는 단계를 포함하는, 방법.
- 제5항에 있어서,상기 생체 데이터는, 심전도 신호를 포함하고,상기 제1 질병은, 심근경색 및 허혈성심장질환 중 어느 하나를 포함하며,상기 제1 종류는, ST분절의 상승 심근경색(STEMI: ST elevation myocardial infarction) 및 ST 분절의 비상승 심근경색(NSTEMI: non-ST elevation myocardial infarction) 중 어느 하나를 포함하는, 방법.
- 제1항에 있어서,상기 제1 및 제2 신경망 모델은,동일한 심전도 신호에 대해 각각 상이한 기준으로 설정된 복수의 라벨이 부여된 학습 데이터에 기초하여 기 학습되고,상기 복수의 라벨은,상기 제1 질병의 유무에 대응하는 제1 유형의 라벨과 상기 제1 질병의 제1 종류에 대응하는 제2 유형의 라벨을 포함하는, 방법.
- 제1항에 있어서,상기 제1 질병에 관한 제1 정보를 생성하는 단계 및 상기 제1 질병에 관한 제2 정보를 생성하는 단계는, 병렬적으로 수행되는, 방법.
- 전자 장치에 있어서,통신 인터페이스;기 학습된 제1 및 제2 신경망 모델이 저장된 메모리;상기 통신 인터페이스를 통해 사용자의 생체 데이터를 획득하고, 상기 획득된 생체 데이터를 기 학습된 제1 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제1 정보를 생성하고, 상기 획득된 생체 데이터를 기 학습된 제2 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제2 정보를 생성하고, 상기 제1 및 제2 정보에 기초하여, 상기 제1 유형의 제1 질병에 관한 상기 사용자의 진단 결과를 생성하는 하나 이상의 프로세서를 포함하는, 전자 장치.
- 적어도 하나의 프로세서를 포함하는 전자 장치에 의해 수행되는, 방법으로서,사용자의 생체 데이터를 획득하는 단계;상기 획득된 생체 데이터를 기 학습된 제1 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제1 정보를 생성하는 단계;상기 획득된 생체 데이터를 기 학습된 제2 신경망 모델에 입력하여 상기 사용자의 제1 질병에 관한 제2 정보를 생성하는 단계; 및상기 제1 및 제2 정보에 기초하여, 상기 제1 유형의 제1 질병에 관한 상기 사용자의 진단 결과를 생성하는 단계;를 포함하는, 방법.
Priority Applications (3)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| JP2025526507A JP2025539032A (ja) | 2022-11-11 | 2023-11-10 | 生体信号に基づいて使用者の疾病を診断する電子装置及びその制御方法 |
| EP23889210.3A EP4599753A4 (en) | 2022-11-11 | 2023-11-10 | ELECTRONIC DEVICE FOR DIAGNOSING A USER'S ILLNESS BASED ON A BIOLOGICAL SIGNAL AND CONTROL METHOD THERE FOR IT |
| CN202380078088.6A CN120201960A (zh) | 2022-11-11 | 2023-11-10 | 以生物信号为基础诊断使用者的疾病的电子装置及其控制方法 |
Applications Claiming Priority (4)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| KR10-2022-0150322 | 2022-11-11 | ||
| KR20220150322 | 2022-11-11 | ||
| KR10-2023-0154684 | 2023-11-09 | ||
| KR1020230154684A KR102853556B1 (ko) | 2022-11-11 | 2023-11-09 | 생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| WO2024101954A1 true WO2024101954A1 (ko) | 2024-05-16 |
Family
ID=91033338
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| PCT/KR2023/018074 Ceased WO2024101954A1 (ko) | 2022-11-11 | 2023-11-10 | 생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 |
Country Status (4)
| Country | Link |
|---|---|
| EP (1) | EP4599753A4 (ko) |
| JP (1) | JP2025539032A (ko) |
| CN (1) | CN120201960A (ko) |
| WO (1) | WO2024101954A1 (ko) |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060117546A (ko) * | 2005-05-11 | 2006-11-17 | 인하대학교 산학협력단 | 신경망을 이용한 심전도 기반의 심장질환 진단장치 및방법 |
| KR20100128083A (ko) * | 2009-05-27 | 2010-12-07 | 충북대학교 산학협력단 | 심장질환 진단장치 및 방법 |
| KR20200071183A (ko) * | 2018-12-10 | 2020-06-19 | 순천향대학교 산학협력단 | 심층 신경망을 이용한 부정맥 분류 시스템 및 방법 |
| US20200205745A1 (en) * | 2018-12-26 | 2020-07-02 | Analytics For Life Inc. | Methods and systems to configure and use neural networks in characterizing physiological systems |
| KR20220008447A (ko) * | 2020-07-14 | 2022-01-21 | 주식회사 바디프랜드 | 딥러닝 기반 심전도를 이용한 심장 질환 진단 장치 및 그 방법 |
Family Cites Families (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US11763943B2 (en) * | 2018-03-02 | 2023-09-19 | Preventice Solutions, Inc. | Automated ventricular ectopic beat classification |
| CN110276411B (zh) * | 2019-06-28 | 2022-11-18 | 腾讯科技(深圳)有限公司 | 图像分类方法、装置、设备、存储介质和医疗电子设备 |
| WO2021163331A1 (en) * | 2020-02-12 | 2021-08-19 | Irhythm Technologies, Inc | Non-invasive cardiac monitor and methods of using recorded cardiac data to infer a physiological characteristic of a patient |
| KR102725596B1 (ko) * | 2021-01-04 | 2024-11-05 | 주식회사 메디컬에이아이 | 시차를 둔 단일유도 심전도를 통한 딥러닝기반의 복수 유도 심전도 생성 시스템 |
-
2023
- 2023-11-10 EP EP23889210.3A patent/EP4599753A4/en active Pending
- 2023-11-10 JP JP2025526507A patent/JP2025539032A/ja active Pending
- 2023-11-10 CN CN202380078088.6A patent/CN120201960A/zh active Pending
- 2023-11-10 WO PCT/KR2023/018074 patent/WO2024101954A1/ko not_active Ceased
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| KR20060117546A (ko) * | 2005-05-11 | 2006-11-17 | 인하대학교 산학협력단 | 신경망을 이용한 심전도 기반의 심장질환 진단장치 및방법 |
| KR20100128083A (ko) * | 2009-05-27 | 2010-12-07 | 충북대학교 산학협력단 | 심장질환 진단장치 및 방법 |
| KR20200071183A (ko) * | 2018-12-10 | 2020-06-19 | 순천향대학교 산학협력단 | 심층 신경망을 이용한 부정맥 분류 시스템 및 방법 |
| US20200205745A1 (en) * | 2018-12-26 | 2020-07-02 | Analytics For Life Inc. | Methods and systems to configure and use neural networks in characterizing physiological systems |
| KR20220008447A (ko) * | 2020-07-14 | 2022-01-21 | 주식회사 바디프랜드 | 딥러닝 기반 심전도를 이용한 심장 질환 진단 장치 및 그 방법 |
Non-Patent Citations (1)
| Title |
|---|
| See also references of EP4599753A4 * |
Also Published As
| Publication number | Publication date |
|---|---|
| EP4599753A1 (en) | 2025-08-13 |
| CN120201960A (zh) | 2025-06-24 |
| JP2025539032A (ja) | 2025-12-03 |
| EP4599753A4 (en) | 2025-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2022019402A1 (ko) | 시계열 생체 신호 기반의 인공신경망 모델 학습 컴퓨터 프로그램 및 방법 | |
| WO2019031794A1 (ko) | 피검체의 치명적 증상의 발생을 조기에 예측하기 위한 예측 결과를 생성하는 방법 및 이를 이용한 장치 | |
| KR102853556B1 (ko) | 생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 | |
| WO2022270755A1 (ko) | 결측 심전도 신호 재생성 장치 및 방법 | |
| WO2021162488A2 (ko) | 질환을 예측하는 방법 및 이를 수행하는 장치 | |
| WO2023048463A1 (ko) | 딥러닝 기반으로 연속적인 신체 상태를 계측하는 방법, 컴퓨터 프로그램 및 장치 | |
| WO2023048486A1 (ko) | 심전도 신호의 오류 복원 방법, 프로그램 및 장치 | |
| WO2024101954A1 (ko) | 생체 신호에 기초하여 사용자의 질병을 진단하는 전자 장치 및 그 제어 방법 | |
| WO2024191190A1 (ko) | 신경망 모델에 기반하여 노이즈를 포함한 심전도 신호를 선별하는 방법, 프로그램 및 장치 | |
| WO2023048437A1 (ko) | 의료 데이터를 기반으로 하는 딥러닝 모델의 학습 및 추론 방법, 프로그램 및 장치 | |
| WO2024096290A1 (ko) | 심전도의 추가 측정 유도 방법 | |
| WO2024025326A1 (ko) | 심전도 판독 서비스를 제공하는 방법, 프로그램 및 장치 | |
| WO2025080026A1 (ko) | 신경망 모델에 기반하여 환자의 질병을 예측하는 방법, 프로그램 및 장치 | |
| WO2024191200A1 (ko) | 인공 지능 모델에 기반하여 잠재적 심방 세동을 식별하는 방법, 프로그램 및 장치 | |
| WO2025211677A1 (ko) | 심전도 데이터에 기초하여 부르가다 증후군을 진단하는 방법, 장치 및 프로그램 | |
| WO2025198378A1 (ko) | 비정상 심전도 신호에 기초하여 사용자의 상태를 파악하는 방법, 장치 및 프로그램 | |
| WO2026054587A1 (ko) | 복수의 학습 데이터 셋에 기초하여 사용자의 상태를 파악하도록 학습된 신경망 모델을 획득하는 장치, 방법 및 프로그램 | |
| WO2025159552A1 (ko) | 딥 러닝 모델 기반의 생체 데이터로부터 환자의 중증도를 판단하는 방법, 장치 및 프로그램 | |
| WO2025053522A1 (ko) | 생체 데이터를 분석하기 위하여 설정된 판단 기준의 성능을 평가하는 장치 및 그 제어 방법 | |
| WO2024225785A1 (ko) | 생체 정보를 식별하는 사용자 맞춤형 신경망 모델을 획득하는 방법, 프로그램 및 장치 | |
| WO2024248501A1 (ko) | 생체 정보를 식별하는 사용자 맞춤형 신경망 모델을 획득하는 방법, 프로그램 및 장치 | |
| WO2026014825A1 (ko) | 심전도 데이터에 기초하여 임상 시험을 수행하는 장치, 방법 및 컴퓨터 프로그램 | |
| WO2025127719A1 (ko) | 심전도 데이터 기반의 질병을 예측하는 신경망 모델을 획득하는 방법, 장치 및 컴퓨터 프로그램 | |
| Saikouk et al. | Towards the implementation of smartphone-based self-testing of COVID-19 using AI | |
| WO2026014648A1 (ko) | 사용자 환경에 맞는 신경망 모델을 구축하는 방법, 장치 및 컴퓨터 프로그램 |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| 121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 23889210 Country of ref document: EP Kind code of ref document: A1 |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2023889210 Country of ref document: EP |
|
| ENP | Entry into the national phase |
Ref document number: 2025526507 Country of ref document: JP Kind code of ref document: A |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 2025526507 Country of ref document: JP |
|
| WWE | Wipo information: entry into national phase |
Ref document number: 202547045044 Country of ref document: IN Ref document number: 202380078088.6 Country of ref document: CN |
|
| ENP | Entry into the national phase |
Ref document number: 2023889210 Country of ref document: EP Effective date: 20250507 |
|
| WWP | Wipo information: published in national office |
Ref document number: 202547045044 Country of ref document: IN |
|
| NENP | Non-entry into the national phase |
Ref country code: DE |
|
| WWP | Wipo information: published in national office |
Ref document number: 202380078088.6 Country of ref document: CN |
|
| WWP | Wipo information: published in national office |
Ref document number: 2023889210 Country of ref document: EP |