Electrocardiogram generation system and method based on deep learning algorithm

Abstract

The present invention relates to an electrocardiogram generating system and method based on a deep learning algorithm. According to the present invention, the system includes a data input unit for inputting 12-lead electrocardiograms measured on multiple patients, a data extraction unit for extracting learning data from the input 12-lead electrocardiograms, a learning unit for inputting the extracted learning data into multiple learning models to learn electrocardiogram characteristics, an electrocardiogram generating unit for inputting one or more reference electrocardiograms from the subject and inputting the input reference electrocardiograms into multiple learning models that have completed learning to generate a virtual electrocardiogram, and a control unit for synchronizing the reference electrocardiogram and the generated virtual electrocardiogram with each other and outputting waveforms for the synchronized reference electrocardiogram and virtual electrocardiogram.

Description

Technical Field

[0001] The present invention relates to an electrocardiogram generation system and method based on a deep learning algorithm, and more particularly, to an electrocardiogram generation system and method for generating a plurality of electrocardiograms from one or more lead electrocardiograms using a deep learning algorithm.

Background Art

[0002] The standard 12-lead electrocardiogram used in hospitals attaches 6 electrodes to the entire chest and 3 electrodes (4 electrodes including the ground electrode) to each limb, and then collects all 12-lead information and synthesizes it to diagnose diseases.

[0003] A 12-lead electrocardiogram records the cardiac potential in 12 electrical directions centered on the heart, and by judging the state of the heart in multiple directions, it is possible to accurately interpret heart diseases limited to one part.

[0004] Measuring the cardiac potential in multiple directions is meaningful in that the characteristics of the heart can be grasped in each direction. For this reason, medicine recommends measuring the standard 12-lead electrocardiogram for the diagnosis of heart diseases (such as myocardial infarction).

[0005] However, in order to take a 12-lead electrocardiogram, the chest must be exposed to attach the chest electrodes, and it is difficult for the general public to attach 9 electrodes (3 on the limbs and 6 on the chest) to the correct positions. Also, after attaching 10 electrodes, it is difficult to move, so it is not suitable for real-time monitoring.

[0006] Therefore, recently, devices that can measure a 1-lead electrocardiogram or two or more lead electrocardiograms have been developed so that they can be used in daily life.

[0007] First, a single-lead electrocardiogram device using two electrodes includes a wristwatch-type electrocardiogram mechanism (Apple Watch or Galaxy Watch). The wristwatch-type electrocardiogram mechanism measures the potential difference between the two electrodes by having the back of the watch in contact with the left wrist and the right finger in contact with the crown of the watch, so that the left arm electrode and the right arm electrode are in contact, and utilizing the potential difference between the two electrodes to measure a single-lead electrocardiogram.

[0008] Also, the wristwatch-type electrocardiogram mechanism is worn on the left arm, and a single-lead measurement is taken by contacting the crown with the right hand. A double-lead measurement is taken by contacting the crown with the right hand with the watch placed on the abdomen, and a triple-lead measurement is taken by contacting the crown with the left hand with the watch placed on the abdomen. Subsequently, a V1-6 lead electrocardiogram is measured by contacting the back of the watch at the V1-6 electrode positions with the left hand in contact with the watch crown.

[0009] The above method has a usability issue in that the user must accurately contact the electrocardiogram at the V1-6 positions. In the case of the V1-6 lead electrocardiogram, different from the standard chest lead electrocardiogram where the potential difference between the virtual center point and the V electrode must be illustrated even when contacting this area, there is a problem in that the standard chest lead cannot be implemented in that the potential difference between the right arm electrode and the V electrode is illustrated.

[0010] Also, it is possible to measure two or more lead electrocardiograms by holding the electrodes with both hands and contacting the electrodes on the back of the device with the leg or ankle, and with the electrodes in contact with the left arm, right arm, and left leg respectively. However, the above method cannot be used for long-term monitoring (such as 1 hour, 24 hours, 7 days, etc.), and has the drawback that the device itself must have three electrodes.

[0011] The background technology of the present invention is disclosed in Korean Registered Patent No. 10-2180135 (announced on November 17, 2020).

Summary of the Invention

Problems to be Solved by the Invention

[0012] Thus, the present invention provides an electrocardiogram generation system and method for generating a plurality of electrocardiograms from one or more lead electrocardiograms using a deep learning algorithm. **Means for Solving the Problem**

[0013] According to an embodiment of the present invention for achieving such a technical problem, an electrocardiogram generation system based on a deep learning algorithm includes a data input unit to which 12-lead electrocardiograms measured from a plurality of patients and patient information corresponding to the 12-lead electrocardiograms are input, a data extraction unit that classifies and stores the input 12-lead electrocardiograms according to the patient information and extracts learning data from the stored 12-lead electrocardiograms, a learning unit that inputs the extracted learning data into one or more learning models to learn the characteristics of electrocardiograms, an electrocardiogram generation unit that inputs one or more reference electrocardiograms from the person to be measured and inputs the input reference electrocardiograms into one or more learning models that have completed learning to generate virtual electrocardiograms, and a control unit that synchronizes the reference electrocardiogram and the generated virtual electrocardiogram with each other and outputs waveforms for the synchronized reference electrocardiogram and virtual electrocardiogram.

[0014] The patient information may include at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram.

[0015] The learning unit can input electrocardiograms of n leads, which are part of the 6-lead electrocardiogram of the limbs and the 6-lead electrocardiogram of the chest, into the first learning model, and learn to discriminate the potential vector for the electrocardiogram input to the first learning model and generate 12 - n electrocardiograms.

[0016] The learning unit constructs 12 second learning models for the 6-lead electrocardiogram of the limbs and the 6-lead electrocardiogram of the chest. When the 12 - n generated electrocardiograms output from the constructed first learning model are input into the second learning model, it can be learned to convert the input lead electrocardiogram into a style having the corresponding potential vector and output it as n virtual electrocardiograms.

[0017] The first learning model and the second learning model can be constructed by mixing an adversarial generation network and an autoencoder method.

[0018] The electrocardiogram generation unit inputs the input reference electrocardiogram into the first learning model, extracts a potential vector for the reference electrocardiogram to generate a plurality of remaining virtual electrocardiograms, and inputs the generated virtual electrocardiograms into the second learning model learned by the extracted potential vector, thereby generating a virtual electrocardiogram of the same induction as the reference electrocardiogram again.

[0019] The control unit matches and synchronizes the reference electrocardiogram and the plurality of virtual electrocardiograms, and outputs the synchronized plurality of electrocardiograms via a monitor.

[0020] Also, according to an embodiment of the present invention, an electrocardiogram generation method using an electrocardiogram generation system includes: a step of inputting a 12-lead electrocardiogram measured from a plurality of patients and patient information corresponding to the 12-lead electrocardiogram into an electrocardiogram generation system based on a deep learning algorithm; a step of classifying and storing the input 12-lead electrocardiogram according to the patient information, and extracting learning data from the stored 12-lead electrocardiogram; a step of inputting the extracted learning data into one or more learning models to learn the characteristics of the electrocardiogram; a step of inputting one or more reference electrocardiograms from the measurement subject, and inputting the input reference electrocardiogram into one or more learning models that have completed learning to generate a virtual electrocardiogram; and a step of synchronizing the reference electrocardiogram and the generated virtual electrocardiogram with each other, and outputting waveforms for the synchronized reference electrocardiogram and virtual electrocardiogram.

Advantages of the Invention

[0021] Thus, according to the present invention, a plurality of synchronized electrocardiograms can be generated from two electrocardiograms measured at different locations using a deep learning algorithm, so that heart diseases can be accurately interpreted like a 12-lead electrocardiogram.

[0022] In addition, according to the present invention, since two electrocardiograms are used, it can be measured even at home or in daily life, can be measured even in a moving state, and real-time monitoring is possible. Since synchronized electrocardiograms are generated by utilizing electrocardiogram information measured at different times, medical staff can interpret the synchronized electrocardiogram information according to the corresponding beats, and more accurate diagnosis is possible based on this.

Brief Description of the Drawings

[0023]

Figure 1

Figure 2

Figure 3

Figure 4

Figure 5

Modes for Carrying Out the Invention

[0024] An electrocardiogram generation system 100 according to an embodiment of the present invention includes a data input unit 110, a data extraction unit 120, a learning unit 130, an electrocardiogram generation unit 140, and a control unit 150.

[0025] First, the data input unit 110 is input with 12-lead electrocardiograms measured from a plurality of patients and patient information corresponding to the 12-lead electrocardiograms.

[0026] Here, the 12-lead electrocardiogram includes 6-lead limb electrocardiograms and 6-lead chest electrocardiograms, and the patient information includes at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram.

[0027] The data extraction unit 120 classifies 12-lead electrocardiograms using the input patient information and stores them in a database. Then, the data extraction unit 120 randomly extracts a plurality of 12-lead electrocardiograms stored in the database to generate training data.

[0028] After that, the learning unit 130 trains a learning model constructed using the training data. More specifically, the learning unit 130 extracts potential vector information about the input lead electrocardiogram and constructs a first learning model that generates a 12 - n lead virtual electrocardiogram from the reference electrocardiogram of n leads, and a second learning model that generates a virtual electrocardiogram of n leads from the 12 - n lead virtual electrocardiogram.

[0029] Then, the learning unit 130 inputs the training data generated by the constructed first learning model or second learning model to train the first learning model or the second learning model.

[0030] The electrocardiogram generation unit 140 acquires one or more reference electrocardiograms from the measurement subject. Then, the electrocardiogram generation unit 140 inputs the acquired reference electrocardiogram to the first learning model that has completed learning, and generates a first virtual electrocardiogram based on the potential vector for the reference electrocardiogram. The electrocardiogram generation unit 140 inputs the first virtual electrocardiogram generated by the learned second learning model so as to output a second virtual electrocardiogram for the second learning model.

[0031] Finally, the control unit 150 synchronizes the reference electrocardiogram and the virtual electrocardiogram, and outputs a plurality of synchronized electrocardiograms from the reference electrocardiogram and the virtual electrocardiogram via a monitor device.

[0032] Mode for Carrying Out the Invention

[0033] Hereinafter, preferred embodiments of the present invention will be described in detail based on the accompanying drawings. Here, the thickness of the lines and the size of the components shown in the drawings may be exaggerated for clarity and convenience of explanation.

[0034] In addition, the terms described below are terms defined in consideration of the functions in the present invention, and these can be changed according to the intentions or conventions of users and operators. Therefore, the definitions of such terms should be based on the content throughout this specification.

[0035] Hereinafter, based on FIG. 1, a electrocardiogram generation system based on a deep learning algorithm according to an embodiment of the present invention will be described in more detail.

[0036] FIG. 1 is a configuration diagram for explaining an electrocardiogram generation system according to an embodiment of the present invention.

[0037] As shown in FIG. 1, an electrocardiogram generation system 100 according to an embodiment of the present invention includes a data input unit 110, a data extraction unit 120, a learning unit 130, an electrocardiogram generation unit 140, and a control unit 150.

[0038] First, the data input unit 110 is input with 12-lead electrocardiograms measured from a plurality of patients and patient information corresponding to the 12-lead electrocardiograms.

[0039] Here, the 12-lead electrocardiogram includes 6-lead limb electrocardiograms and 6-lead chest electrocardiograms, and the patient information includes at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram.

[0040] The data extraction unit 120 classifies the input 12-lead electrocardiograms using the input patient information and stores them in a database. Then, the data extraction unit 120 randomly extracts a plurality of 12-lead electrocardiograms stored in the database to generate learning data.

[0041] Thereafter, the learning unit 130 learns a learning model constructed using the learning data. More specifically, the learning unit 130 extracts potential vector information about the input lead electrocardiogram, and constructs a first learning model that generates a 12 - n lead virtual electrocardiogram from the reference electrocardiogram of n leads, and a second learning model that generates a virtual electrocardiogram of n leads from the 12 - n lead virtual electrocardiogram.

[0042] Then, the learning unit 130 inputs the generated learning data into the constructed first learning model or second learning model to train the first learning model or the second learning model.

[0043] The electrocardiogram generation unit 140 acquires one or more reference electrocardiograms from the subject to be measured. Then, the electrocardiogram generation unit 140 inputs the acquired reference electrocardiogram into the first learning model that has completed learning, and generates a first virtual electrocardiogram based on the potential vector for the reference electrocardiogram. The electrocardiogram generation unit 140 inputs the first virtual electrocardiogram generated by the learned second learning model so as to output a second virtual electrocardiogram for the second learning model.

[0044] Finally, the control unit 150 synchronizes the reference electrocardiogram and the virtual electrocardiogram, and outputs a plurality of synchronized electrocardiograms from the reference electrocardiogram and the virtual electrocardiogram via the monitoring device.

[0045] Hereinafter, based on FIGS. 2 to 5, an electrocardiogram generation method using the electrocardiogram generation system 100 according to an embodiment of the present invention will be described in more detail.

[0046] FIG. 2 is a flowchart for explaining an electrocardiogram generation method using the electrocardiogram generation system according to an embodiment of the present invention.

[0047] As shown in FIG. 2, the electrocardiogram generation method using the electrocardiogram generation system according to an embodiment of the present invention is divided into a stage of training a learning model and a stage of generating an electrocardiogram using the trained learning model.

[0048] In the stage of training the learning model, first, the electrocardiogram generation system 100 is input with the induced electrocardiograms measured in a plurality of patients and the information of the patients (S210).

[0049] First, the 12-lead electrocardiogram records the potentials of the heart in 12 electrical directions centered on the heart.

[0050] When examining a general method of taking an electrocardiogram, first, four electrodes (limb leads) are attached to both arms and both legs of a patient. Here, the electrocardiogram measuring device measures the potential with the electrodes on both arms and the left leg electrode, and the right leg electrode serves as a ground electrode.

[0051] Figure 3 is an exemplary diagram illustrating a method of measuring a general 12 - lead electrocardiogram.

[0052] As shown in Figure 3, the electrocardiogram measuring device subtracts the potential of the right arm from the potential of the left arm to generate a lead I electrocardiogram, subtracts the potential of the right arm from the potential of the left leg to generate a lead II electrocardiogram, and subtracts the potential of the left arm from the potential of the left leg to generate a lead III electrocardiogram.

[0053] The electrocardiogram measuring device obtains the average of the potentials of the electrodes on both arms and the left leg to obtain the potential of a virtual center point. Then, the electrocardiogram measuring device subtracts the potential of the virtual center point from the potential of the left arm to generate an electrocardiogram of lead aVL, subtracts the potential of the virtual center point from the potential of the right arm to generate an electrocardiogram of lead aVR. Also, the electrocardiogram measuring device subtracts the potential of the virtual center point from the potential of the left leg electrode to generate an electrocardiogram of lead aVF. As described above, the electrocardiogram measuring device uses three (four including the ground electrode) limb electrodes to generate a total of six - lead (six - column) electrocardiograms.

[0054] After that, the electrocardiogram measuring device uses the difference between the potential of the virtual center point determined by the limb leads and the potentials of the six electrodes attached to the chest to generate a six - lead chest electrocardiogram. That is, six electrodes of V1, V2, V3, V4, V5, and V6 are attached to the chest from the front side at a predetermined position to the left chest.

[0055] Then, the electrocardiogram measuring device subtracts the potential measured by the V1 electrode from the potential of the virtual center point obtained by the average of the limb electrodes to generate a V1 - lead electrocardiogram.

[0056] As described above, the electrocardiogram measuring device generates a 12 - lead electrocardiogram using the electrodes attached to a plurality of patients. Then, the generated 12 - lead electrocardiogram is transmitted to the electrocardiogram generation system 100.

[0057] Here, the electrocardiogram generation system 100 further inputs the generated 12-lead electrocardiogram and the corresponding patient information.

[0058] Here, the patient information includes at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram.

[0059] When the S210 stage is completed, the electrocardiogram generation system 100 extracts learning data using the collected 12-lead electrocardiogram and patient information (S220).

[0060] To explain this further, the data extraction unit 120 classifies the collected 12-lead electrocardiogram according to patient information and stores it in the database. Then, the data extraction unit 120 randomly extracts from the stored multiple 12-lead electrocardiograms to generate learning data.

[0061] After that, the learning unit 130 learns the first learning model and the second learning model respectively using the generated learning data (S230).

[0062] First, the learning unit 130 inputs the learning data composed of 12-lead electrocardiograms into the first learning model and the second learning model to learn the characteristics of the electrocardiogram. To explain this further, the induced electrocardiogram has different directions of electric current flow due to the potential vector, and the electrocardiogram style is affected by the patient's age and gender. That is, as people get older, the heart muscle decreases, so the amplitude of the electrocardiogram tends to decrease. In the case of women, the position of the electrocardiogram electrode is lowered by the breast or the distance between the heart and the electrode becomes larger, resulting in deformation of the electrocardiogram shape.

[0063] Also, in the case of having chronic obstructive pulmonary disease (late-onset obstructive pulmonary disease), the lung volume increases, and the heart between the lungs is erected in the vertical direction, so the electrical flow of the heart changes in the vertical direction in three-dimensional space.

[0064] Therefore, the learning unit 130 inputs the patient information, that is, the age, gender, health level of the patient, and the lead electrocardiogram separately stored according to the potential vector, into the first learning model and the second learning model. Then, the first learning model and the second learning model learn the characteristics of the input lead electrocardiogram and extract the potential vector information of the input lead electrocardiogram. However, it is also possible to receive and learn the electrocardiogram without being classified by patient information, and it is also possible to learn a learning model that is not limited to one characteristic by this method.

[0065] In addition, the learning unit 130 constructs 12 second learning models corresponding to the 12-lead electrocardiogram. Then, the learning unit 130 makes each second learning model learn according to the characteristics of the electrocardiogram.

[0066] To explain this further, the first learning model, based on the deep learning algorithm, learns the correlation between the input electrocardiogram and the 12-lead electrocardiogram, extracts at least one of the characteristics of the input electrocardiogram, that is, age, gender, and potential vector, and thereby generates a virtual electrocardiogram of 12 - n leads from the reference electrocardiogram of n leads. On the other hand, the second learning model generates a virtual electrocardiogram of n leads from the virtual electrocardiogram of 12 - n leads. For example, assuming that a V1-lead electrocardiogram is to be generated, the learning unit 130 learns the style in the first learning model to generate virtual electrocardiograms of leads I, II, III, aVL, aVR, aVF, V2, V3, V4, V5, V6 from the V1-lead electrocardiogram, and the second learning model learns the style to generate a virtual V1-lead electrocardiogram from the virtual electrocardiograms of leads I, II, III, aVL, aVR, aVF, V2, V3, V4, V5, V6. Then, the second learning model converts the input electrocardiogram into the V1-lead style to generate a virtual electrocardiogram.

[0067] For the first learning model, the electrocardiogram of which lead the input electrocardiogram is is input and learned together so as to determine which lead the input electrocardiogram is. After being learned in this way, when it is used, any electrocardiogram with an unknown lead can be input to generate a 12-lead electrocardiogram.

[0068] Here, the first learning model and the second learning model are based on a deep learning algorithm composed of an autoencoder or a generative adversarial network. The deep learning algorithm can be implemented using one selected from the autoencoder or the generative adversarial network, or can be implemented by mixing the autoencoder and the generative adversarial network.

[0069] When the learning of the learning model is completed in the S210 stage and the S230 stage, the electrocardiogram generation system 100 according to the embodiment of the present invention generates an electrocardiogram using the learned learning model.

[0070] First, the electrocardiogram generation system 100 receives a reference electrocardiogram measured via electrodes attached to the body of the measurement subject (S240).

[0071] Here, the reference electrocardiogram does not include information about the potential vector.

[0072] Thereafter, the electrocardiogram generation unit 140 inputs the input reference electrocardiogram into the first learning model and the second learning model to generate a plurality of virtual electrocardiograms (S250).

[0073] First, the electrocardiogram generation unit 140 inputs the reference electrocardiogram into the first learning model to extract the characteristics of the reference electrocardiogram.

[0074] Here, the characteristics include at least one of the age, gender, and potential vector of the measurement subject.

[0075] Thereby, a first virtual electrocardiogram is generated from the reference electrocardiogram.

[0076] Thereafter, the electrocardiogram generation unit 140 inputs the generated first virtual electrocardiogram into the second learning model. Then, the second learning model generates a second virtual electrocardiogram based on the input first virtual electrocardiogram.

[0077] FIG. 4 is an exemplary diagram for explaining the step S240 shown in FIG. 2.

[0078] As shown in FIG. 4, the first electrocardiogram shows a reference electrocardiogram. Here, it is assumed that the characteristics of the reference electrocardiogram are L1 for the first learning model. Then, the electrocardiogram generation unit 140 generates virtual electrocardiograms of the remaining 11 leads based on the L1 electrocardiogram by the first learning model. After inputting the generated virtual electrocardiograms of the 11 leads into the second learning model, a virtual L1 lead electrocardiogram is generated. Each learning model is configured with a generator for generating an electrocardiogram and a discriminator for determining whether the generated electrocardiogram is accurately generated and for improving the accuracy of the generated electrocardiogram by feedback.

[0079] When the step S250 is completed, the control unit 150 outputs the reference electrocardiogram and the 11 virtual electrocardiograms via the monitoring device (S260).

[0080] FIG. 5 is an exemplary diagram for explaining the step S260 shown in FIG. 2.

[0081] The control unit 150 outputs the reference electrocardiogram and the 11 virtual electrocardiograms. Then, as shown in FIG. 5, the control unit 150 synchronizes the output reference electrocardiogram and the 11 virtual electrocardiograms to determine whether there is any health abnormality in the measurement subject. Here, since the virtual electrocardiograms of the leads identical to the reference electrocardiogram are also generated by the second learning model, it is also possible to discriminate by the 12 virtual electrocardiograms.

[0082] As described above, the electrocardiogram is output with different slopes, amplitudes, etc. depending on the age, gender, and presence or absence of health abnormality of the measurement subject. Therefore, the control unit 150 outputs the 12 electrocardiograms synchronized by the reference electrocardiogram and the 11 virtual electrocardiograms via the monitoring device.

[0083] Thus, according to the present invention, since a plurality of synchronized electrocardiograms can be generated from two electrocardiograms measured at different locations using a deep learning algorithm, heart diseases can be accurately interpreted like a 12-lead electrocardiogram.

[0084] In addition, since the electrocardiogram generation system according to the present invention uses two-lead electrocardiograms, it can be measured even at home or in daily life, can be measured even in a moving state, enables real-time monitoring, and utilizes electrocardiogram information measured at different times to generate synchronized electrocardiograms. Therefore, medical staff can interpret the synchronized lead electrocardiogram information according to the corresponding beats, and more accurate diagnosis is possible based on this.

[0085] The present invention has been described based on the embodiments shown in the drawings, but these are merely exemplary, and it will be understandable that those having ordinary knowledge in the field to which the technology belongs will be able to make various modifications and equivalent other embodiments in the future. Therefore, the true technical protection scope of the present invention will have to be determined by the technical idea of the following claims.

Industrial Applicability

[0086] Since the present invention can generate a plurality of synchronized electrocardiograms from two electrocardiograms measured at different locations using a deep learning algorithm, it can accurately interpret heart diseases like a 12-lead electrocardiogram, and is industrially applicable to electrocardiogram generation systems based on various deep learning algorithms.

Explanation of Signs

[0087] 100 Electrocardiogram generation system 110 Data input unit 120 Data extraction unit 130 Learning unit 140 Electrocardiogram generation unit 150 Control unit

Claims

1. An electrocardiogram generation system based on a deep learning algorithm, A data input unit to which 12-lead electrocardiograms measured from a plurality of patients are input, A data extraction unit that extracts learning data from the input 12-lead electrocardiograms, A learning unit that inputs the extracted learning data into a plurality of learning models and learns the characteristics of electrocardiograms based on the deep learning algorithm, The plurality of learning models include a first learning model that generates one or more virtual electrocardiograms of 12 - n leads from an electrocardiogram of n leads, and a second learning model that generates a virtual electrocardiogram of n leads from the virtual electrocardiogram of 12 - n leads generated by the first learning model, The first learning model and the second learning model are models learned based on the deep learning algorithm, When one or more reference electrocardiograms are input from a measured subject, the input reference electrocardiogram is input into the first learning model that has completed learning to generate a virtual electrocardiogram of 12 - n leads, and then, the virtual electrocardiogram of 12 - n leads generated by the first learning model is input into the second learning model that has completed learning to generate a virtual electrocardiogram of n leads, an electrocardiogram generation unit, A control unit that synchronizes the reference electrocardiogram with the virtual electrocardiograms generated by the first learning model and the second learning model and outputs waveforms for the synchronized reference electrocardiogram and virtual electrocardiograms, An electrocardiogram generation system.

2. Patient information corresponding to the 12-lead electrocardiogram is input to the data input unit, The patient information includes at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram, The electrocardiogram generation system according to claim 1.

3. The learning unit inputs electrocardiograms of n leads, which are part of the four-limb six-lead electrocardiogram and the chest six-lead electrocardiogram, into the first learning model, discriminates the potential vector for the input electrocardiogram for the first learning model, and generates a virtual electrocardiogram of 12 - n leads so as to train the first learning model. The electrocardiogram generation system according to claim 2.

4. When the learning unit inputs the electrocardiogram of 12 - n leads generated from the first learning model into the second learning model, it trains the second learning model to convert the input lead electrocardiogram into a style having the potential vector and output it as a virtual electrocardiogram of n leads. The electrocardiogram generation system according to claim 3.

5. The first learning model and the second learning model are constructed by mixing or using respectively the adversarial generation network and the autoencoder method. The electrocardiogram generation system according to claim 1.

6. The electrocardiogram generation unit inputs the input reference electrocardiogram into the first learning model to extract the potential vector for the reference electrocardiogram, generates a first virtual electrocardiogram, and inputs the first virtual electrocardiogram into the second learning model to generate a second virtual electrocardiogram. The electrocardiogram generation system according to claim 5.

7. The control unit matches and synchronizes the reference electrocardiogram with the first virtual electrocardiogram or the second virtual electrocardiogram, and outputs the synchronized plurality of electrocardiograms via a monitor. The electrocardiogram generation system according to claim 6.

8. An electrocardiogram generation method using an electrocardiogram generation system, comprising: inputting a 12-lead electrocardiogram measured from a plurality of patients and patient information corresponding to the 12-lead electrocardiogram; classifying and storing the input 12-lead electrocardiograms according to the patient information, and extracting learning data from the stored 12-lead electrocardiograms; Inputting the extracted training data into a plurality of learning models to learn the characteristics of the electrocardiogram; The plurality of learning models include: a first learning model that generates one or more virtual electrocardiograms of 12 - n leads from an electrocardiogram of n leads; and a second learning model that generates a virtual electrocardiogram of n leads from the virtual electrocardiogram of 12 - n leads generated by the first learning model. When one or more reference electrocardiograms are input from a measured subject, inputting the input reference electrocardiogram into the first learning model that has completed learning to generate a virtual electrocardiogram of 12 - n leads, and subsequently, inputting the virtual electrocardiogram of 12 - n leads generated by the first learning model into the second learning model that has completed learning to generate a virtual electrocardiogram of n leads; Synchronizing the reference electrocardiogram with the virtual electrocardiograms generated by the first learning model and the second learning model, and outputting waveforms for the synchronized reference electrocardiogram and virtual electrocardiograms. An electrocardiogram generation method.

9. The patient information includes at least one of gender, age, presence or absence of heart disease, and the potential vector of the measured electrocardiogram. The electrocardiogram generation method according to claim 8.

10. The step of learning the characteristics of the electrocardiogram is to input an electrocardiogram of n leads, which is a part of the six - limb - lead electrocardiogram and the six - chest - lead electrocardiogram, into the first learning model, and learn the first learning model to discriminate the potential vector for the input electrocardiogram and generate a virtual electrocardiogram of 12 - n leads. The electrocardiogram generation method according to claim 8.

11. The step of learning the characteristics of the electrocardiogram is to input the virtual electrocardiogram of 12 - n leads generated by the first learning model into the second learning model, and learn the second learning model to convert the input electrocardiogram into a style having the potential vector and output it as a virtual electrocardiogram of n leads. The electrocardiogram generation method according to claim 10.

12. The first learning model and the second learning model are constructed by mixing or using respectively an adversarial generation network and an autoencoder method, The electrocardiogram generation method according to Claim 8.

13. The step of generating the virtual electrocardiogram includes inputting the input reference electrocardiogram into the first learning model to extract a potential vector for the reference electrocardiogram, generating a first virtual electrocardiogram, and inputting the first virtual electrocardiogram into the second learning model to generate a second virtual electrocardiogram. The electrocardiogram generation method according to Claim 12.

14. The step of outputting a plurality of synchronized electrocardiograms includes matching and synchronizing the reference electrocardiogram with the virtual electrocardiogram or the second virtual electrocardiogram, and outputting the plurality of synchronized electrocardiograms via a monitor. The electrocardiogram generation method according to Claim 13.

Images