KR102818092B1 - Method, program, and apparatus for restoring error of ecg signal - Google Patents

Abstract

According to one embodiment of the present disclosure, a method for error recovery of an electrocardiogram signal performed by a computing device may include a step of obtaining an electrocardiogram signal, a step of estimating an error of the obtained electrocardiogram signal based on a correlation between leads for measuring the obtained electrocardiogram signal, and a step of restoring the electrocardiogram signal whose error is estimated based on restoration information for correcting an error of the electrocardiogram signal.

Description

{METHOD, PROGRAM, AND APPARATUS FOR RESTORING ERROR OF ECG SIGNAL}

The present disclosure relates to a method for restoring errors in electrocardiogram signals, and more particularly, to a method for detecting errors that occur in the process of measuring electrocardiogram signals based on correlations of electrocardiogram signals and restoring electrocardiogram signals.

Clinical examinations and imaging tests are used to examine the presence or absence of heart abnormalities. In particular, for the early diagnosis of heart disease, a method of measuring electrocardiograms, displaying the measured electrocardiogram signals in the form of graphs, and determining the presence or absence of heart abnormalities in the patient based on the graphs is widely used.

Electrocardiography (ECG) records the activity current generated in the myocardium in the form of waves according to the heartbeat. The 12-lead ECG method measures signals by attaching 10 electrodes to the patient's body. At this time, 6 of the 10 electrodes, which are chest lead electrodes (V1 to V6), are attached to specific anatomical locations according to the monopolar thoracic lead method. Then, the remaining 4 electrodes are attached to the patient's limbs, that is, the left hand and foot, and the right hand and foot, respectively. However, the electrodes attached to the left hand and foot and the right hand and foot, respectively, may be attached to the upper left and right and the lower left and right of the chest, depending on the case.

In general, in order to measure an electrocardiogram signal, 10 electrocardiogram electrode attachment locations must be found on the patient's body and attached one by one. However, it is not easy to find the exact locations to attach the 10 electrodes on the patient, and even when used by skilled workers such as medical staff, the electrodes are often attached incorrectly, resulting in signals that cannot be diagnosed or incorrect diagnoses.

Therefore, in order to measure the correct signal by attaching the ECG electrode, a method is needed that can quickly provide a normal ECG signal by determining the reversed position of the ECG electrode and restoring the signal in which lead inversion has occurred.

Korean Patent Publication No. 10-2020-0068161 (Title of the invention: System and method for predicting heart disease using a machine learning model, Publication date: 2020.06.15)

The present disclosure has been made in response to the aforementioned background technology, and aims to provide a method for determining an error, such as a lead reversal case caused by an incorrectly attached electrocardiogram electrode, based on an electrocardiogram signal, and restoring a signal determined to be an error into a correct signal.

The present disclosure aims to provide a method for classifying an electrocardiogram signal in which an error, such as a lead inversion case, has occurred by using a learned deep learning model.

However, the problems to be solved in the present disclosure are not limited to the problems mentioned above, and other problems not mentioned can be clearly understood based on the description below.

According to one embodiment of the present disclosure for realizing the task as described above, a method for error recovery of an electrocardiogram signal performed by a computing device may include a step of obtaining an electrocardiogram signal, a step of estimating an error of the obtained electrocardiogram signal based on a correlation between leads for measuring the obtained electrocardiogram signal, and a step of restoring the electrocardiogram signal whose error is estimated based on restoration information for correcting an error of the electrocardiogram signal.

Alternatively, the step of estimating an error of the acquired electrocardiogram signal based on the correlation between the leads for measuring the acquired electrocardiogram signal may include the step of generating a correlation matrix representing the electrocardiogram signal relationship between the leads based on the acquired electrocardiogram signal, and the step of detecting an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals based on the correlation matrix.

Alternatively, the step of detecting an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals based on the correlation matrix may include a step of classifying an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals by inputting the correlation matrix into a pre-learned first neural network model.

Alternatively, the step of estimating an error of the acquired electrocardiogram signal based on the correlation between the leads for measuring the acquired electrocardiogram signal may include a step of inputting the electrocardiogram signal into a pre-learned self-attention based second neural network model to classify an electrocardiogram signal in which a lead inversion has occurred among the electrocardiogram signals.

Alternatively, the step of estimating an error of the acquired electrocardiogram signal based on a correlation between leads for measuring the acquired electrocardiogram signal may include a step of extracting a target signal for analyzing the correlation from the acquired electrocardiogram signal using a band-pass filter.

Alternatively, the restoration information may include a restoration table indicating a transformation relationship between an electrocardiogram signal of a normal lead and an electrocardiogram signal in which lead inversion has occurred.

Alternatively, the electrodes attached to the body to acquire the electrocardiogram signal may include at least one of a plurality of electrodes attached to an extremity of a limb, or a plurality of electrodes attached to a preset location on the chest.

Alternatively, the acquired electrocardiogram signal may include at least one of a first electrocardiogram signal generated based on a first electrode corresponding to a right arm and a second electrode corresponding to a left arm, a second electrocardiogram signal generated based on a third electrode corresponding to the first electrode and a left leg, or a third electrocardiogram signal generated based on the second electrode and the third electrode, or chest electrocardiogram signals corresponding to each of a plurality of chest electrodes.

According to one embodiment of the present disclosure for realizing the task as described above, a computer program for performing operations for error restoration of an electrocardiogram signal may include an operation of acquiring an electrocardiogram signal, an operation of estimating an error of the acquired electrocardiogram signal based on a correlation between leads for measuring the acquired electrocardiogram signal, and an operation of restoring an electrocardiogram signal whose error is estimated based on restoration information for correcting an error of the electrocardiogram signal.

According to one embodiment of the present disclosure for realizing the task as described above, a computing device for error recovery of an electrocardiogram signal may include a processor including at least one core, a memory including program codes executable by the processor, and a network unit for acquiring an electrocardiogram signal. The processor may acquire an electrocardiogram signal, estimate an error of the acquired electrocardiogram signal based on a correlation between leads for measuring the acquired electrocardiogram signal, and restore an electrocardiogram signal whose error is estimated based on restoration information for correcting the error of the electrocardiogram signal.

The present disclosure can detect a lead inversion case even when an electrode for measuring an electrocardiogram signal is incorrectly attached, restore a correct electrocardiogram signal, and provide the restored signal to the user.

Through this disclosure, it is possible to specifically identify and correct a case of lead inversion when an electrocardiogram signal is measured incorrectly, thereby preventing misdiagnosis.

The present disclosure can restore and provide an electrocardiogram signal to a user without additional measurement when lead inversion occurs due to incorrect attachment of an electrocardiogram electrode.

FIG. 1 is a block diagram of a computing device according to one embodiment of the present disclosure.
FIG. 2 is a block diagram illustrating an error recovery process of an electrocardiogram signal according to one embodiment of the present disclosure.
FIG. 3 is a drawing illustrating an electrocardiogram electrode attachment location and a corresponding lead according to one embodiment of the present disclosure.
FIG. 4 is a table explaining signals generated during lead reversal according to one embodiment of the present disclosure.
FIG. 5 is a diagram illustrating a neural network model according to one embodiment of the present disclosure.
FIG. 6 is a flowchart illustrating a method for restoring an electrocardiogram signal error according to one embodiment of the present disclosure.

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

Throughout the specification of the present disclosure, the same or similar drawing reference numerals refer to the same or similar components. In addition, in order to clearly describe the present disclosure, drawing reference numerals of parts that are not related to the description of the present disclosure may be omitted in the drawings.

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

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

The terms "comprises" and/or "comprising" as used herein should be understood to mean the presence of particular features and/or components. However, it should be understood that the terms "comprises" and/or "comprising" do not exclude the presence or addition of one or more other features, other components, and/or combinations thereof.

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

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

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

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

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

The term "data" used in the present disclosure may include "images", signals, etc. The term "image" used in the present disclosure may refer to multidimensional data composed of discrete image elements. In other words, "image" may be understood as a term referring to a digital representation of an object that can be seen with the human eye. For example, "image" may refer to multidimensional data composed of elements corresponding to pixels in a two-dimensional image. "Image" may refer to multidimensional data composed of elements corresponding to voxels in a three-dimensional image.

As used herein, "lead reversal" may mean lead misplacement, which may occur when electrodes for electrocardiogram measurement are attached to incorrect locations. For example, in the case of lead reversal where an electrode other than the neutral electrode among the limb electrodes is misplaced, a specific lead signal may be changed into a different lead signal, another specific lead signal may be inverted or rotated, and another specific lead signal may not be changed. For example, in the case of a misplacement of the neutral electrode among the limb electrodes, not only the limb leads signal but also the chest leads (precordial leads) signal may be distorted. In this case, the lead signal may appear as another lead signal or the signal size may be reduced to almost 0.

The explanation of the terms set forth above is intended to aid in understanding the present disclosure. Therefore, if the terms set forth above are not explicitly stated as matters limiting the contents of the present disclosure, it should be noted that they are not used to limit the technical ideas of the contents of the present disclosure.

FIG. 1 is a block diagram of a computing device according to one embodiment of the present disclosure.

The computing device (100) according to one embodiment of the present disclosure may be a hardware device or a 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. For example, the computing device (100) may be a server that performs an intensive data processing function and shares resources, or may be a client that shares resources through interaction with a server. In addition, the computing device (100) may be a cloud system in which a plurality of servers and clients interact to comprehensively process data. Since the above description is only one example related to the type of the computing device (100), the type of the computing device (100) may be configured in various ways within a category that can be understood by those skilled in the art based on the contents of the present disclosure.

Referring to FIG. 1, a computing device (100) according to one embodiment of the present disclosure may include a processor (110), a memory (120), and a network unit (130). However, FIG. 1 is only an example, and the computing device (100) may include other configurations for implementing a computing environment. In addition, only some of the configurations disclosed above may be included in the computing device (100).

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

The processor (110) can detect an error in an electrocardiogram signal caused by an electrode being placed in an incorrect position for measuring an electrocardiogram. An electrode being placed in an incorrect position can be understood as an electrode for measuring an electrocardiogram signal of a specific lead being placed out of its normal position. If the electrode is placed in an incorrect position, an electrocardiogram signal that cannot be diagnosed is measured, or the probability of a misdiagnosis occurring due to the measured electrocardiogram signal increases. In order to prevent such a problem, the processor (110) can detect an error in an electrocardiogram signal caused by an electrode being placed in an incorrect position based on the measured electrocardiogram signal.

For example, the processor (110) can analyze an electrocardiogram signal to detect a lead inversion case that occurs due to an electrode being placed in an incorrect position. Specifically, the processor (110) can analyze a correlation between leads for measuring an electrocardiogram signal based on the electrocardiogram signal. At this time, the correlation between leads can be understood as an analysis result indicating what kind of relationship the electrocardiogram signals between leads have based on the measured lead-specific signals. The processor (110) can detect a lead inversion case using the correlation between leads analyzed based on the measured electrocardiogram signal. The processor (110) can detect which lead's signal was measured in an inverted form due to an electrode being placed in an incorrect position based on the correlation between leads. At this time, the computational process for detecting a lead inversion case can be performed according to a rule-based logic defined by a program code, or can be performed based on a deep learning algorithm based on the program code. Through this computational process, the processor (110) can easily detect which inversion case occurred in the measurement signal of which lead for all leads used for measuring the electrocardiogram signal.

The processor (110) can not only detect errors in the electrocardiogram signal, but also restore errors in the detected electrocardiogram signal. The processor (110) can improve errors in the detected electrocardiogram signal based on restoration information determined in advance to correct errors in the electrocardiogram signal. At this time, the restoration information may be predefined information for converting a lead-by-lead signal measured when the electrode is placed in an incorrect position into a lead-by-lead signal measured when the electrode is placed in a normal position. That is, the processor (110) can use the restoration information to convert an electrocardiogram signal in which an error is detected into a signal measured when the electrode is placed in a normal position.

The error detection and restoration process of the electrocardiogram signal performed through the processor (110) can minimize the resources of the hospital environment required to secure the electrocardiogram signal that can be normally used for diagnosing diseases, etc. For example, even if the signal is measured with the electrode settings for measuring the electrocardiogram signal incorrectly, the above-described error detection and restoration process can minimize the hassle of having to reset the electrode settings and re-measure the signal.

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

The memory (120) can manage, by structuring and organizing, data required for the processor (110) to perform operations, combinations of data, and program codes executable by the processor (110). For example, the memory (120) can store electrocardiogram signals or electrocardiogram data received through the network unit (130) described below. The memory (120) can store program codes that operate a neural network model to perform learning by receiving an electrocardiogram signal or various data related to an electrocardiogram signal, program codes that operate a neural network model to perform inference according to the purpose of use of the computing device (100) by receiving an electrocardiogram signal or various data related to an electrocardiogram signal, and processed data generated as the program codes are executed.

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

The network unit (130) can receive data required for the processor (110) to perform calculations through wired or wireless communication with any system or any client, etc. In addition, the network unit (130) can transmit data generated through the calculation of the processor (110) through wired or wireless communication with any system or any client, etc. For example, the network unit (130) can receive electrocardiogram-related data through communication with an electrocardiogram signal measuring device, a database in a hospital environment, a cloud server, or a computing device, etc. The network unit (130) can transmit output data of a neural network model, intermediate data, processed data, etc. derived from the calculation process of the processor (110), etc. through communication with the aforementioned database, server, or computing device, etc.

FIG. 2 is a block diagram illustrating an error recovery process of an electrocardiogram signal according to one embodiment of the present disclosure. FIG. 4 is a table explaining a signal generated during lead inversion according to one embodiment of the present disclosure.

Referring to FIG. 2, a computing device (200) for error recovery of an electrocardiogram signal may include a reverse detection module (220) for detecting an error of input data (210) and a recovery module (230) for correcting an error of the input data (210). In the present disclosure, the input data (210) may include an electrocardiogram signal corresponding to at least one lead. The input data (210) may be generated in a device for measuring an electrocardiogram and transmitted to the computing device (200). When the computing device (200) includes an electrode for measuring an electrocardiogram signal, the input data (210) may be generated directly in the computing device (200).

The reversal detection module (220) may include a filter unit (222), a correlation analysis unit (224), and a case classification unit (226).

The filter unit (222) can filter out noise of the input data (210) to facilitate analysis of the input data (210). For example, the filter unit (222) can include a band pass filter that passes only components of a specific frequency band to be analyzed in the input data (210). The band pass filter can pass only signals of 0.5 Hz to 50 Hz in the input data (210). However, since the value of 0.5 Hz to 50 Hz is only an example, the present disclosure is not limited to such a value. The filter unit (222) can further include a notch filter, etc. to reduce mains hum, which is power noise generated from a power cable, etc.

The correlation analysis unit (224) can analyze correlations for each lead. The correlation analysis of the correlation analysis unit (224) can be performed based on the input data (210) or the output data of the filter unit (222). The correlation analysis unit (224) can analyze what kind of correlations the electrocardiogram signals between the leads have for each lead, thereby generating a correlation matrix of size N X N (N is a natural number) representing the electrocardiogram signal relationships between the leads. For example, when the input data (210) includes 12-lead electrocardiogram signals, the correlation analysis unit (224) can analyze what kind of morphological correlations the 12 leads have, and generate a correlation matrix of size 12X12. The computational process for generating the correlation matrix of size 12X12 can be referred to the following [Mathematical Formula 1].

[Mathematical formula 1]

X = X - E[X] ∈ R ^TC (X ∈ R ^TC ,E[X] ∈ R ^C )

A = X ^T X ∈ R ^CC

In the above mathematical expression 1, X may represent an electrocardiogram signal, C may represent a channel or lead, T may represent time, and A may represent an adjacency matrix. E[X] may be the expected value of the random variable X.

In this way, the correlation analysis unit (224) can organize the correlations among all leads into a single matrix based on the electrocardiogram signal. The correlation analysis unit (224) can express the signal relationships of all leads as a correlation matrix so that the case classification unit (226) can easily interpret the complex signal relationships. That is, the case classification unit (226) can easily perform feature interpretation of the signal for each lead using the matrix generated by the correlation analysis unit (224) and can effectively detect lead inversion.

The case classification unit (226) can determine whether a signal included in the input data (210) is lead inverted based on the correlation matrix or correlation information generated by the correlation analysis unit (224), and classify a lead inversion case. For example, the case classification unit (226) can determine whether a lead inversion exists by comparing the degree of agreement between the information included in the correlation matrix and a predetermined normal case. Then, the case classification unit (226) can classify a lead inversion case in the correlation matrix based on the determination result. The case classification unit (226) can also classify a lead inversion case existing in the input data (210) based on the correlation matrix by using a first neural network model. The first neural network model can include a convolutional neural network, etc. for receiving a visualized matrix and detecting a lead inversion case. The first neural network model can be learned based on supervised learning using a matrix of a normal case in which a lead inversion does not exist as a label. However, neural network models can be trained based on learning methods other than supervised learning, such as unsupervised learning and reinforcement learning.

Meanwhile, according to an alternative embodiment of the present disclosure, the operation of the correlation analysis unit (224) and the operation of the case classification unit (226) may be replaced with a second neural network model. For example, the second neural network model may receive input data (210) or output data of the filter unit (222) and classify a lead inversion case. The second neural network model may be trained to extract features for interpreting the correlation of a lead-by-lead signal based on the input data (210) or output data of the filter unit (22), and to classify a lead inversion case based on the features. At this time, the second neural network model may include a neural network that performs a self-attention-based operation.

The restoration module (230) can restore the electrocardiogram signal whose error is detected by the inversion detection module (220) based on predetermined restoration information. The restoration module (230) can correct the lead inversion case detected by the case classification unit (226) according to the restoration information. The restoration module (230) can use a corresponding signal table (table) (hereinafter, restoration table) for each lead inversion case as restoration information for restoration of the electrocardiogram signal. The restoration table can be information in the form of a table representing the conversion relationship between the electrocardiogram signal of a normal lead and the electrocardiogram signal in which lead inversion has occurred. In this case, normal can be understood as a state in which the arrangement of the electrode is attached to a position determined for normal lead signal measurement. The restoration table can be referred to FIG. 4. For example, according to the restoration table of Fig. 4, in the case of LA/RA leads where the left arm electrode and the right arm electrode are attached in reverse, Lead I is an inverted signal, and Lead II and Lead III can be changed. (Lead I-> -(Lead I), Lead II -> Lead III, Lead III -> Lead II)

In the table of FIG. 4, "-" indicates a signal with reversed polarity, and "≒" may indicate a nearly identical signal. CW may indicate rotation in the direction of RA->LA->LL->RA, and CCW may indicate rotation in the direction of RA->LL->LA->RA. The restoration module (230) may restore an ECG signal with an error by converting an ECG signal with lead inversion into an ECG signal of a normal lead based on a restoration table including the table of FIG. 4. Meanwhile, the ECG signal error restoration method of the present disclosure has been described focusing on the limb leads among the 12 leads, but the application target is not limited thereto, and an ECG measurement method of a 5-lead, 3-lead, or other lead type may also be included.

Referring again to FIG. 2, the output data (240) may be data that restores an electrocardiogram signal in which lead inversion has occurred to an electrocardiogram signal of a normal lead. The output data (240) may be transmitted to various types of clients and displayed through a display device.

The computing device (200) for recovering errors in an electrocardiogram signal may be used in a form combined with or included in a device for measuring an electrocardiogram signal, or may be implemented in a separate device. In addition, the recovery method performed by the computing device (200) may be implemented in a server or computing device physically separated from the electrocardiogram signal measuring device. In addition, the recovery method performed by the computing device (200) may be implemented in a form of receiving a measured signal through a network, analyzing it, and transmitting the analysis result again through the network.

FIG. 3 is a drawing illustrating an electrocardiogram electrode attachment location and a corresponding lead according to one embodiment of the present disclosure.

Referring to FIG. 3, a standard 12-lead (or 12-lead) ECG may include limb leads and chest leads (precordial leads). The limb leads may include four electrodes attached to the limbs (hereinafter, limb electrodes). And, the chest leads may include six electrodes attached to the chest (hereinafter, chest electrodes).

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 locations corresponding to the right arm, left arm, right leg, and left leg, respectively.

The chest electrodes (or prethoracic electrodes) may include a first chest electrode (V1), a second chest electrode (V2), a third chest electrode (V3), a fourth chest electrode (V4), a fifth chest electrode (V5), and a sixth chest electrode (V6).

The limb leads may include standard limb leads such as Lead I, II, and II, and amplified limb leads such as aVR, aVL, and aVF. Among the limb leads, the standard limb lead is a bipolar lead, so that a voltage difference between two electrodes may represent a trace recorded on an electrocardiogram recorder. For example, the first lead (L1) may represent a voltage difference between a right arm electrode (RA) and a left arm electrode (LA), the second lead (L2) may represent a voltage difference between a right arm electrode (RA) and a left leg electrode (LL), and the third lead (L3) may represent a voltage difference between a left arm electrode (LA) and a left leg electrode (LL). The first electrocardiogram signal or the first channel signal may be a signal generated from the first lead (L1), the second electrocardiogram signal or the second channel signal may be a signal generated from the second lead (L2), and the third electrocardiogram signal or the third channel signal may be a signal generated from the third lead (L3). Among the limb leads, the amplified limb lead is a unipolar lead and thus can generate the fourth to sixth electrocardiogram signals corresponding to the right arm electrode (RA), the left arm electrode (LA), and the left leg electrode (LL), respectively.

Since the chest lead is a unipolar lead, it can generate a seventh to twelfth electrocardiogram signal corresponding to each of the chest electrodes.

The above description is only one example for standard 12-lead electrocardiogram measurements, and the present disclosure is not limited to this example.

FIG. 5 is a diagram illustrating a neural network model according to one embodiment of the present disclosure.

Referring to FIG. 5, a neural network model (320) for detecting lead inversion can classify a lead inversion case (330) in an electrocardiogram signal by inputting correlation data (310) representing the relevance of all leads for measuring an electrocardiogram signal. At this time, the correlation data (310) may be a correlation matrix generated in the correlation analysis unit (224) of FIG. 2.

The neural network model (320) can receive a correlation matrix representing the relationship between electrocardiogram signals between leads and interpret the characteristics of the relationship between electrocardiogram signals for each lead. The neural network model (320) can determine whether lead inversion exists among the signals present in the input during the feature interpretation process. In addition, the neural network model (320) can classify lead inversion cases among the input signals based on the judgment result.

For example, the neural network model (320) can classify lead reversal cases for limb leads among 12 leads based on the correlation matrix. An algorithm for classifying lead reversal cases for limb leads among 12 leads can refer to the following [Mathematical Formula 2].

[Mathematical formula 2]

Z = flatten(A[:3,:])∈ R ³⁶

Y = f(Z,θ)∈ R ⁸

Z can be a correlation matrix for three limb leads among the ECG signals for judging lead inversion. θ can be a parameter of a neural network model that classifies lead inversion cases. For example, flatten() can be a function that flattens a multidimensional array space to one dimension.

FIG. 6 is a flowchart illustrating a method for restoring an electrocardiogram signal error according to one embodiment of the present disclosure.

Referring to FIGS. 1 and 6, a computing device (100) according to one embodiment of the present disclosure can obtain an electrocardiogram signal (S110). The computing device (100) can utilize an electrocardiogram signal measured by being connected to an electrocardiogram measuring device or an electrocardiogram signal transmitted through a network. At this time, the electrocardiogram signal obtained by the computing device (100) can include an electrocardiogram signal measured in a standard 12-lead manner, an electrocardiogram signal measured in a 5-lead manner, a 3-lead manner, or a variety of lead manners.

For example, the electrodes attached to the body to acquire an electrocardiogram signal may include at least one of a plurality of limb electrodes attached to the distal portion of a limb according to a standard 12-lead method and a plurality of chest electrodes attached to preset locations on the chest.

For example, the electrocardiogram signal may include at least one of a first electrocardiogram signal generated based on a first electrode corresponding to a right arm and a second electrode corresponding to a left arm according to a standard 12-lead method, a second electrocardiogram signal generated based on the first electrode and a third electrode corresponding to a left leg, or a third electrocardiogram signal generated based on the second electrode and the third electrode.

The computing device (100) can estimate an error of an electrocardiogram signal based on the correlation between leads (S120). The computing device (100) can generate a correlation matrix representing the electrocardiogram signal relationship between leads based on the acquired electrocardiogram signal. In addition, the computing device (100) can detect an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals based on the correlation matrix. The computing device (100) can preprocess an input signal in the process of calculating the correlation between the electrocardiogram signals for each lead, or generate a correlation matrix by considering weights, etc.

The computing device (100) can classify the ECG signal in which the lead inversion has occurred among the acquired ECG signals by inputting the correlation matrix into a pre-learned first neural network model in the process of detecting the ECG signal in which the lead inversion has occurred among the acquired ECG signals based on the correlation matrix. The computing device (100) can classify the ECG signal in which the lead inversion has occurred among the ECG signals by inputting the ECG signal into a pre-learned self-attention based second neural network model.

The computing device (100) can extract a target signal for analyzing the correlation from the acquired electrocardiogram signal by using a bandpass filter to remove noise from the electrocardiogram signal.

The computing device (100) can restore the electrocardiogram signal in which an error is estimated and provide it to the user (S130). The computing device (100) can restore the electrocardiogram signal in which lead inversion is estimated by utilizing a restoration table representing the conversion relationship between the electrocardiogram signal of a normal lead and the electrocardiogram signal in which lead inversion has occurred. For example, the restoration table can include the table described in FIG. 4.

Through the ECG signal error recovery method according to one embodiment of the present disclosure, even when lead inversion occurs due to incorrect attachment of the ECG electrode, information about the incorrectly attached electrode and the lead inversion case can be accurately determined. In addition, through the ECG signal error recovery method according to one embodiment of the present disclosure, the ECG signal can be quickly provided without reattaching the electrode or remeasuring the signal.

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

Claims (10)

A method for error recovery of an electrocardiogram signal, performed by a computing device including at least one processor,
A step of acquiring an electrocardiogram signal including a plurality of leads from an electrocardiogram electrode;
A step of estimating an error of the acquired electrocardiogram signal based on the correlation between the leads; and
A step of restoring an electrocardiogram signal whose error is estimated based on restoration information for correcting an error of the electrocardiogram signal;
Including,
method.
In paragraph 1,
Based on the correlation between the above leads, the step of estimating the error of the acquired electrocardiogram signal is:
A step of generating a correlation matrix representing the electrocardiogram signal relationship between the leads based on the acquired electrocardiogram signal; and
A step of detecting an electrocardiogram signal in which a lead reversal has occurred among the acquired electrocardiogram signals based on the above correlation matrix;
Including,
method.
In the second paragraph,
Based on the above correlation matrix, the step of detecting an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals is as follows.
A step of classifying an electrocardiogram signal in which a lead inversion has occurred among the acquired electrocardiogram signals by inputting the correlation matrix into a pre-learned first neural network model;
Including,
method.
In paragraph 1,
Based on the correlation between the above leads, the step of estimating the error of the acquired electrocardiogram signal is:
A step of inputting the electrocardiogram signal into a pre-learned second neural network model and classifying the electrocardiogram signal in which lead inversion has occurred among the electrocardiogram signals;
Including,
method.
In paragraph 1,
Based on the correlation between the above leads, the step of estimating the error of the acquired electrocardiogram signal is:
A step of extracting a target signal for analyzing the correlation from the acquired electrocardiogram signal using a band pass filter;
Including,
method.
In paragraph 1,
The above restoration information is,
A restoration table (table) including a transformation relationship between an electrocardiogram signal of a normal lead and an electrocardiogram signal in which lead inversion has occurred.
method.
In paragraph 1,
The electrodes attached to the body to obtain the above electrocardiogram signals are:
Comprising at least one of a plurality of electrodes attached to the distal end of a limb, or a plurality of electrodes attached to a predetermined location on the chest;
method.
In paragraph 1,
The above acquired electrocardiogram signal is,
A first electrocardiogram signal generated based on a first electrode corresponding to a right arm and a second electrode corresponding to a left arm, a second electrocardiogram signal generated based on a third electrode corresponding to the first electrode and a left leg, a third electrocardiogram signal generated based on the second electrode and the third electrode, or at least one of chest electrocardiogram signals corresponding to each of a plurality of chest electrodes.
method.
A computer program stored in a computer-readable storage medium, wherein the computer program, when executed on one or more processors, performs operations for error recovery of an electrocardiogram signal.
The above actions are,
An act of acquiring an electrocardiogram signal including multiple leads from an electrocardiogram electrode;
An operation of estimating an error of the acquired electrocardiogram signal based on the correlation between the leads; and
An operation of restoring an electrocardiogram signal whose error is estimated based on restoration information for correcting an error of the electrocardiogram signal;
Including,
Computer program.
As a computing device for error recovery of electrocardiogram signals,
A processor comprising at least one core;
a memory containing program codes executable by the processor; and
A network unit for acquiring an electrocardiogram signal;
Including,
The above processor,
A method for obtaining an electrocardiogram signal including a plurality of leads from an electrocardiogram electrode, estimating an error of the obtained electrocardiogram signal based on a correlation between the leads, and restoring the electrocardiogram signal whose error is estimated based on restoration information for correcting the error of the electrocardiogram signal, characterized in that:
device.

Images