Method, system and non-transient computer-readable medium for estimating arrhythmias using complex artificial neural networks

Abstract

According to one aspect of the present invention, there is provided a method for estimating arrhythmia using a composite artificial neural network, the method comprising the steps of estimating a class corresponding to a bit segment included in a first section of an electrocardiogram signal using a first artificial neural network, estimating a class corresponding to the first section of the electrocardiogram signal using a second artificial neural network, and verifying between the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal.

Description

[Technical Field]

[0001] The present invention relates to a method, system, and non-transitory computer-readable medium for estimating arrhythmias using complex artificial neural networks. [Background technology]

[0002] Recently, the quality of life of all humanity has improved due to the rapid development of science and technology, and many changes have occurred in the medical environment. In particular, wearable monitoring devices that can analyze electrocardiogram signals and predict arrhythmias in daily life without going to the hospital have become widely used.

[0003] Typically, such wearable monitoring devices are equipped with an artificial intelligence model to estimate arrhythmias from electrocardiogram signals, but conventionally, such artificial intelligence models have typically been implemented based on an artificial neural network trained to estimate which type of arrhythmia a given section of an electrocardiogram signal corresponds to.

[0004] However, artificial neural networks trained to estimate which arrhythmia a given section of an electrocardiogram signal corresponds to have a limitation in that they cannot accurately estimate arrhythmias that can be estimated on a bit-segment basis (e.g., atrial premature contraction (APC), ventricular premature contraction (VPC), left bundle branch block (LBBB), right bundle branch block (RBBB), etc.). Therefore, conventional wearable monitoring devices have the problem of being unable to accurately determine the number or severity of arrhythmias that can be estimated on a bit-segment basis within a given electrocardiogram signal section. Summary of the Invention [Problem to be solved by the invention]

[0005] An object of the present invention is to solve all of the problems of the prior art mentioned above.

[0006] Another object of the present invention is to improve the accuracy of arrhythmia estimation by combining an artificial neural network trained to estimate what arrhythmia a bit segment included in a given electrocardiogram signal section corresponds to and an artificial neural network trained to estimate what arrhythmia a given electrocardiogram signal section corresponds to. [Means for solving the problem]

[0007] A typical configuration of the present invention to achieve the above object is as follows.

[0008] According to one aspect of the present invention, there is provided a method for estimating arrhythmia using a composite artificial neural network, the method including the steps of estimating a class corresponding to a bit segment included in a first section of an electrocardiogram signal using a first artificial neural network, estimating a class corresponding to the first section of the electrocardiogram signal using a second artificial neural network, and verifying the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal with the class estimated to correspond to the first section of the electrocardiogram signal.

[0009] According to another aspect of the present invention, there is provided a system for estimating arrhythmia using a composite artificial neural network, the system including: a first estimation unit for estimating a class corresponding to a bit segment included in a first section of an electrocardiogram signal using a first artificial neural network; a second estimation unit for estimating a class corresponding to the first section of the electrocardiogram signal using a second artificial neural network; and a verification unit for verifying the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal.

[0010] In addition, other methods and systems for embodying the present invention, and a non-transitory computer-readable recording medium having a computer program for carrying out the methods are also provided. [Effects of the Invention]

[0011] According to the present invention, the accuracy of arrhythmia estimation can be improved by combining an artificial neural network trained to estimate what arrhythmia a bit segment included in a given electrocardiogram signal section corresponds to and an artificial neural network trained to estimate what arrhythmia a given electrocardiogram signal section corresponds to. [Brief explanation of the drawings]

[0012] [Figure 1] 1 is a diagram showing a schematic configuration of an entire system for predicting arrhythmia using a complex artificial neural network according to an embodiment of the present invention. [Figure 2] 1 is a diagram illustrating in detail the internal configuration of an arrhythmia estimation system according to an embodiment of the present invention. [Figure 3] 1 is a diagram illustrating a cross-validation process according to an embodiment of the present invention; [Explanation of symbols]

[0013] 100:Communication network 200: Arrhythmia Prediction System 210: 1st estimation part 220:Second estimation part 230: Verification Department 240: Communications Department 250: Control unit 300:Device DETAILED DESCRIPTION OF THE INVENTION

[0014] The following detailed description of the present invention refers to the accompanying drawings, which illustrate, by way of example, specific embodiments in which the present invention may be practiced. These embodiments are described in detail to enable those skilled in the art to fully practice the present invention. It should be understood that although the various embodiments of the present invention are different from one another, they are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be embodied in different embodiments without departing from the spirit and scope of the present invention. It should also be understood that the location or arrangement of individual components within each embodiment may be changed without departing from the spirit and scope of the present invention. Therefore, the following detailed description should not be taken in a limiting sense, and the scope of the present invention should be understood to encompass the scope of the appended claims and all equivalents thereof. Like reference numerals in the drawings indicate the same or similar components throughout the various aspects.

[0015] DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings so that those skilled in the art can easily practice the present invention.

[0016] Overall system configuration

[0017] FIG. 1 is a diagram showing a schematic configuration of an entire system for predicting arrhythmia using a complex artificial neural network according to an embodiment of the present invention.

[0018] As shown in FIG. 1, the overall system according to one embodiment of the present invention may include a communication network 100, an arrhythmia estimation system 200, and a device 300.

[0019] First, the communication network 100 according to an embodiment of the present invention may be configured regardless of the communication mode, such as wired communication or wireless communication, and may be configured as various communication networks, such as a local area network (LAN), a metropolitan area network (MAN), or a wide area network (WAN). Preferably, the communication network 100 referred to in this specification may be the well-known Internet or World Wide Web (WWW). However, the communication network 100 is not necessarily limited thereto, and may include at least a portion of a well-known wired / wireless data communication network, a well-known telephone network, or a well-known wired / wireless television communication network.

[0020] For example, communication network 100 may be a wireless data communication network that at least partially implements a conventional communication method such as WiFi communication, WiFi-Direct communication, Long Term Evolution (LTE) communication, 5G communication, Bluetooth communication (including Bluetooth Low Energy (BLE) communication), infrared communication, ultrasonic communication, etc. As another example, communication network 100 may be an optical communication network that at least partially implements a conventional communication method such as LiFi (Light Fidelity), etc.

[0021] Next, the arrhythmia detection system 200 according to an embodiment of the present invention can communicate with a device 300 (described later) through the communication network 100. The arrhythmia detection system 200 according to an embodiment of the present invention can estimate a class corresponding to a bit segment included in a first section of an electrocardiogram signal using a first artificial neural network, estimate a class corresponding to the first section of the electrocardiogram signal using a second artificial neural network, and verify the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal with the class estimated to correspond to the first section of the electrocardiogram signal. Meanwhile, the arrhythmia detection system 200 can be a digital device equipped with a memory means and a microprocessor and having computing capabilities, and can be, for example, a server system operated on the communication network 100.

[0022] The configuration and functions of the arrhythmia estimation system 200 according to an embodiment of the present invention will be explained in detail below.

[0023] Next, the device 300 according to one embodiment of the present invention is a digital device having a function of being able to communicate after being connected to the arrhythmia estimation system 200, such as a smart patch, smart watch, smart band, smart glasses, etc., which is a digital device having a memory means, a microprocessor, and computing capabilities, and may be a wearable monitoring device including a sensing means (e.g., contact electrodes, etc.) for measuring a predetermined biosignal (e.g., an electrocardiogram signal) from the human body and a display means for providing a user with various information regarding the measurement of the biosignal.

[0024] According to an embodiment of the present invention, the device 300 may further include an application program for performing functions according to the present invention. Such an application may exist in the form of a program module within the device 300. The nature of such a program module may generally be similar to the first estimation unit 210, the second estimation unit 220, the verification unit 230, the communication unit 240, and the control unit 250 of the arrhythmia estimation system 200, which will be described later. Here, at least a portion of the application may be replaced, if necessary, with a hardware or firmware device capable of performing substantially the same or equivalent functions.

[0025] Arrhythmia prediction system configuration

[0026] Hereinafter, the internal configuration and functions of each component of the arrhythmia estimation system 200, which performs important functions for implementing the present invention, will be discussed in detail.

[0027] FIG. 2 is a diagram illustrating in detail the internal configuration of an arrhythmia estimation system 200 according to an embodiment of the present invention.

[0028] As shown in FIG. 2 , an arrhythmia estimation system 200 according to an embodiment of the present invention may include a first estimation unit 210, a second estimation unit 220, a verification unit 230, a communication unit 240, and a control unit 250. According to an embodiment of the present invention, the first estimation unit 210, the second estimation unit 220, the verification unit 230, the communication unit 240, and the control unit 250 of the arrhythmia estimation system 200 may be program modules, at least some of which communicate with an external system (not shown). Such program modules may be included in the arrhythmia estimation system 200 in the form of an operating system, application program module, or other program modules, and may be physically stored in various known storage devices. Furthermore, such program modules may be stored in a remote storage device that can communicate with the arrhythmia estimation system 200. Meanwhile, such program modules include, but are not limited to, routines, subroutines, programs, objects, components, data structures, etc. that perform specific tasks or execute specific abstract data types, as described below, according to the present invention.

[0029] Meanwhile, although the arrhythmia estimation system 200 has been described above, this description is merely illustrative, and it will be apparent to those skilled in the art that at least some of the components or functions of the arrhythmia estimation system 200 may be embodied within the device 300 or a server (not shown) or included in an external system (not shown), as necessary.

[0030] First, the first estimation unit 210 according to an embodiment of the present invention may perform a function of estimating a class corresponding to a bit segment included in a first section of an electrocardiogram signal using a first artificial neural network.

[0031] Here, the first artificial neural network according to an embodiment of the present invention may be an artificial neural network trained to estimate to which class of classes indicating a first type of arrhythmia a bit segment (here, a bit segment may refer to a QRS complex shown in an electrocardiogram signal; detecting a bit segment in an electrocardiogram signal may be performed by the first artificial neural network or by other means or methods other than the first artificial neural network) included in a predetermined section of the electrocardiogram signal corresponds. According to an embodiment of the present invention, the first type of arrhythmia may include arrhythmias that can be estimated on a bit segment basis, for example, atrial premature contraction (APC), ventricular premature contraction (VPC), left bundle branch block (LBBB), right bundle branch block (RBBB), etc.

[0032] Specifically, the first estimation unit 210 according to one embodiment of the present invention can use a first artificial neural network to estimate which class among classes indicating a first type of arrhythmia at least one bit segment included in a first section of an electrocardiogram signal corresponds to, and if the bit segment does not correspond to any of the classes indicating a first type of arrhythmia, can estimate that the class corresponding to the bit segment is a class indicating a normal electrocardiogram.

[0033] For example, assuming that an electrocardiogram signal is input to the first artificial neural network, the first estimation unit 210 can use the first artificial neural network to estimate that the fourth bit segment of the five bit segments included in the first section of the electrocardiogram signal corresponds to a class indicating a premature atrial contraction (APC), and that the first, second, third, and fifth bit segments of the five bit segments included in the first section of the electrocardiogram signal correspond to a class indicating a normal electrocardiogram.

[0034] Meanwhile, according to one embodiment of the present invention, the first artificial neural network includes an input layer, a hidden layer, and an output layer, and may be implemented as a convolutional neural network (CNN), a recurrent neural network (RNN), etc., but is not necessarily limited thereto.

[0035] Next, the second estimating unit 220 according to an embodiment of the present invention may perform a function of estimating a class corresponding to a first section of the electrocardiogram signal using a second artificial neural network.

[0036] Here, the second artificial neural network according to an embodiment of the present invention may be an artificial neural network trained to estimate which class of classes indicating a second type of arrhythmia a predetermined section of an electrocardiogram signal corresponds to. According to an embodiment of the present invention, the second type of arrhythmia may include arrhythmias that can be estimated based on rhythm changes between consecutive bit segments, such as atrial fibrillation (AFib), paroxysmal supraventricular tachycardia (SVT), atrioventricular block (AV block), etc.

[0037] Specifically, the second estimation unit 220 according to one embodiment of the present invention can use a second artificial neural network to estimate which class of classes indicating a second type of arrhythmia the first section of the electrocardiogram signal corresponds to, and if the first section does not correspond to any of the classes indicating a second type of arrhythmia, can estimate that the class corresponding to the first section is a class indicating a normal electrocardiogram.

[0038] For example, assuming that an electrocardiogram signal is input to the second artificial neural network, the second estimation unit 220 can use the second artificial neural network to estimate that a first section of the electrocardiogram signal corresponds to a class indicating atrial fibrillation (AFib) or a class indicating a normal electrocardiogram.

[0039] Meanwhile, according to one embodiment of the present invention, a second artificial neural network may be configured in parallel with a first artificial neural network, and the first and second artificial neural networks configured in parallel may receive the same electrocardiogram signal. That is, for the same electrocardiogram signal, the first artificial neural network can estimate which of the classes indicating a first type of arrhythmia a bit segment included in a first section of the corresponding electrocardiogram signal corresponds to, and the second artificial neural network can estimate which of the classes indicating a second type of arrhythmia the corresponding first section of the corresponding electrocardiogram signal corresponds to. According to one embodiment of the present invention, the second artificial neural network, like the first artificial neural network, includes an input layer, a hidden layer, and an output layer, and may be embodied as a convolutional neural network (CNN), a recurrent neural network (RNN), etc., but is not limited thereto.

[0040] Next, the verification unit 230 according to one embodiment of the present invention can perform the function of verifying between the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal.

[0041] According to one embodiment of the present invention, a case may occur in which a class estimated to correspond to a bit segment included in a first section of an electrocardiogram signal and a class estimated to correspond to the first section of the electrocardiogram signal are incompatible. For example, even though premature atrial contractions (APCs) do not appear in a section of the electrocardiogram signal where atrial fibrillation (AFib) occurs, a case may occur in which a class corresponding to the first section of the electrocardiogram signal is estimated to represent atrial fibrillation (AFib), and a class corresponding to a bit segment included in the first section of the electrocardiogram signal is estimated to represent a premature atrial contraction (APC). As in the above case, the class estimation for the first section of the electrocardiogram signal or the bit segment included in the first section of the electrocardiogram signal may be erroneous, but the present invention can resolve such errors through a cross-validation process.

[0042] Specifically, the verification unit 230 according to one embodiment of the present invention can mutually verify the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal, and if the verification result determines that the class estimation for either the bit segment included in the first section of the electrocardiogram signal or the first section of the electrocardiogram signal was made incorrectly (i.e., if the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal are incompatible with each other), the verification unit 230 can correct either the class estimated to correspond to the bit segment included in the first section of the electrocardiogram signal or the class estimated to correspond to the first section of the electrocardiogram signal based on the other.

[0043] For example, as shown in FIG. 3, assuming that the second artificial neural network estimates the class corresponding to the first section of the electrocardiogram signal as a class indicating atrial fibrillation (AFib) (S100), and the first artificial neural network estimates the classes corresponding to 11 of the 19 bit segments included in the first section of the electrocardiogram signal as a class indicating premature atrial contractions (APC) (denoted by "S"), and the classes corresponding to 8 of the bit segments as a class indicating a normal electrocardiogram (denoted by "N") (S200), the verification unit 230 can correct the classes estimated to correspond to the 11 of the 19 bit segments included in the first section of the electrocardiogram signal, i.e., the class indicating premature atrial contractions (APC), to a class indicating a normal electrocardiogram (S→N) based on the class estimated to correspond to the first section of the electrocardiogram signal (i.e., the class indicating atrial fibrillation (AFib)) (S300).

[0044] Next, the communication unit 240 according to an embodiment of the present invention may perform a function that enables data transmission / reception to / from the first estimating unit 210, the second estimating unit 220, and the verifying unit 230.

[0045] Finally, the control unit 250 according to an embodiment of the present invention may perform a function of controlling the flow of data between the first estimating unit 210, the second estimating unit 220, the verifying unit 230, and the communication unit 240. That is, the control unit 250 according to the present invention may control the first estimating unit 210, the second estimating unit 220, the verifying unit 230, and the communication unit 240 to perform their respective unique functions by controlling the flow of data from / to the outside of the arrhythmia estimation system 200 or the flow of data between each component of the arrhythmia estimation system 200.

[0046] The above-described embodiments of the present invention may be embodied in the form of program instructions that can be executed by various computer components and stored on a computer-readable recording medium. The computer-readable recording medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions stored on the computer-readable recording medium may be specially designed and constructed for the present invention, or may be readily available to those skilled in the art of computer software. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tapes, optical media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specially configured to store and execute program instructions, such as ROM, RAM, and flash memory. Examples of program instructions include not only machine language code, such as that produced by a compiler, but also high-level language code that can be executed by a computer using an interpreter, etc. A hardware device may be replaced by one or more software modules to perform the processes of the present invention, and vice versa.

[0047] Although the present invention has been described above using specific details such as specific components and limited examples and drawings, this is merely provided to facilitate a more general understanding of the present invention, and the present invention is not limited to the above examples. Those skilled in the art to which the present invention pertains may attempt various modifications and changes from such descriptions.

[0048] Therefore, the concept of the present invention should not be limited to the above-described embodiments, and all scopes equivalent to or modified equivalently from the scope of the claims, as well as the scope of the claims described below, can be said to fall within the scope of the concept of the present invention.

Claims

1. 1. A method implemented in a system for estimating arrhythmia utilizing a complex artificial neural network, comprising: the system includes a first estimator, a second estimator, and a verifyer; The method comprises: a step of estimating, by the first estimation unit, to which class of classes indicating a first type of arrhythmia each bit segment included in a first section of the electrocardiogram signal corresponds using a first artificial neural network, wherein the first type of arrhythmia includes arrhythmias that can be estimated on a bit segment-by-bit segment basis; a step of estimating, by the second estimation unit, to which class of classes indicating a second type of arrhythmia the first section of the electrocardiogram signal corresponds using a second artificial neural network, the second type of arrhythmia including arrhythmia that can be estimated based on a rhythm change between consecutive bit segments; the verifying unit verifies the class estimated to correspond to each bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal by determining whether the class estimated to correspond to each bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal are mutually exclusive, the first artificial neural network and the second artificial neural network are configured in parallel, and the same electrocardiogram signal is input to the first artificial neural network and the second artificial neural network; a method of correcting a class estimated to correspond to the bit segment, which is determined in the verification step to be incompatible with a class estimated to correspond to the first section of the electrocardiogram signal, based on the class estimated to correspond to the first section of the electrocardiogram signal.

2. A non-transitory computer-readable recording medium having a computer program recorded thereon for executing the method of claim 1.

3. 1. A system for estimating arrhythmia utilizing a complex artificial neural network, comprising: a first estimation unit that estimates to which class each bit segment included in a first section of the electrocardiogram signal corresponds among classes indicating a first type of arrhythmia using a first artificial neural network, the first type of arrhythmia including arrhythmias that can be estimated in units of bit segments; a second estimation unit that estimates to which class the first section of the electrocardiogram signal corresponds among classes indicating a second type of arrhythmia using a second artificial neural network, the second type of arrhythmia including arrhythmia that can be estimated based on a rhythm change between consecutive bit segments; a verification unit that verifies the class estimated to correspond to each bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal by determining whether the class estimated to correspond to each bit segment included in the first section of the electrocardiogram signal and the class estimated to correspond to the first section of the electrocardiogram signal are mutually exclusive, the first artificial neural network and the second artificial neural network are configured in parallel, and the same electrocardiogram signal is input to the first artificial neural network and the second artificial neural network; The verification unit is configured to correct the class estimated to correspond to the bit segment, which is determined to be incompatible with the class estimated to correspond to the first section of the electrocardiogram signal, based on the class estimated to correspond to the first section of the electrocardiogram signal.

Images