Method and system for estimating risk of cardiac arrest and death

Abstract

According to an aspect of the present invention, provided is a method for estimating the risk of cardiac arrest and death, comprising the steps of: obtaining non-real-time information and real-time information associated with a bio-signal of a subject; applying a first classification model and a second classification model to the non-real-time information and the real-time information to generate output information based on the non-real-time information and output information based on the real-time information, respectively; and combining the output information based on the non-real-time information and the output information based on the real-time information to estimate the risk of at least one of cardiac arrest and death of the subject.

Description

Method and system for estimating the risk of cardiac arrest and death

[0001] The present invention relates to a method and system for estimating the risk of cardiac arrest and death.

[0002] As the shortage of medical personnel within healthcare institutions has become increasingly severe, various technologies are being developed to overcome these human resource limitations. One such technology, which estimates (or predicts) a subject's risk of cardiac arrest and death using only discontinuous biosignals, has been introduced. However, this technology suffers from a significant number of false alarms, limiting its clinically meaningful application.

[0003] The purpose of the present invention is to solve all of the problems of the above-mentioned prior art.

[0004] In addition, another purpose of the present invention is to enable the risk of cardiac arrest and death of a subject to be estimated with high accuracy by allowing non-real-time information and real-time information related to the subject's bio-signals to be input as continuous information into a deep learning or artificial intelligence-based model, and further, to enable the estimation results to be clinically meaningfully utilized.

[0005] A representative configuration of the present invention to achieve the above purpose is as follows.

[0006] According to one aspect of the present invention, a method for estimating a risk of cardiac arrest and death is provided, comprising the steps of: obtaining non-real-time information and real-time information associated with a bio-signal of a subject; applying a first classification model and a second classification model to the non-real-time information and the real-time information, respectively, to generate output information based on the non-real-time information and output information based on the real-time information; and combining the output information based on the non-real-time information and the output information based on the real-time information to estimate a risk of at least one of cardiac arrest and death of the subject.

[0007] According to another aspect of the present invention, a system for estimating a risk of cardiac arrest and death is provided, comprising: an acquisition unit for acquiring non-real-time information and real-time information associated with a bio-signal of a subject; a generation unit for applying a first classification model and a second classification model to the non-real-time information and the real-time information, respectively, to generate output information based on the non-real-time information and output information based on the real-time information; and an estimation unit for estimating a risk of at least one of cardiac arrest and death of the subject by combining the output information based on the non-real-time information and the output information based on the real-time information.

[0008] In addition, a non-transitory computer-readable recording medium recording another method for implementing the present invention, another system, and a computer program for executing the method are further provided.

[0009] According to the present invention, by inputting non-real-time information and real-time information related to the subject's bio-signals as continuous information into a deep learning or artificial intelligence-based model, the subject's risk of cardiac arrest and death can be estimated with high accuracy, and further, the estimation results can be clinically meaningfully utilized.

[0010] FIG. 1 is a diagram schematically illustrating the configuration of an entire system for estimating the risk of cardiac arrest and death according to one embodiment of the present invention.

[0011] FIG. 2 is a drawing showing in detail the internal configuration of an estimation system according to one embodiment of the present invention.

[0012] FIG. 3 is a diagram exemplarily showing a process for estimating the risk of cardiac arrest and death according to the present invention.

[0013] <Explanation of symbols>

[0014] 100: Communications network

[0015] 200: Estimation System

[0016] 210: Acquisition Department

[0017] 220: Generation Unit

[0018] 230: Estimation

[0019] 240: Communications Department

[0020] 250: Control Unit

[0021] 300: Server

[0022] 400: Device

[0023] The following detailed description of the present invention refers to the accompanying drawings, which illustrate specific embodiments in which the present invention may be practiced. These embodiments are described in sufficient detail to enable those skilled in the art to practice the present invention. It should be understood that the various embodiments of the present invention, while different from each other, are not necessarily mutually exclusive. For example, specific shapes, structures, and characteristics described herein may be modified and implemented from one embodiment to another without departing from the spirit and scope of the present invention. Furthermore, it should be understood that the positions or arrangements of individual components within each embodiment may also be modified without departing from the spirit and scope of the present invention. Accordingly, the following detailed description is not to be taken in a limiting sense, and the scope of the present invention is to be construed to encompass the scope of the claims and all equivalents thereof. Like reference numerals in the drawings represent the same or similar elements throughout the several aspects.

[0024] Hereinafter, various preferred embodiments of the present invention will be described in detail with reference to the attached drawings so that a person having ordinary skill in the art to which the present invention pertains can easily practice the present invention.

[0025] Composition of the entire system

[0026] FIG. 1 is a diagram schematically illustrating the configuration of an entire system for estimating the risk of cardiac arrest and death according to one embodiment of the present invention.

[0027] As illustrated in FIG. 1, the entire system according to one embodiment of the present invention may include a communication network (100), an estimation system (200), a server (300), and a device (400).

[0028] First, the communication network (100) according to one embodiment of the present invention can be configured regardless of the communication mode such as wired communication or wireless communication, and can be configured with various communication networks such as a local area network (LAN), a metropolitan area network (MAN), and a wide area network (WAN). Preferably, the communication network (100) referred to herein may be the well-known Internet or the 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 or wireless data communication network, a well-known telephone network, or a well-known wired or wireless television communication network.

[0029] For example, the communication network (100) may be a wireless data communication network that implements conventional communication methods 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., at least in part.

[0030] Next, the estimation system (200) according to one embodiment of the present invention can communicate with the server (300) and the device (400) described later through the communication network (100). In addition, the estimation system (200) according to one embodiment of the present invention can obtain non-real-time information and real-time information related to the bio-signals of the subject of measurement, apply a first classification model and a second classification model to the non-real-time information and the real-time information, respectively, to generate output information based on the non-real-time information and output information based on the real-time information, and perform a function of estimating the risk of at least one of cardiac arrest and death of the subject of measurement by combining the output information based on the non-real-time information and the output information based on the real-time information. Meanwhile, the estimation system (200) may be a digital device having a memory means and a microprocessor equipped with a computing capability, and may be, for example, a server system operating on the communication network (100).

[0031] The configuration and function of the estimation system (200) according to one embodiment of the present invention will be described in detail below.

[0032] Next, a server (300) according to one embodiment of the present invention can be installed within a medical institution such as a hospital or public health center, and can perform a function of managing various events occurring within the medical institution.

[0033] In particular, the server (300) according to one embodiment of the present invention can perform a function of storing and managing (e.g., updating) the electronic medical record (EMR) of the subject.

[0034] Here, according to one embodiment of the present invention, the subject's heart rate (HR), respiratory rate (RR), body temperature (BT), blood pressure (BP), and other bio-signals may be recorded in the electronic medical record (EMR). According to one embodiment of the present invention, the bio-signals may be recorded in the electronic medical record (EMR) at predetermined time intervals, but may not be recorded in real time, and thus, the bio-signals recorded in the electronic medical record (EMR) may include measurement gaps.

[0035] Meanwhile, according to one embodiment of the present invention, the server (300) can communicate with the estimation system (200) through the communication network (100), and during this communication process, an electronic medical record (EMR) (or a bio-signal recorded in the electronic medical record (EMR)) can be transmitted from the server (300) to the estimation system (200).

[0036] Next, a device (400) according to one embodiment of the present invention may be configured to include a sensing means (not shown) (e.g., a contact electrode, etc.) attached to the body of a subject to measure a biosignal (e.g., an electrocardiogram (ECG), saturation of percutaneous oxygen (SpO2), etc.) from the subject in real time, or may be linked with the sensing means.

[0037] For example, a device (400) according to one embodiment of the present invention may be configured to include a sensing means for measuring an electrocardiogram (ECG) based on a single or multiple leads by being attached to the body of a subject, or may be linked to the sensing means. According to one embodiment of the present invention, such a device (400) may be configured in the form of a patient monitor (PM), a Holter, a patch, a wristwatch, etc.

[0038] For another example, a device (400) according to one embodiment of the present invention may be configured to include a sensing means for measuring oxygen saturation (SpO2) or the like based on a photoplethysmogram (PPG) by being attached to the body of a subject (specifically, a peripheral body part such as a fingertip) or may be linked to the sensing means. According to one embodiment of the present invention, such a device (400) may be configured in the form of a clip or the like.

[0039] Meanwhile, according to one embodiment of the present invention, the device (400) can communicate with the estimation system (200) through the communication network (100), and during this communication process, a biosignal measured by the sensing means can be transmitted from the device (400) to the estimation system (200).

[0040] Composition of the estimation system

[0041] Below, the internal configuration of the estimation system (200) will be examined with reference to FIG. 2, and the process of estimating the risk of cardiac arrest and death as the function of the estimation system (200) is realized will be examined with reference to FIG. 3.

[0042] FIG. 2 is a diagram illustrating in detail the internal configuration of an estimation system (200) according to one embodiment of the present invention. Additionally, FIG. 3 is a diagram exemplarily illustrating a process for estimating the risk of cardiac arrest and death according to the present invention.

[0043] As illustrated in FIG. 2, an estimation system (200) according to one embodiment of the present invention may include an acquisition unit (210), a generation unit (220), an estimation unit (230), a communication unit (240), and a control unit (250). According to one embodiment of the present invention, at least some of the acquisition unit (210), the generation unit (220), the estimation unit (230), the communication unit (240), and the control unit (250) of the estimation system (200) may be program modules that communicate with an external system (not shown). These program modules may be included in the estimation system (200) in the form of an operating system, an application program module, or other program modules, and may be physically stored in various known memory devices. In addition, these program modules may also be stored in a remote memory device that can communicate with the estimation system (200). Meanwhile, these 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 later in accordance with the present invention.

[0044] Meanwhile, although the estimation system (200) has been described as above, this description is exemplary, and it is obvious to those skilled in the art that at least some of the components or functions of the estimation system (200) may be realized within the server (300) and / or device (400) or included within an external system (not shown) as needed.

[0045] First, the acquisition unit (210) according to one embodiment of the present invention can perform a function of acquiring non-real-time information (I1) and real-time information (I2) associated with the bio-signal of the subject of measurement.

[0046] For example, the acquisition unit (210) according to one embodiment of the present invention can acquire non-real-time information (I1) associated with the bio-signal of the subject from the server (300). Here, according to one embodiment of the present invention, the non-real-time information (I1) may include the bio-signal recorded in the electronic medical record (EMR) of the subject. According to one embodiment of the present invention, the bio-signal recorded in the electronic medical record (EMR) of the subject may include the heart rate (HR), respiratory rate (RR), body temperature (BT), blood pressure (BP), etc. of the subject as described above. According to one embodiment of the present invention, since such bio-signals may include measurement gaps in the recording process as described above, they may be understood to lack real-time characteristics and thus be included in the non-real-time information (I1).

[0047] For another example, the acquisition unit (210) according to one embodiment of the present invention can acquire real-time information (I2) associated with the bio-signal of the subject from the device (400). Here, according to one embodiment of the present invention, the real-time information (I2) may include the bio-signal of the subject measured by the device (400). According to one embodiment of the present invention, the bio-signal of the subject measured by the device (400) may include the electrocardiogram (ECG), oxygen saturation (SpO2), etc. of the subject as described above. According to one embodiment of the present invention, since such bio-signals can be measured in real time by the device (400) as described above, it can be understood that they are included in the real-time information (I2) as real-time property is satisfied.

[0048] Meanwhile, according to one embodiment of the present invention, the acquisition unit (210) can receive non-real-time information (I1) and real-time information (I2) from the server (300) and the device (400), respectively. At this time, it may be understood that the acquisition unit (210) receives the non-real-time information (I1) and real-time information (I2) as a single module, but it may also be understood that the first acquisition unit (not shown) and the second acquisition unit (not shown) included in the acquisition unit (210) independently receive the non-real-time information (I1) and real-time information (I2) as separate modules. For example, it may be understood that the non-real-time information (I1) is received by the first acquisition unit, and the real-time information (I2) is received by the second acquisition unit.

[0049] On the other hand, as described above, non-real-time information (I1) may include measurement gaps, which may be replaced with valid information (or valid values) through a predetermined processing step (which may correspond to a preprocessing step).

[0050] Specifically, the acquisition unit (210) according to one embodiment of the present invention can replace the measurement gap with valid information using at least one statistical method in response to the inclusion of a measurement gap in the non-real-time information (I1).

[0051] For example, the acquisition unit (210) according to one embodiment of the present invention can replace a measurement gap with valid information by using a method such as using a trend of a plurality of bio-signals recorded at a time point before the time point when a measurement gap occurs, a trend of a plurality of bio-signals recorded at a time point after the time point when a measurement gap occurs, an average of a plurality of bio-signals recorded within a predetermined time interval including the time point when a measurement gap occurs, etc.

[0052] However, according to one embodiment of the present invention, the method of replacing a measurement gap with valid information is not necessarily limited to the above-mentioned methods, and it is to be noted that various methods may be used, such as a method of replacing a measurement gap with the same information as a bio-signal recorded at a point immediately before the occurrence of the measurement gap (forward fill), a method of replacing a measurement gap with the same information as a bio-signal recorded at a point immediately after the occurrence of the measurement gap (backward fill), etc.

[0053] Next, the generation unit (220) according to one embodiment of the present invention can apply the first classification model (M1) and the second classification model (M2) to the non-real-time information (I1) and real-time information (I2) obtained as described above, respectively, to generate output information based on the non-real-time information (I1) and output information based on the real-time information (I2).

[0054] Specifically, the generation unit (220) according to one embodiment of the present invention can input non-real-time information (I1) as input data for the first classification model (M1), and can input real-time information (I2) as input data for the second classification model (M2).

[0055] In particular, when real-time information (I2) is input to a second classification model (M2), the generation unit (220) according to one embodiment of the present invention can divide the real-time information (I2) (e.g., divide it into fixed sizes) to form time-series elements (sequences), and input these time-series elements to the second classification model (M2). Here, according to one embodiment of the present invention, the time-series elements can be configured to be lower dimensional than the real-time information (I2).

[0056] According to one embodiment of the present invention, the first classification model (M1) and the second classification model (M2) may correspond to different deep learning (or artificial intelligence)-based classification models. For example, the first classification model (M1) may be a classification model based on at least one of a recurrent neural network (RNN) and a convolutional neural network (CNN), and the second classification model (M2) may be a classification model based on at least one of a convolutional neural network (CNN) and a vision transformer (ViT).

[0057] Meanwhile, according to one embodiment of the present invention, the first classification model (M1) and the second classification model (M2) may correspond to different deep learning-based classification models as described above, but may be models learned to generate (or output) a risk level related to cardiac arrest and death in common.

[0058] Specifically, according to one embodiment of the present invention, as a result of the first classification model (M1) processing the non-real-time information (I1), output information based on the non-real-time information (I1) may be generated, and as a result of the second classification model (M2) processing the real-time information (I2), output information based on the real-time information (I2) may be generated. Each of the output information generated by the first classification model (M1) and the second classification model (M2) may include a risk level regarding cardiac arrest and death.

[0059] That is, according to one embodiment of the present invention, the first classification model (M1) and the second classification model (M2) can receive information having different attributes (i.e., non-real-time or real-time) and generate information having the same attributes.

[0060] Next, the estimation unit (230) according to one embodiment of the present invention can perform a function of finally estimating the risk of at least one of cardiac arrest and death of the subject by combining output information based on non-real-time information (I1) and output information based on real-time information (I2).

[0061] Specifically, the estimation unit (230) according to one embodiment of the present invention can estimate the risk of at least one of cardiac arrest and death of a subject by an ensemble model (M3) that combines the respective output information generated from the first classification model (M1) and the second classification model (M2) (or combines the first classification model (M1) and the second classification model (M2)).

[0062] For example, the estimation unit (230) according to one embodiment of the present invention may estimate the risk of at least one of cardiac arrest and death of the subject by using a voting method or the like for information output from the first classification model (M1) (i.e., output information based on non-real-time information (I1)) and information output from the second classification model (M2) (i.e., output information based on real-time information (I2)).

[0063] According to one embodiment of the present invention, although FIG. 3 illustrates that the ensemble model (M3) is placed after the first classification model (M1) and the second classification model (M2), it may also be understood that the ensemble model (M3) includes the first classification model (M1) and the second classification model (M2).

[0064] Continuing, the estimation unit (230) according to one embodiment of the present invention can convert each risk estimated from the ensemble model (M3) into a score within a predetermined range and output it.

[0065] Specifically, according to one embodiment of the present invention, each risk estimated in the ensemble model (M3) can be expressed in the form of a probability, and the estimation unit (230) can convert each risk expressed in the form of a probability into the form of a score and output it as a score (S1) regarding the risk of cardiac arrest and a score (S2) regarding the risk of death.

[0066] Here, according to one embodiment of the present invention, the score converted by the estimation unit (230) can be converted to a specific value within a range set between 0 and 100, but it is to be noted that the range is not necessarily limited to the above-described range and can be variously changed within a range that can achieve the purpose of the present invention.

[0067] Next, the communication unit (240) according to one embodiment of the present invention can perform a function that enables data transmission and reception from / to the acquisition unit (210), the generation unit (220), and the estimation unit (230).

[0068] Finally, the control unit (250) according to one embodiment of the present invention can perform a function of controlling the flow of data between the acquisition unit (210), the generation unit (220), the estimation unit (230), and the communication unit (240). That is, the control unit (250) according to one embodiment of the present invention can control the flow of data from / to the outside of the estimation system (200) or the flow of data between each component of the estimation system (200), thereby controlling the acquisition unit (210), the generation unit (220), the estimation unit (230), and the communication unit (240) to perform their respective unique functions.

[0069] The embodiments of the present invention described above may be implemented in the form of program commands that can be executed through various computer components and recorded on a computer-readable recording medium. The computer-readable recording medium may include program commands, data files, data structures, etc., either singly or in combination. The program commands recorded on the computer-readable recording medium may be specially designed and configured for the present invention or may be known and 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 recording media such as CD-ROMs and DVDs, magneto-optical media such as floptical disks, and hardware devices specifically configured to store and execute program commands, such as ROMs, RAMs, and flash memories. Examples of program commands include not only machine language codes generated by a compiler, but also high-level language codes that can be executed by a computer using an interpreter, etc. Hardware devices may be changed into one or more software modules to perform processing according to the present invention, and vice versa.

[0070] Although the present invention has been described above with specific details such as specific components and limited examples and drawings, these are provided only to help a more general understanding of the present invention, and the present invention is not limited to the above examples, and those with ordinary knowledge in the technical field to which the present invention pertains can make various modifications and changes based on this description.

[0071] Therefore, the idea of ​​the present invention should not be limited to the embodiments described above, and not only the scope of the patent claims described below but also all scopes equivalent to or equivalently modified from the scope of the patent claims are considered to fall within the scope of the idea of ​​the present invention.

Claims

1. A method for estimating the risk of cardiac arrest and death, A step of acquiring non-real-time information and real-time information related to the subject's bio-signals, A step of applying a first classification model and a second classification model to the non-real-time information and the real-time information, respectively, to generate output information based on the non-real-time information and output information based on the real-time information, and A step of estimating the risk of at least one of cardiac arrest and death of the subject by combining output information based on the non-real-time information and output information based on the real-time information. method.

2. In paragraph 1, The above non-real-time information includes bio-signals recorded in the subject's electronic medical record (EMR). The above real-time information includes biosignals measured by a device attached to the body of the subject. method.

3. In paragraph 1, In the above acquisition step, in response to the non-real-time information including a measurement gap, at least one statistical method is used to replace the measurement gap with valid information. method.

4. In paragraph 1, The above first classification model includes a classification model based on at least one of a recurrent neural network (RNN) and a convolutional neural network (CNN), The second classification model includes a classification model based on at least one of a convolutional neural network (CNN) and a vision transformer (ViT). method.

5. In paragraph 1, The risk of at least one of cardiac arrest and death of the above subject is estimated using an ensemble model. method.

6. In paragraph 1, The above estimated risk is converted into a score within a certain range and output. method.

7. A non-transitory computer-readable recording medium recording a computer program for executing the method according to paragraph 1.

8. A system for estimating the risk of cardiac arrest and death. An acquisition unit that acquires non-real-time information and real-time information related to the subject's bio-signals; A generation unit that applies a first classification model and a second classification model to the non-real-time information and the real-time information, respectively, to generate output information based on the non-real-time information and output information based on the real-time information, and An estimation unit that estimates the risk of at least one of cardiac arrest and death of the subject by combining output information based on the non-real-time information and output information based on the real-time information. System.

9. In paragraph 8, The above non-real-time information includes bio-signals recorded in the subject's electronic medical record (EMR). The above real-time information includes biosignals measured by a device attached to the body of the subject. System.

10. In paragraph 8, The above acquisition unit, in response to the non-real-time information including a measurement gap, replaces the measurement gap with valid information using at least one statistical method. System.

11. In paragraph 8, The above first classification model includes a classification model based on at least one of a recurrent neural network (RNN) and a convolutional neural network (CNN), The second classification model includes a classification model based on at least one of a convolutional neural network (CNN) and a vision transformer (ViT). System.

12. In paragraph 8, The risk of at least one of cardiac arrest and death of the above subject is estimated using an ensemble model. System.

13. In paragraph 8, The above estimated risk is converted into a score within a certain range and output. System.

Images