KR102471086B1 - AI-based ECG reading system - Google Patents

Abstract

The embodiment includes an electrocardiogram measuring device for obtaining electrocardiogram data; and a server that reads the specific diseases using the electrocardiogram data, wherein the server includes a time-series data analysis unit that primarily reads the specific diseases through P, Q, R, S, and T peak detection of the electrocardiogram data; An AI-based electrocardiogram reading system including an image data analysis unit for secondary reading of detailed diseases through an AI-based image data analysis algorithm, and a data analysis result integration unit for generating a reading result by integrating the primary reading and secondary reading Initiate.

Description

AI-based ECG reading system}

The present invention relates to an electrocardiogram reading system based on artificial intelligence. Specifically, the present invention relates to a reading system capable of providing highly accurate electrocardiogram reading results using only single-channel electrocardiogram data.

The recent increase in sudden deaths is accelerating. According to statistics from the Korea Centers for Disease Control and Prevention, the number of people who died suddenly from heart problems increased by more than 6,000 from 12,271 in 2008 to 18,658 in 2015, and 18,261 people died in 2017.

According to the World Health Organization, cardiovascular disease is the number one cause of death worldwide. It is estimated that by 2030 the number of deaths will reach 23 million.

Along with population aging, the number of heart disease patients and medical expenses in Korea are gradually increasing, requiring attention to prevent and manage cardiovascular disease. do.

The heartbeat is one of the vital signals that the body transmits. When uncomfortable symptoms such as palpitations or shortness of breath appear, an accurate diagnosis can be made only by taking an electrocardiogram at the time of the symptoms, keeping in mind the possibility that a heart abnormality has occurred.

As such, there is a need for home wellness health management medical devices that can easily measure ECG anytime, anywhere to measure and diagnose an electrocardiogram immediately when a heart abnormality is felt. Major products currently on the market include a service (AliveCor, etc.) that transmits single-channel electrocardiogram measurement data to hospitals so that doctors remotely diagnose and determine the need for a patient visit (such as AliveCor), and Apple, which provides ECG graphs and atrial fibrillation notification functions. Position 4 is representative. These electrocardiogram devices can diagnose arrhythmia diseases that are relatively simple to diagnose, and are being serviced overseas as a form of telemedicine through electrocardiogram data monitoring.

The present invention obtains an ECG signal of similar quality to an ECG reading device used in a hospital, even though only a single channel signal is obtained using two relatively light electrodes, rather than an electrocardiogram reading device generally used in hospitals. It aims to provide an electrocardiogram reading system with similar reading accuracy.

In particular, the present invention solves the problem of accuracy degradation that may occur by using only two electrodes by using and integrating two methods of time-series analysis and image analysis together, and provides a user-portable electrocardiogram reading system aims to

A relatively lightweight AI-based electrocardiogram reading system capable of reading single-channel electrocardiogram data through AI is disclosed. An electrocardiogram reading system according to an embodiment of the present invention includes a single-channel electrocardiogram measuring device that obtains single-channel electrocardiogram data using two electrodes, outputs data obtained from the single-channel electrocardiogram measuring device in real time, and sends the data to a server. A time-series data analysis unit that acquires single-channel electrocardiogram measurement data from the terminal and performs primary reading through an AI-based time-series data analysis algorithm; and an AI independent of the primary reading. and a server including an image data analysis unit that performs secondary reading through a base image data analysis algorithm, and a data analysis result integration unit that generates an integrated reading result by integrating the primary reading and secondary reading.

An electrocardiogram reading system according to an embodiment of the present invention can easily read an electrocardiogram even for ordinary people using only two electrodes like a general electrocardiogram reading system.

In addition, the electrocardiogram reading system according to an embodiment of the present invention can increase reading accuracy by using an AI-based reading algorithm.

In addition, the electrocardiogram reading system according to an embodiment of the present invention uses a reading algorithm in which big data on the specialist's reading result is reflected in the reading procedure, and can increase reading accuracy compared to conventional reading systems that judge only with peaks. .

1 is a conceptual diagram illustrating an electrocardiogram reading system according to an embodiment of the present invention.
2 is a block diagram for explaining a terminal according to an embodiment of the present invention.
3 is a block diagram showing the configuration of a server.
4 is a flowchart illustrating an electrocardiogram reading procedure according to an embodiment of the present invention.
5 shows an example of a user interface including a final read result displayed on a terminal.

Embodiments according to the present invention can be applied with various changes and can have various forms, so the embodiments are illustrated in the drawings and described in detail in this specification. However, this is not intended to limit the embodiments according to the present invention to specific disclosed forms, and includes all changes, equivalents, or substitutes included in the spirit and scope of the present invention.

It is understood that when an element is referred to as being "connected" or "connected" to another element, it may be directly connected or connected to the other element, but other elements may exist in the middle. It should be. On the other hand, when a component is referred to as “directly connected” or “directly connected” to another component, it should be understood that no other component exists in the middle. Other expressions describing the relationship between components, such as "between" and "directly between" or "adjacent to" and "directly adjacent to", etc., should be interpreted similarly.

Terms used in this specification are used only to describe specific embodiments, and are not intended to limit the present invention. Singular expressions include plural expressions unless the context clearly dictates otherwise. In this specification, terms such as "comprise" or "having" are intended to indicate that there is a feature, number, step, operation, component, part, or combination thereof described in this specification, but one or more other features It should be understood that it does not preclude the possibility of the presence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Unless defined otherwise, all terms used herein, including technical and scientific terms, have the same meaning as commonly understood by one of ordinary skill in the art to which the present invention belongs. Terms such as those defined in commonly used dictionaries should be interpreted as having a meaning consistent with the meaning in the context of the related art, and unless explicitly defined in this specification, it should not be interpreted in an ideal or excessively formal meaning. don't “And/or” includes each and every combination of one or more of the recited items.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

1 is a conceptual diagram illustrating an electrocardiogram reading system according to an embodiment of the present invention.

As shown in FIG. 1 , the electrocardiogram reading system according to an embodiment of the present invention may include a terminal 100 , a single-channel electrocardiogram measurement device 200 and a server 300 .

In general, electrocardiogram measurement devices used in large hospitals measure electrocardiogram data by attaching a plurality of electrodes to a patient's body. In the case of a standard 12-lead ECG, a total of 10 electrodes, 6 on the front of the chest and 4 on the extremities, are attached to the patient. Not all of the 10 electrodes directly indicate the ECG, and some electrodes are used by doctors to improve the reading accuracy of the ECG. to provide additional ECG data.

However, in the case of an electrocardiogram measuring device used in a large hospital, since it uses a plurality of electrodes, it is inevitably large, and of course cannot be carried, and the electrocardiogram can be measured only when a patient arrives at the hospital.

Accordingly, the electrocardiogram reading system according to an embodiment of the present invention includes the single-channel electrocardiogram measuring device 200. The single-channel electrocardiogram measuring device 200 of the present invention includes only two electrodes, and thus may be portable. The single-channel electrocardiogram measuring device 200 acquires single-channel electrocardiogram data through contact with a patient's body.

The terminal 100 obtains ECG data from the single-channel ECG measuring device 200, outputs the obtained ECG data, and transmits the ECG data to the server 300. The terminal 100 provides an ECG reading service to the patient based on the ECG reading result obtained from the server 300 . An application that performs a series of procedures for providing electrocardiogram reading data to a patient may be installed and executed in the terminal 100 .

In one embodiment, the terminal 100 may be connected to the single-channel electrocardiogram measurement device 200 through short-range wireless communication. In one embodiment, the terminal 100 may be connected to the server through wired or wireless communication.

The server 300 receives electrocardiogram data from the terminal 100 and generates electrocardiogram reading results according to a series of reading algorithms. Specifically, the server 300 may generate an electrocardiogram reading result through an artificial intelligence-based reading algorithm. The server 300 may build a large amount of electrocardiogram reading database for artificial intelligence electrocardiogram reading. The server 300 may update the ECG reading database through learning whenever the ECG is read.

2 is a block diagram for explaining a terminal according to an embodiment of the present invention.

As shown in FIG. 2 , the terminal 100 according to an embodiment of the present invention includes a communication unit 110, an A/V input unit 120, a user input unit 130, a memory 140, and an output unit 150. , a power supply unit 160 and a control unit 170 may be included. Meanwhile, since the components shown in FIG. 2 are not essential, a terminal having more or fewer components may be implemented.

As described above, the terminal 100 according to an embodiment of the present invention may be a wearable terminal.

Hereinafter, the above components are examined in turn.

The communication unit 110 may include at least one module enabling communication with the server 200 described in FIG. 1 or another terminal including a joystick. For example, the communication unit 110 may include at least one of an internet module 111, a short distance communication module 112, and a location information module 113.

The Internet module 111 refers to a module for wired or wireless Internet access, and may be built into or external to the terminal 100 . Wireless Internet technologies include wireless LAN (WLAN) (Wi-Fi), wireless broadband (Wibro), world interoperability for microwave access (Wimax), high speed downlink packet access (HSDPA), and the like.

The short-distance communication module 112 refers to a module for short-distance communication. As a short range communication technology, Bluetooth, Radio Frequency Identification (RFID), Infrared Data Association (IrDA), Ultra Wideband (UWB), ZigBee, or the like may be used.

The location information module 113 is a module for acquiring the location of the terminal 100, and a representative example thereof is a Global Positioning System (GPS) module.

Referring to FIG. 2 , an audio/video (A/V) input unit 120 is for inputting an audio signal or a video signal, and may include a camera 121 and a microphone 122. The camera 121 processes an image frame such as a still image or a moving image obtained by an image sensor in a video call mode or a photographing mode. The processed image frame may be displayed on the display unit 151 .

An image frame processed by the camera 121 may be stored in the memory 140 or transmitted to the outside through the communication unit 110 . Two or more cameras 121 may be provided according to the use environment.

The microphone 122 receives an external sound signal through a microphone and processes it into electrical voice data. Various noise cancellation algorithms for removing noise generated in the process of receiving an external sound signal may be implemented in the microphone 122 .

The user input unit 130 generates input data for the user to control the operation of the terminal. The user input unit 130 may include a joystick, a key pad, a dome switch, a touch pad (static pressure/capacity), a jog wheel, a jog switch, a mouse, and the like.

The memory 140 may store programs for operation of the control unit 170 and may temporarily store input/output data. The memory 140 may store data related to vibration and sound of various patterns output when a touch is input on the touch screen.

The memory 140 may be a flash memory type, a hard disk type, a multimedia card micro type, a card type memory (eg SD or XD memory, etc.), RAM (Random Access Memory, RAM), SRAM (Static Random Access Memory), ROM (Read-Only Memory, ROM), EEPROM (Electrically Erasable Programmable Read-Only Memory), PROM (Programmable Read-Only Memory), magnetic memory, magnetic It may include at least one type of storage medium among a disk and an optical disk. The terminal 100 may operate in relation to a web storage performing a storage function of the memory 140 on the Internet.

The output unit 150 is for generating an output related to sight, hearing, or touch, and includes a display unit 151, a sound output module 152, an alarm unit 153, and a haptic module 154. can Here, the haptic module 154 may also be included in the user input unit 130.

The display unit 151 displays (outputs) information processed by the terminal 100 .

The display unit 151 may include a liquid crystal display (LCD), a thin film transistor-liquid crystal display (TFT LCD), an organic light-emitting diode (OLED), and a flexible display. display) and a 3D display.

Some of these displays may be of a transparent type or a light transmission type so that the outside can be seen through them. This may be referred to as a transparent display, and a representative example of the transparent display is TOLED (Transparent OLED) and the like. A rear structure of the display unit 151 may also be configured as a light transmissive structure. With this structure, the user can see an object located behind the terminal body through the area occupied by the display unit 151 of the terminal body.

Depending on the implementation form of the terminal 100, two or more display units 151 may exist. For example, in the terminal 100, a plurality of display units may be spaced apart or integrally disposed on one surface, or may be respectively disposed on different surfaces.

The audio output module 152 may output audio data received from the communication unit 110 or stored in the memory 140 . The sound output module 152 also outputs a sound signal related to data acquired by the terminal 100 (eg, a failure notification signal, etc.). The sound output module 152 may include a receiver, a speaker, a buzzer, and the like.

The alarm unit 153 outputs a signal for notifying occurrence of an event in the terminal 100 . Examples of events generated in the mobile terminal include call signal reception, message reception, key signal input, and touch input. The alarm unit 153 may output a signal for notifying the occurrence of an event in a form other than a video signal or an audio signal, for example, vibration. The video signal or audio signal may be output through the display unit 151 or the audio output module 152, so they 151 and 152 may be classified as part of the alarm unit 153.

A haptic module 154 generates various tactile effects that a user can feel. A representative example of the tactile effect generated by the haptic module 154 is vibration. The strength and pattern of vibration generated by the haptic module 154 can be controlled. For example, different vibrations may be synthesized and output or sequentially output.

In addition to vibration, the haptic module 154 responds to stimuli such as a pin array vertically moving with respect to the contact skin surface, air blowing force or suction force through a nozzle or suction port, rubbing against the skin surface, contact of an electrode, electrostatic force, and the like. It is possible to generate various tactile effects, such as the effect of heat absorption and the effect of reproducing the feeling of coolness and warmth using an element capable of absorbing heat or generating heat.

The haptic module 154 not only transmits a tactile effect through direct contact, but also can be implemented so that a user can feel a tactile effect through a muscle sense such as a finger or an arm. Two or more haptic modules 154 may be provided according to the configuration of the monitoring device 400 .

The power supply unit 160 receives external power and internal power under the control of the controller 170 and supplies power required for operation of each component.

A controller 170 controls overall operations of the terminal 100 in general.

Various embodiments described herein may be embodied in a recording medium readable by a computer or similar device using, for example, software, hardware, or a combination thereof.

According to hardware implementation, the embodiments described herein include application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable gate arrays (FPGAs), It may be implemented using at least one of processors, controllers, micro-controllers, microprocessors, and other electrical units for performing functions. These may be implemented by the controller 170.

According to software implementation, embodiments such as procedures or functions may be implemented together with a separate software module to perform at least one function or operation. The software code may be implemented by a software application written in any suitable programming language. The software code may be stored in the memory 140 and executed by the controller 170 .

3 is a block diagram showing the configuration of a server.

As shown in FIG. 3 , the server 300 may include a data transmission/reception unit 310 , a data analysis unit 320 , an electrocardiogram database 330 and a data learning unit 340 .

The data transmission/reception unit 310 may include the same configuration as the communication unit 110 described in FIG. 2 . The data transmission/reception unit 310 may receive single-channel electrocardiogram data from the terminal 100 . The data transmission/reception unit 310 may transmit the electrocardiogram reading result to the terminal 100 .

The data analyzer 320 pre-processes and analyzes the single-channel electrocardiogram data obtained from the data transceiver 310 to generate an electrocardiogram reading result. Specifically, the data analysis unit 320 may include a time series data analysis unit 321 , an image data analysis unit 322 and a data analysis result integration unit 323 .

The time-series data analysis unit 321 pre-processes single-channel electrocardiogram data and diagnoses specific diseases using the PQRST peak. The time-series data analysis unit 321 may diagnose detailed diseases according to a judgment algorithm that enhances the existing diagnosis method using the PQRST peak with a deep learning-based algorithm. In a specific embodiment, the time series data analysis unit 321 may increase the accuracy of diagnosis by continuously reinforcing the criterion based on the interval between QRSs or the width of QRSs with a deep learning-based algorithm.

The time-series data analyzer 321 may acquire raw data of the electrocardiogram and perform electrocardiogram diagnosis through region segmentation, data standardization and feature extraction, and PQST peak detection based on deep learning.

The image data analyzer 322 analyzes the single-channel electrocardiogram waveform image to diagnose specific diseases. The image data analyzer 322 diagnoses detailed diseases according to a reading algorithm generated based on expert judgment data on the electrocardiogram pattern image. Specifically, the results of actual doctors reading ECG pattern images over the past decades have been accumulated, and these big data can be databased through machine learning and deep learning algorithms. The image data analyzer 322 may read detailed diseases based on the electrocardiogram reading database.

The data analysis result integration unit 323 integrates the analysis result of the time series data analysis unit 321 and the analysis result of the image data analysis unit 322 . Specifically, the data analysis result integrator 323 synthesizes the detailed diseases determined by the time series data analysis unit 321 and the detailed diseases determined by the image data analysis unit 322 to derive a single detailed disease reading result.

In one embodiment, the data analysis result integration unit 323 determines whether there is an overlapping disease between one or more detailed diseases analyzed by the time series data analysis unit 321 and one or more detailed diseases analyzed by the image data analysis unit 322. It is possible to create a detailed disease reading list by determining. The detailed disease reading list may be a list including a predetermined number of next-ranked detailed diseases from the most probable detailed disease based on the current electrocardiogram data. The data analysis result integration unit 323 may generate an integrated list based on overlapping detailed diseases and ranks of detailed diseases in each reading result.

Meanwhile, as a result of comparison in the data analysis result integrator 323, there may be cases in which ranking conflicts occur between different diseases. For example, as a result of electrocardiogram reading, the time series data analysis unit 321 reads disease a as the second priority and reads disease b as the third priority. In the image data analyzer 322, disease a is read as the third priority, and disease b The disease can be read in second order.

In this case, in one embodiment, the data analysis result integrator 323 may utilize weights for each disease and create an integrated list according to the weights. The weight may be determined according to the severity of the disease. Here, weights for each disease can be strengthened through learning.

In another embodiment, the data analysis result integration unit 323 may utilize weights between two analysis units for each disease. For example, for disease a, the analysis result of the time series data analyzer 321 may be more reliable than the analysis result of the image data analyzer 322. In this case, the time series data analyzer 321 is ) may have a higher weight. Weights between analysis units may be strengthened through learning.

The ECG database 330 may have the same configuration as the memory 140 described in FIG. 2 or may refer to data stored in the memory. The electrocardiogram database 330 may store big data used for image data analysis. Big data here may be electrocardiogram images accumulated for decades and results of experts' interpretation of them.

In addition, the electrocardiogram database 330 may store electrocardiogram reading big data using the PQST peak utilized by the time series data analyzer 321 . In addition, the electrocardiogram database 330 may store one or more weight data used by the data analysis result integrator 323 .

The data learning unit 340 learns and reinforces the time series data analysis result and the image data analysis result through a learning algorithm including machine learning or deep learning. In a specific embodiment, the time series data analysis unit 321 may increase the accuracy of diagnosis by continuously reinforcing the criterion based on the interval between QRSs or the width of QRSs with a deep learning-based algorithm.

The enhanced decision algorithm may be stored in the electrocardiogram database 330 or provided to the time series data analyzer 321 and the image data analyzer 322 . In addition, the data learning unit 340 may learn and reinforce the weights used by the data analysis result integration unit 323 according to the reading result. Likewise, the reinforced weight data may be stored in the electrocardiogram database 330 or provided to the data analysis result integration unit.

In addition, the data analysis unit 320 and the data learning unit 340 may be operated by a control unit included in the server 300, and the control unit of the server refers to the description of the control unit 170 described in FIG. 2. do.

Meanwhile, among the above configurations, the data analysis unit 320 or the data learning unit 340 may be a component of the terminal 100 other than the server 300 or a module executed by an application running on the terminal 100. may be In other words, the data analysis unit 320 or the data learning unit 340 may be implemented in the terminal 100 or the server 300 according to a specific embodiment.

4 is a flowchart illustrating an electrocardiogram reading procedure according to an embodiment of the present invention.

Each step described in FIG. 4 may be performed through the control unit of the terminal 100 or the server 300 according to an embodiment. Hereinafter, each procedure will be described according to the embodiment of FIG. 3 .

The server 300 acquires single-channel electrocardiogram data (S101). The single-channel ECG data obtained by the server 300 is ECG data obtained from the above-described single-channel ECG measurement device 200, and is obtained through only two electrodes, unlike general 12-channel ECG measurement data. The single-channel electrocardiogram measuring device 200 may acquire electrocardiogram data from both arms through two electrodes.

The server 300 may directly receive single-channel ECG data from the single-channel ECG measurement device 200 or through the terminal 100 . If electrocardiogram data is received through the terminal, the electrocardiogram data may be output to the terminal 100 in real time and displayed to the user.

The server 300 pre-processes the acquired single-channel ECG data (S103). Specifically, the server 300 may remove noise from acquired single-channel ECG data and convert the data into 2D image data. For example, the server 300 may remove noise and amplify a signal using Fourier, wavelet transform, or Taken's embedding. In addition, the server 300 may create a 2D multi-channel image by combining the converted 2D image data results.

The server 300 performs area division on the preprocessed data (S105). Basically, the server 300 creates beat data, which is electrocardiogram data between R peaks, using R peaks that are easy to find by an algorithm.

In addition, the server 300 may use a CPD algorithm for more detailed area division. You can estimate the approximate extent of each peak. In this case, the server 300 may divide the ECG data into data having a length smaller than bits.

The server 300 standardizes the divided data and extracts features (S107). Specifically, the server 300 formats and standardizes data. In addition, the server 300 extracts features required for data analysis using a feature extraction algorithm. The features extracted here may be P, Q, S, and T peaks that are difficult to find compared to R peaks, and the server 300 can improve the accuracy of extracting P, Q, S, and T peaks based on deep learning. have.

The server 300 performs primary reading based on deep learning (S109). Here, the primary reading is a reading by time-series data analysis based on the P, Q, S, and T peaks. In addition, as mentioned above, the server 300 may analyze time series data with higher accuracy based on deep learning through the time series data analysis unit 321 .

As a result of the primary reading, one or more main subdiseases and subordinate subdiseases may be read for electrocardiogram data.

The server 300 performs secondary reading based on the expert reading data (S111). Here, the secondary reading is reading by analyzing preprocessed 2D image data. In detail, the server 300 may perform secondary reading through the image data analyzer 322 . Similarly, as a result of the secondary reading, a main detailed disease and one or more subordinate detailed diseases may be read for the electrocardiogram data. In addition, the first reading and the second reading are independent reading procedures and do not affect each other's judgment.

The server 300 integrates the first reading result and the second reading result (S113). In detail, the server 300 integrates the first reading result and the second reading result through the data analysis result integration unit 323 to generate an integrated list. In the integrated list, probable diseases may be listed in descending order from the most probable disease based on electrocardiogram data, and the number of diseases included in the list may be separately determined.

The server 300 outputs the integrated result through the terminal 100 (S115). Specifically, the server 300 transmits the integrated analysis result to the terminal 100, and the terminal 100 outputs the analysis result through the display unit 151. At this time, the terminal 100 may further output additional information as well as reading disorders, which will be described with reference to FIG. 5 .

5 shows an example of a user interface including a final read result displayed on a terminal.

As shown in FIG. 5, the user interface including the final reading result is the state of the patient's electrocardiogram (normal, consultation required, immediate consultation), display of the disease zone (QRS width (0.32 sec), text description of the disease (ventricular description of tachycardia), a reading probability (95.5%) of a disease, a reading date, an electrocardiogram graph, or a cardiologist connection menu In addition, the user interface including the final reading result may include the patient's electrocardiogram A certain number of diseases suspected in the data may be displayed, and may be displayed in order of high probability.

In the case of the electrocardiogram state, whether the patient's electrocardiogram is normal or abnormal, and whether it is an urgent condition requiring consultation with a specialist is displayed in an easy-to-understand manner for the general public. Status information of the patient's electrocardiogram may be transferred from the server 300 along with a final reading result.

In one embodiment, the cardiologist connection menu may be deactivated in connection with the patient's electrocardiogram status information, when the cardiologist is in a normal state. In another embodiment, in connection with the patient's electrocardiogram status information, if immediate counseling is required or it is determined to be very serious, the terminal may automatically connect the cardiologist.

The above-described present invention can be implemented as computer readable code on a medium on which a program is recorded. The computer-readable medium includes all types of recording devices in which data that can be read by a computer system is stored. Examples of computer-readable media include Hard Disk Drive (HDD), Solid State Disk (SSD), Silicon Disk Drive (SDD), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. , and also includes those implemented in the form of a carrier wave (eg, transmission over the Internet).

Although the present invention has been described with reference to the embodiments shown in the drawings, this is only exemplary, and those skilled in the art will understand that various modifications and equivalent other embodiments are possible therefrom. Therefore, the true technical protection scope of the present invention should be determined by the technical spirit of the appended claims.

Claims (9)

an electrocardiogram measuring device that acquires electrocardiogram data; and
A server for reading detailed diseases using the electrocardiogram data;
The server,
A time-series data analysis unit that first reads detailed diseases through P, Q, R, S, and T peak detection of the electrocardiogram data;
An image data analysis unit that secondaryly reads detailed diseases through an AI-based image data analysis algorithm, and
An AI-based electrocardiogram reading system comprising a data analysis result integrator for generating a reading result by integrating the first reading and the second reading.
According to claim 1,
The server,
An electrocardiogram database including big data used for the first reading and the second reading, and
The AI-based electrocardiogram reading system further comprising a data learning unit configured to enhance the electrocardiogram database by learning analysis results.
According to claim 2,
The AI-based electrocardiogram reading system for analyzing the characteristics of the detected P, Q, R, S, and T peaks by the time series data analysis unit using a reading algorithm generated based on big data stored in the electrocardiogram database.
According to claim 3,
The characteristics of the P, Q, R, S, T peaks include the width value of the QRS, the PR interval value, the interval deviation value between QRSs, the PR interval deviation value, the presence or absence of P waves, and the correlation value of P and QRS numbers. AI based electrocardiogram reading system.
According to claim 1,
The server outputs the reading result to the terminal,
The information output to the terminal includes an electrocardiogram status of the patient, an electrocardiogram graph, a display of a disease section, a reading probability of the disease, and a text description of the disease.
According to claim 5,
The information output to the terminal includes at least one suspected disease information, and the one or more suspected disease information is displayed in order of highest probability among suspected diseases.
According to claim 5,
The information output to the terminal is an AI-based electrocardiogram reading system including a cardiologist connection menu.
According to claim 1,
The data analysis result integrator integrates the first read and ranked subdiseases and the second read and ranked subdiseases to create a detailed disease reading list including subdiseases with the highest probability to the next ranked subdiseases. AI-based electrocardiogram reading system.
According to claim 8,
The data analysis result integrator compares the first read and ranked detailed disease with the second read and ranked detailed disease, and if there is a ranking conflict between different diseases, a detailed disease read list is generated according to the weight of each disease. AI-based electrocardiogram reading system.

Images