CN111657921A - Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium - Google Patents
Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN111657921A CN111657921A CN202010623467.7A CN202010623467A CN111657921A CN 111657921 A CN111657921 A CN 111657921A CN 202010623467 A CN202010623467 A CN 202010623467A CN 111657921 A CN111657921 A CN 111657921A
- Authority
- CN
- China
- Prior art keywords
- data
- electrocardio
- real
- time
- abnormal
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B5/00—Measuring for diagnostic purposes; Identification of persons
- A61B5/72—Signal processing specially adapted for physiological signals or for diagnostic purposes
- A61B5/7235—Details of waveform analysis
- A61B5/7264—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems
- A61B5/7267—Classification of physiological signals or data, e.g. using neural networks, statistical classifiers, expert systems or fuzzy systems involving training the classification device
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/044—Recurrent networks, e.g. Hopfield networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Theoretical Computer Science (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Mathematical Physics (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Computing Systems (AREA)
- Data Mining & Analysis (AREA)
- Software Systems (AREA)
- Computational Linguistics (AREA)
- Psychiatry (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physiology (AREA)
- Fuzzy Systems (AREA)
- Signal Processing (AREA)
- Pathology (AREA)
- Heart & Thoracic Surgery (AREA)
- Medical Informatics (AREA)
- Surgery (AREA)
- Animal Behavior & Ethology (AREA)
- Public Health (AREA)
- Veterinary Medicine (AREA)
- Measurement And Recording Of Electrical Phenomena And Electrical Characteristics Of The Living Body (AREA)
Abstract
The application relates to a real-time electrocardio abnormality monitoring method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring real-time electrocardiogram data and clinical characteristic data; updating the cache data according to the real-time electrocardiogram data to obtain updated cache data; inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data; judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report; and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired. By adopting the method, the accuracy of the electrocardio abnormal probability prediction data can be improved, and the monitoring efficiency can be improved.
Description
Technical Field
The application relates to the technical field of artificial intelligence, in particular to a real-time electrocardio abnormality monitoring method and device, computer equipment and a storage medium.
Background
The electrocardiosignal analysis and real-time monitoring technology has great demand in the scenes of sickbed electrocardio monitoring, 120 ambulance electrocardio monitoring and the like, but the existing technical method still has a plurality of defects.
The existing electrocardiosignal analysis and real-time monitoring technologies are mainly divided into two categories. The first category is based on electrocardiographic signal feature engineering and relies on electrocardiographic medical domain knowledge. The method identifies each heart Beat (Beat) in a section of electrocardio signals through a digital signal processing method, analyzes the characteristics of each heart Beat, and finally realizes the diagnosis of the electrocardio abnormality by combining the knowledge in the medical field of electrocardio. The second category directly classifies the electrocardio-abnormality of a section of electrocardio-signals by utilizing a deep learning method.
The existing first-class electrocardiosignal analysis and real-time monitoring technology has the following defects: the method has the advantages that characteristic engineering is required, the process is complex, the workload is large, and the result quality is easily influenced by interference signals, so that the method is poor in robustness and expandability, depends on a large amount of knowledge in the medical field of electrocardio, has a natural bottleneck in analysis effect, is high in implementation difficulty and long in development period.
The existing second type electrocardiosignal analysis and real-time monitoring technology has the following defects: the characteristic wave information is not considered when modeling the electrocardiosignal, and the characteristic wave information can not be provided for a doctor, the characteristic wave information is important in diagnosis of the doctor, and the accuracy of diagnosis can be improved by considering the characteristic wave information. The method comprises a recurrent neural network, data calculation cannot be performed in parallel, and time overhead is large. Differences in other clinical characteristics of patients such as age, gender, etc. are not generally considered. No erroneous prediction information is utilized, and the electrocardiosignal analysis and real-time monitoring method is continuously improved.
Disclosure of Invention
In view of the above, it is necessary to provide a real-time monitoring method, an apparatus, a computer device and a storage medium for cardiac electrical anomalies, which can solve at least one of the problems.
A real-time monitoring method for electrocardio abnormality comprises the following steps:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
Preferably, the updating the cache data includes:
deleting part of the cached data according to the length of the real-time electrocardiogram data to obtain part of the cached data;
and splicing the real-time electrocardiogram data with part of the cache data to obtain the updated cache data.
Preferably, the cache data is electrocardiographic data obtained in advance and time-continuous with the real-time electrocardiographic data.
Preferably, the length of the real-time electrocardiographic data is shorter than that of the cached data.
Preferably, the training of the neural network for recognizing the cardiac electrical abnormality specifically includes:
acquiring electrocardiogram data and clinical characteristic data related to the electrocardiogram data, and preprocessing the electrocardiogram data to obtain electrocardiogram abnormal labels;
inputting the electrocardio data and the clinical characteristic data into the electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
calculating an error function according to the electrocardio abnormal probability prediction data and the electrocardio abnormal label to obtain a prediction error;
and optimizing the electrocardio abnormality recognition neural network according to the prediction error until the prediction error reaches the expectation, finishing training and obtaining the trained electrocardio abnormality recognition neural network.
Preferably, the step of inputting the electrocardiographic data and the clinical characteristic data into the electrocardiographic abnormality recognition neural network to obtain electrocardiographic abnormality probability prediction data includes:
inputting the electrocardio data into a trained electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
combining the characteristic segmentation data and the electrocardiogram data to obtain first combined data;
inputting the first combined data into an electrocardio high-level feature extraction neural network to obtain electrocardio high-level feature data;
combining the electrocardio high-level characteristic data and the clinical characteristic data to obtain second combined data;
and performing full-connection layer operation on the second combined data to obtain an operation result, inputting the operation result into a Softmax logistic regression model, and obtaining electrocardio abnormal probability prediction data.
Preferably, the training of the electrocardiographic feature wave segmentation network specifically includes:
acquiring electrocardiogram data, and preprocessing the electrocardiogram data to obtain an electrocardiogram characteristic wave label;
inputting the electrocardio data into the electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
performing error calculation according to the characteristic segmentation data and the electrocardio characteristic wave label to obtain a segmentation error;
and optimizing the electrocardio characteristic wave segmentation network according to the segmentation error until the segmentation error reaches the expectation, finishing training and obtaining the trained electrocardio characteristic wave segmentation network.
A device for real-time monitoring of cardiac electrical abnormalities, the device comprising:
the data acquisition module is used for acquiring real-time electrocardiogram data and clinical characteristic data;
the data updating module is used for updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
the abnormal probability prediction acquisition module is used for inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormal recognition neural network to obtain electrocardio abnormal probability prediction related to the real-time electrocardio data;
the abnormality judgment module is used for predicting and judging whether the real-time electrocardio data is abnormal according to the electrocardio abnormal probability, and if the judgment result is abnormal, performing an abnormal report;
and the ending instruction acquisition module is used for acquiring the real-time electrocardio data again for monitoring until an electrocardio monitoring ending instruction is acquired.
A computer device comprising a memory and a processor, the memory storing a computer program, the processor implementing the following steps when executing the computer program:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
A computer-readable storage medium, on which a computer program is stored which, when executed by a processor, carries out the steps of:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
According to the real-time electrocardio abnormality monitoring method, the real-time electrocardio data and the previously acquired electrocardio data are spliced, and then the spliced data are input into the trained electrocardio abnormality recognition neural network, so that accurate electrocardio abnormality probability prediction data are obtained. And the obtained electrocardio abnormal probability prediction data is more accurate by considering other different clinical characteristic data of each patient.
Drawings
FIG. 1 is a schematic flow chart of a real-time monitoring method for cardiac electrical anomalies in one embodiment;
FIG. 2 is a schematic diagram of a process for training an ECG anomaly recognition neural network according to an embodiment;
FIG. 3 is a schematic diagram of a process for obtaining prediction data of abnormal cardiac electrical probability in one embodiment;
FIG. 4 is a schematic diagram of a neural network for recognizing cardiac electrical anomalies in another embodiment;
FIG. 5 is a block diagram of an embodiment of a device for real-time monitoring of an abnormal electrocardiogram;
FIG. 6 is a diagram illustrating an internal structure of a computer device according to an embodiment.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the present application is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the present application and are not intended to limit the present application.
As shown in fig. 1, a real-time monitoring method for electrocardiographic abnormality is provided, which includes:
103, inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
104, judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and reporting the abnormality if the judgment result is abnormal;
and 105, acquiring the real-time electrocardio data again for monitoring until an electrocardio monitoring ending instruction is acquired.
In step 101, the acquired real-time electrocardiographic data is acquired standard 12-lead electrocardiographic data. The obtained clinical characteristic data refers to the age, the sex and the like of individuals with different electrocardiosignals.
In step 102, the buffered data is previously acquired electrocardiographic data that is temporally continuous with the real-time electrocardiographic data. And when the real-time electrocardiogram data is acquired, the acquisition length is shorter than that of the cache data.
Correspondingly deleting partial data of the cache data according to the length of the real-time electrocardiogram data to obtain partial cache data; and splicing the real-time electrocardiogram data with part of the cache data to obtain updated cache data.
When the cache data is updated, the part of the cache data which is earlier consistent with the length of the real-time electrocardiogram data is deleted, and then the real-time electrocardiogram data is spliced with the rest cache data to obtain new cache data.
In step 103, the electrocardiographic abnormality probability prediction data includes an electrocardiographic abnormality prediction probability for the segment of real-time electrocardiographic data.
In step 104, the prediction data of the electrocardiographic abnormality probability may be determined by a threshold method. For example, if the preset threshold is 50%, the number of abnormal probability predictions is greater than 50%, and it is determined that the electrocardiographic abnormality is positive. And if the prediction of the abnormal probability is less than 50%, judging that the electrocardio abnormality is negative.
After the judgment, if the judgment result is positive, the abnormity is reported.
In step 105, the monitoring process is to continuously loop steps 101 to 104 until an instruction to stop monitoring is received, stop monitoring and no longer obtain real-time electrocardiographic data.
In this embodiment, if an error prediction feedback is received after an abnormality report, indicating that the electrocardiographic abnormality probability prediction data is erroneous, the cache data and the electrocardiographic abnormality probability prediction data corresponding to the abnormality report are recorded. And after a certain amount of error prediction feedback is collected, retraining all error electrocardio abnormal probability prediction data to the electrocardio abnormal recognition neural network, thereby reducing the error prediction probability of the electrocardio abnormal recognition neural network and improving the accuracy of the electrocardio abnormal recognition neural network.
As shown in fig. 2, the specific steps of training the neural network for recognizing cardiac electrical abnormality include:
and 204, optimizing the electrocardio abnormality recognition neural network according to the prediction error until the prediction error reaches the expectation, finishing training and obtaining the trained electrocardio abnormality recognition neural network.
In step 201, the acquired electrocardiographic data and the clinical characteristic are not necessarily acquired in real time, and they have a correlation.
All the electrocardio data are subjected to noise suppression processing. The electrocardio abnormality is marked as vector data, and the length of the vector is the number of the electrocardio abnormality types. For example, if there are 10 types of abnormal electrocardiographic types, the abnormal electrocardiographic is marked with a length of 10, and if there are several types of abnormal electrocardiographic types in the section of electrocardiographic data, the abnormal electrocardiographic is marked at a position corresponding to the abnormal electrocardiographic marking vector.
In step 202, as shown in fig. 3-4, inputting the electrocardiographic data and the clinical characteristic data into the electrocardiographic abnormality recognition neural network to obtain electrocardiographic abnormality probability prediction data includes:
and 305, performing full-connection layer operation on the second combined data to obtain an operation result, inputting the operation result into a Softmax logistic regression model, and obtaining electrocardio abnormal probability prediction data.
In step 301, the size of the electrocardiographic data is 1 × 12 × L, where 12 is the number of leads, L is the time for acquiring the electrocardiographic data, i.e., the length of the electrocardiographic data, and the obtained feature segmentation data is 1 × 12 × L.
In step 302, the first combined data of 2 × 12 × L is obtained by combining the electrocardiographic data and the characteristic style data using a Depth Concat function.
In step 303, the high-level electrocardiographic feature data is a vector with length M.
In the embodiment, the electrocardio high-level feature extraction neural network can adopt two-dimensional ResNet 50.
In step 304, the clinical characteristic data is a vector combination with a length of K, and the value of K is the number of categories of clinical characteristics. And combining the high-level electrocardio data with the length of M and the clinical characteristic data with the length of K by adopting a Concat function to obtain second data combined data with the length of M + K.
In step 305, the prediction data of the electrocardiographic abnormality probability includes prediction probabilities corresponding to various electrocardiographic abnormalities in the section of electrocardiographic data.
In this embodiment, the trained electrocardiographic feature wave segmentation network is used to extract feature segmentation data, and in other embodiments, other methods may be used to obtain the feature segmentation data.
In this embodiment, before training the electrocardiogram anomaly recognition neural network, the electrocardiogram characteristic wave segmentation network is trained first, so that the electrocardiogram characteristic wave segmentation network has the capability of extracting characteristic segmentation data.
The training of the electrocardio characteristic wave segmentation network specifically comprises the following steps: acquiring electrocardiogram data, and preprocessing the electrocardiogram data to obtain electrocardiogram characteristic wave labels; inputting the electrocardio data into an electrocardio characteristic wave segmentation network to obtain characteristic segmentation data; performing error calculation according to the feature segmentation data and the electrocardio feature wave label to obtain a segmentation error; and optimizing the electrocardio characteristic wave segmentation network according to the segmentation error until the segmentation error reaches the expectation, finishing training and obtaining the trained electrocardio characteristic wave segmentation network.
In this embodiment, the electrocardiographic data includes a plurality of signal waves, each signal wave is divided into different stages, and the electrocardiographic characteristic wave is labeled to label a stage of interest in the signal wave. And the electrocardio data used for training the electrocardio characteristic wave style network is consistent with the electrocardio abnormal category contained in the electrocardio data used for training the electrocardio abnormal recognition neural network.
Before training by using the electrocardiogram data, the electrocardiogram data is subjected to noise suppression processing.
And when error calculation is carried out according to the feature segmentation data and the electrocardio feature wave standard, a multi-classification cross entropy function is adopted for calculation to obtain a feature error.
In step 203, the classification cross entropy may be used to calculate the prediction error.
In step 204, the electrocardiogram anomaly recognition neural network is optimized according to the prediction error by using an optimizer.
In this embodiment, when the cardiac anomaly recognition neural network is trained, the weights of the cardiac characteristic wave segmentation network inside the cardiac anomaly recognition neural network are initialized.
According to the real-time monitoring method for the electrocardio abnormality, clinical information of patients such as age and sex is fused when the electrocardio data are analyzed, so that differences of electrocardiosignals of the patients with different clinical characteristics can be distinguished, and higher accuracy is obtained. And utilizes the characteristic wave information to provide the special detection wave information, and the characteristic wave information is important in the diagnosis of doctors. The characteristic wave information is considered when the electrocardio abnormality recognition neural network is trained, and the accuracy of electrocardio abnormality probability prediction data can be improved.
When the electrocardio abnormity is monitored in real time, the error prediction of each time is recorded, the recorded error information is fed back to the neural network, and the neural network is trained again, so that the error prediction probability of the method is reduced, and the accuracy of the method is improved. When acquiring the electrocardio data, the electrocardio data cache pool is utilized, and the electrocardio signal analysis is started without waiting for acquiring the signal with the length of L except the first section of electrocardio data, so that the timeliness is higher.
It should be understood that although the various steps in the flow charts of fig. 1-3 are shown in order as indicated by the arrows, the steps are not necessarily performed in order as indicated by the arrows. The steps are not performed in the exact order shown and described, and may be performed in other orders, unless explicitly stated otherwise. Moreover, at least some of the steps in fig. 1-3 may include multiple sub-steps or multiple stages that are not necessarily performed at the same time, but may be performed at different times, and the order of performance of the sub-steps or stages is not necessarily sequential, but may be performed in turn or alternating with other steps or at least some of the sub-steps or stages of other steps.
In one embodiment, as shown in fig. 5, there is provided a real-time monitoring device for cardiac electrical abnormality, including: a data obtaining module 501, a data updating module 502, an electrocardiographic abnormality probability prediction obtaining module 503, an abnormality judging module 504 and an ending instruction obtaining module 505, wherein:
the data acquisition module 501 is configured to acquire real-time electrocardiographic data and clinical characteristic data;
a data updating module 502, configured to update the cache data according to the real-time electrocardiographic data to obtain updated cache data;
an abnormal probability prediction obtaining module 503, configured to input the updated cache data and the clinical characteristic data into the trained cardiac electrical abnormality recognition neural network, so as to obtain a cardiac electrical abnormal probability prediction related to the real-time cardiac electrical data;
an abnormality judgment module 504, configured to judge whether the real-time electrocardiographic data is abnormal according to the electrocardiographic abnormal probability prediction, and if the judgment result is abnormal, perform an abnormal report;
and an ending instruction obtaining module 505, configured to obtain the real-time electrocardiographic data again for monitoring until an electrocardiographic monitoring ending instruction is obtained.
For specific limitations of the real-time monitoring device for cardiac electrical abnormalities, reference may be made to the above limitations on the real-time monitoring method for cardiac electrical abnormalities, which are not described herein again. All modules in the electrocardio abnormality real-time monitoring device can be completely or partially realized through software, hardware and a combination thereof. The modules can be embedded in a hardware form or independent from a processor in the computer device, and can also be stored in a memory in the computer device in a software form, so that the processor can call and execute operations corresponding to the modules.
In one embodiment, a computer device is provided, which may be a server, and its internal structure diagram may be as shown in fig. 6. The computer device includes a processor, a memory, a network interface, and a database connected by a system bus. Wherein the processor of the computer device is configured to provide computing and control capabilities. The memory of the computer device comprises a nonvolatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The internal memory provides an environment for the operation of an operating system and computer programs in the non-volatile storage medium. The database of the computer device is used for storing cache data. The network interface of the computer device is used for communicating with an external terminal through a network connection. The computer program is executed by a processor to realize a real-time electrocardio abnormality monitoring method.
Those skilled in the art will appreciate that the architecture shown in fig. 6 is merely a block diagram of some of the structures associated with the disclosed aspects and is not intended to limit the computing devices to which the disclosed aspects apply, as particular computing devices may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
In one embodiment, a computer device is provided, comprising a memory and a processor, the memory having a computer program stored therein, the processor implementing the following steps when executing the computer program:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
acquiring electrocardiogram data and clinical characteristic data related to the electrocardiogram data, and preprocessing the electrocardiogram data to obtain electrocardiogram abnormal labels;
inputting the electrocardio data and the clinical characteristic data into the electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
calculating an error function according to the electrocardio abnormal probability prediction data and the electrocardio abnormal label to obtain a prediction error;
and optimizing the electrocardio abnormality recognition neural network according to the prediction error until the prediction error reaches the expectation, finishing training and obtaining the trained electrocardio abnormality recognition neural network.
In one embodiment, the processor, when executing the computer program, further performs the steps of:
inputting the electrocardio data into a trained electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
combining the characteristic segmentation data and the electrocardiogram data to obtain first combined data;
inputting the first combined data into an electrocardio high-level feature extraction neural network to obtain electrocardio high-level feature data;
combining the electrocardio high-level characteristic data and the clinical characteristic data to obtain second combined data;
and performing full-connection layer operation on the second combined data to obtain an operation result, inputting the operation result into a Softmax logistic regression model, and obtaining electrocardio abnormal probability prediction data.
In one embodiment, a computer-readable storage medium is provided, having a computer program stored thereon, which when executed by a processor, performs the steps of:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
In one embodiment, the computer program when executed by the processor further performs the steps of:
acquiring electrocardiogram data and clinical characteristic data related to the electrocardiogram data, and preprocessing the electrocardiogram data to obtain electrocardiogram abnormal labels;
inputting the electrocardio data and the clinical characteristic data into the electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
calculating an error function according to the electrocardio abnormal probability prediction data and the electrocardio abnormal label to obtain a prediction error;
and optimizing the electrocardio abnormality recognition neural network according to the prediction error until the prediction error reaches the expectation, finishing training and obtaining the trained electrocardio abnormality recognition neural network.
In one embodiment, the computer program when executed by the processor further performs the steps of:
inputting the electrocardio data into a trained electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
combining the characteristic segmentation data and the electrocardiogram data to obtain first combined data;
inputting the first combined data into an electrocardio high-level feature extraction neural network to obtain electrocardio high-level feature data;
combining the electrocardio high-level characteristic data and the clinical characteristic data to obtain second combined data;
and performing full-connection layer operation on the second combined data to obtain an operation result, inputting the operation result into a Softmax logistic regression model, and obtaining electrocardio abnormal probability prediction data.
It will be understood by those skilled in the art that all or part of the processes of the methods of the embodiments described above can be implemented by hardware instructions of a computer program, which can be stored in a non-volatile computer-readable storage medium, and when executed, can include the processes of the embodiments of the methods described above. Any reference to memory, storage, database, or other medium used in the embodiments provided herein may include non-volatile and/or volatile memory, among others. Non-volatile memory can include read-only memory (ROM), Programmable ROM (PROM), Electrically Programmable ROM (EPROM), Electrically Erasable Programmable ROM (EEPROM), or flash memory. Volatile memory can include Random Access Memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms such as Static RAM (SRAM), Dynamic RAM (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDRSDRAM), Enhanced SDRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), Rambus Direct RAM (RDRAM), direct bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).
The technical features of the above embodiments can be arbitrarily combined, and for the sake of brevity, all possible combinations of the technical features in the above embodiments are not described, but should be considered as the scope of the present specification as long as there is no contradiction between the combinations of the technical features.
The above-mentioned embodiments only express several embodiments of the present application, and the description thereof is more specific and detailed, but not construed as limiting the scope of the invention. It should be noted that, for a person skilled in the art, several variations and modifications can be made without departing from the concept of the present application, which falls within the scope of protection of the present application. Therefore, the protection scope of the present patent shall be subject to the appended claims.
Claims (10)
1. The real-time electrocardio abnormality monitoring method is characterized by comprising the following steps:
acquiring real-time electrocardiogram data and clinical characteristic data;
updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
judging whether the real-time electrocardio data are abnormal according to the electrocardio abnormal probability prediction data, and if the judgment result is abnormal, performing an abnormal report;
and re-acquiring the real-time electrocardio data for monitoring until an electrocardio monitoring ending instruction is acquired.
2. The method for monitoring the electrocardiographic abnormality in real time according to claim 1, wherein the updating the cache data comprises:
deleting part of the cached data according to the length of the real-time electrocardiogram data to obtain part of the cached data;
and splicing the real-time electrocardiogram data with part of the cache data to obtain the updated cache data.
3. The method for real-time monitoring of electrocardiographic abnormality according to claim 2, wherein the cache data is electrocardiographic data obtained in advance in time-series with the real-time electrocardiographic data.
4. The method for real-time monitoring of cardiac electrical abnormalities as set forth in claim 2, wherein the real-time cardiac electrical data has a length shorter than the length of the buffered data.
5. The method for monitoring the electrocardiographic abnormality in real time according to claim 1, wherein the training of the electrocardiographic abnormality recognition neural network specifically includes:
acquiring electrocardiogram data and clinical characteristic data related to the electrocardiogram data, and preprocessing the electrocardiogram data to obtain electrocardiogram abnormal labels;
inputting the electrocardio data and the clinical characteristic data into the electrocardio abnormality recognition neural network to obtain electrocardio abnormality probability prediction data;
calculating an error function according to the electrocardio abnormal probability prediction data and the electrocardio abnormal label to obtain a prediction error;
and optimizing the electrocardio abnormality recognition neural network according to the prediction error until the prediction error reaches the expectation, finishing training and obtaining the trained electrocardio abnormality recognition neural network.
6. The method for monitoring the cardiac electrical abnormality in real time according to claim 5, wherein the step of inputting the cardiac electrical data and the clinical characteristic data into the cardiac electrical abnormality recognition neural network to obtain the prediction data of the cardiac electrical abnormality probability comprises:
inputting the electrocardio data into a trained electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
combining the characteristic segmentation data and the electrocardiogram data to obtain first combined data;
inputting the first combined data into an electrocardio high-level feature extraction neural network to obtain electrocardio high-level feature data;
combining the electrocardio high-level characteristic data and the clinical characteristic data to obtain second combined data;
and performing full-connection layer operation on the second combined data to obtain an operation result, inputting the operation result into a Softmax logistic regression model, and obtaining electrocardio abnormal probability prediction data.
7. The method for monitoring the electrocardiographic abnormality in real time according to claim 6, wherein the training of the electrocardiographic characteristic wave division network specifically includes:
acquiring electrocardiogram data, and preprocessing the electrocardiogram data to obtain an electrocardiogram characteristic wave label;
inputting the electrocardio data into the electrocardio characteristic wave segmentation network to obtain characteristic segmentation data;
performing error calculation according to the characteristic segmentation data and the electrocardio characteristic wave label to obtain a segmentation error;
and optimizing the electrocardio characteristic wave segmentation network according to the segmentation error until the segmentation error reaches the expectation, finishing training and obtaining the trained electrocardio characteristic wave segmentation network.
8. An apparatus for real-time monitoring of cardiac electrical anomalies, the apparatus comprising:
the data acquisition module is used for acquiring real-time electrocardiogram data and clinical characteristic data;
the data updating module is used for updating the cache data according to the real-time electrocardiogram data to obtain updated cache data;
the abnormal probability prediction acquisition module is used for inputting the updated cache data and the clinical characteristic data into the trained electrocardio abnormal recognition neural network to obtain electrocardio abnormal probability prediction related to the real-time electrocardio data;
the abnormality judgment module is used for predicting and judging whether the real-time electrocardio data is abnormal according to the electrocardio abnormal probability, and if the judgment result is abnormal, performing an abnormal report;
and the ending instruction acquisition module is used for acquiring the real-time electrocardio data again for monitoring until an electrocardio monitoring ending instruction is acquired.
9. A computer device comprising a memory and a processor, the memory storing a computer program, wherein the processor implements the steps of the method of any one of claims 1 to 7 when executing the computer program.
10. A computer-readable storage medium, on which a computer program is stored, which, when being executed by a processor, carries out the steps of the method of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010623467.7A CN111657921A (en) | 2020-07-01 | 2020-07-01 | Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010623467.7A CN111657921A (en) | 2020-07-01 | 2020-07-01 | Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN111657921A true CN111657921A (en) | 2020-09-15 |
Family
ID=72390750
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010623467.7A Pending CN111657921A (en) | 2020-07-01 | 2020-07-01 | Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111657921A (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022121107A1 (en) * | 2020-12-10 | 2022-06-16 | 青岛大爱慈康智能医疗科技有限公司 | Electrocardiographic signal detection method and device |
| CN115862843A (en) * | 2022-12-12 | 2023-03-28 | 哈尔滨医科大学 | Auxiliary identification system and equipment for myocardial muscle calcium egg rising type and cardiovascular diseases |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106214123A (en) * | 2016-07-20 | 2016-12-14 | 杨平 | A kind of electrocardiogram compressive classification method based on degree of depth learning algorithm |
| CN108968951A (en) * | 2018-08-15 | 2018-12-11 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram detecting method, apparatus and system |
| CN109567789A (en) * | 2018-12-03 | 2019-04-05 | 东软集团股份有限公司 | Division processing method, device and the readable storage medium storing program for executing of ECG data |
| CN110353665A (en) * | 2019-08-14 | 2019-10-22 | 苏州大学 | ECG signal signature waveform detection method is singly led based on improved one-dimensional U-net |
| CN110495878A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | Disease forecasting method, apparatus and electronic equipment based on ECG |
| CN110495877A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | A kind of Multi resolution feature extraction method and device based on ECG |
| CN110584649A (en) * | 2019-09-09 | 2019-12-20 | 武汉大学 | Method and system for identifying abnormal electrocardiosignals in dynamic single-lead electrocardiogram |
| CN110897633A (en) * | 2019-12-17 | 2020-03-24 | 安徽心之声医疗科技有限公司 | Electrocardiosignal segmentation method based on deep neural network |
| CN111242098A (en) * | 2020-02-27 | 2020-06-05 | 西安交通大学 | Electrocardiogram data classification method and system combining feature extraction and initiation network |
-
2020
- 2020-07-01 CN CN202010623467.7A patent/CN111657921A/en active Pending
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN106214123A (en) * | 2016-07-20 | 2016-12-14 | 杨平 | A kind of electrocardiogram compressive classification method based on degree of depth learning algorithm |
| CN108968951A (en) * | 2018-08-15 | 2018-12-11 | 武汉中旗生物医疗电子有限公司 | Electrocardiogram detecting method, apparatus and system |
| CN109567789A (en) * | 2018-12-03 | 2019-04-05 | 东软集团股份有限公司 | Division processing method, device and the readable storage medium storing program for executing of ECG data |
| CN110353665A (en) * | 2019-08-14 | 2019-10-22 | 苏州大学 | ECG signal signature waveform detection method is singly led based on improved one-dimensional U-net |
| CN110495878A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | Disease forecasting method, apparatus and electronic equipment based on ECG |
| CN110495877A (en) * | 2019-08-21 | 2019-11-26 | 中国科学院深圳先进技术研究院 | A kind of Multi resolution feature extraction method and device based on ECG |
| CN110584649A (en) * | 2019-09-09 | 2019-12-20 | 武汉大学 | Method and system for identifying abnormal electrocardiosignals in dynamic single-lead electrocardiogram |
| CN110897633A (en) * | 2019-12-17 | 2020-03-24 | 安徽心之声医疗科技有限公司 | Electrocardiosignal segmentation method based on deep neural network |
| CN111242098A (en) * | 2020-02-27 | 2020-06-05 | 西安交通大学 | Electrocardiogram data classification method and system combining feature extraction and initiation network |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2022121107A1 (en) * | 2020-12-10 | 2022-06-16 | 青岛大爱慈康智能医疗科技有限公司 | Electrocardiographic signal detection method and device |
| CN115862843A (en) * | 2022-12-12 | 2023-03-28 | 哈尔滨医科大学 | Auxiliary identification system and equipment for myocardial muscle calcium egg rising type and cardiovascular diseases |
| CN115862843B (en) * | 2022-12-12 | 2024-02-02 | 哈尔滨医科大学 | Auxiliary identification system and equipment for myocardial troponin elevation type and cardiovascular diseases |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11134880B2 (en) | Automatic method to delineate or categorize an electrocardiogram | |
| US10758139B2 (en) | Automatic method to delineate or categorize an electrocardiogram | |
| US20200337580A1 (en) | Time series data learning and analysis method using artificial intelligence | |
| US11331034B2 (en) | Automatic method to delineate or categorize an electrocardiogram | |
| CN113705685B (en) | Disease feature recognition model training, disease feature recognition method, device and equipment | |
| CN111759304B (en) | Electrocardiogram abnormity identification method and device, computer equipment and storage medium | |
| CN111915584A (en) | Focus follow-up assessment method and system based on CT (computed tomography) image | |
| CN111657921A (en) | Real-time electrocardio abnormality monitoring method and device, computer equipment and storage medium | |
| Mehmood et al. | Systematic framework to predict early-stage liver carcinoma using hybrid of feature selection techniques and regression techniques | |
| CN111956214A (en) | QRS wave automatic detection method based on U-net end-to-end neural network model | |
| Qin et al. | A novel temporal generative adversarial network for electrocardiography anomaly detection | |
| Roy et al. | ECG-NET: A deep LSTM autoencoder for detecting anomalous ECG | |
| CN111985340A (en) | Face recognition method and device based on neural network model and computer equipment | |
| Magesh et al. | Ensemble feature extraction-based prediction of fetal arrhythmia using cardiotocographic signals | |
| CN116864140A (en) | Intracardiac branch of academic or vocational study postoperative care monitoring data processing method and system thereof | |
| Bitarafan et al. | A hybrid deep model for automatic arrhythmia classification based on LSTM recurrent networks | |
| Likitha et al. | Heart Disease Detection using Machine Learning Technique | |
| CN113080847B (en) | Device for diagnosing mild cognitive impairment based on bidirectional long-short term memory model of graph | |
| EP4198997A1 (en) | A computer implemented method, a method and a system | |
| CN117643461B (en) | Heart rate intelligent monitoring system and method based on artificial intelligence | |
| Yang et al. | Superimposed semantic communication for iot-based real-time ecg monitoring | |
| Mamprin et al. | Identification of patients at risk of cardiac conduction diseases requiring a permanent pacemaker following TAVI procedure: a deep-learning approach on ECG signals | |
| KR102592762B1 (en) | Arrhythmia classification method using densenet based on convolution neural network | |
| CN114420292A (en) | Medical health dynamic prediction method, system and equipment | |
| Kale et al. | An Overview of Cardiac Disease Diagnosis using Machine Learning Algorithms |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |