Method and system for detecting human body shaking of defibrillator fusing electrocardio and thoracic impedance

Abstract

The present application relates to the field of signal processing, in order to solve the problem of high false alarm rate of human body shaking under the state of patient movement, low detection accuracy and poor reliability of human body shaking, a defibrillator human body shaking detection method and system are provided, which fuses electrocardiogram and thoracic impedance. The defibrillator human body shaking detection method fusing electrocardiogram and thoracic impedance includes calculating the probability of human body shaking judged by impedance, then combining the probability of electrocardiogram noise to obtain the time-frequency analysis human body shaking probability; through a deep learning model, the human body shaking probability output by the deep learning model is obtained; the confidence of the deep learning model is calculated according to the human body shaking probability output by the deep learning model, the human body shaking probability calculation strategy is adaptively matched to obtain the final human body shaking probability; the final human body shaking probability is compared with the preset judgment threshold to determine the human body shaking state. It can reduce the false alarm rate of human body shaking under the state of patient movement, improve the detection accuracy and reliability of human body shaking.

Description

A method and system for detecting body sway in defibrillators that integrates electrocardiogram and transthoracic impedance. Technical Field

[0001] This invention relates to the field of signal processing, and more particularly to a defibrillator human body sway detection method and system that integrates electrocardiogram and transthoracic impedance. Background Technology

[0002] The statements in this section are merely background information related to the present invention and do not necessarily constitute prior art.

[0003] Automated external defibrillators (AEDs) are crucial devices for rescuing cardiac arrest. However, in mobile environments, the movement of rescuers or patients can degrade signal quality, affecting the accuracy of heart rhythm analysis. Existing methods for detecting motion artifacts in defibrillators generally rely on single-mode ECG signal analysis. However, this single-mode detection faces a trade-off between sensitivity and specificity in defibrillable heart rhythm analysis. Because traditional methods judge motion artifacts based solely on amplitude changes or frequency characteristics of ECG signals, the false alarm rate may be high in complex environments, potentially leading to defibrillation delays and failing to meet the critical time requirement for cardiac arrest resuscitation. Furthermore, single-signal detection also carries a high risk of missed detections, potentially causing false defibrillation and endangering the patient's safety.

[0004] Therefore, existing defibrillator motion detection systems generally lack the ability to fuse and analyze multi-source physiological signals, resulting in an inability to comprehensively assess the patient's motion state. This leads to a high false alarm rate for body sway during patient motion, and low accuracy and reliability in body sway detection. Summary of the Invention

[0005] To address the aforementioned technical problems, this invention provides a defibrillator human motion detection method and system that integrates electrocardiogram and transthoracic impedance, which can reduce the false alarm rate of motion and improve the accuracy and reliability of human motion detection.

[0006] To achieve the above objectives, the present invention adopts the following technical solution:

[0007] The first aspect of the present invention provides a defibrillator human body sway detection method that integrates electrocardiogram and transthoracic impedance.

[0008] In one or more embodiments, a defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance is provided, comprising:

[0009] Simultaneously acquire raw electrocardiogram signals and transthoracic impedance signals and perform preprocessing;

[0010] Autocorrelation analysis was performed on the preprocessed electrocardiogram (ECG) signal to calculate the probability of ECG noise; variance calculation and peak detection were performed on the preprocessed transthoracic impedance signal to obtain the probability of human body swaying based on impedance judgment; and then, combined with the probability of ECG noise, the probability of human body swaying based on time-frequency analysis was obtained.

[0011] The ECG features and impedance features of the preprocessed ECG signal and transthoracic impedance signal are extracted respectively, and the features are fused. Then, the deep learning model is used to obtain the probability of human body sway output by the deep learning model.

[0012] The confidence level of the deep learning model is calculated based on the probability of human body swaying output by the deep learning model. Then, the human body swaying probability calculation strategy is adaptively matched to obtain the final human body swaying probability. The human body swaying probability calculation strategy includes time-frequency analysis strategy, deep learning model strategy, and weighted fusion of time-frequency analysis and deep learning model strategy.

[0013] The final probability of human body sway is compared with a preset judgment threshold to determine the state of human body sway, so as to determine whether to stop defibrillation rhythm analysis.

[0014] As one implementation method, when the confidence level of the deep learning model is greater than or equal to the first preset confidence threshold, the human body sway probability calculation strategy selects the deep learning model strategy, that is, the human body sway probability output by the deep learning model is used as the final human body sway probability.

[0015] As one implementation method, when the confidence level of the deep learning model is less than the second preset confidence threshold, the human body sway probability calculation strategy selects the time-frequency analysis strategy, that is, the human body sway probability of time-frequency analysis is used as the final human body sway probability; wherein, the second preset confidence threshold is less than the first preset confidence threshold.

[0016] As one implementation method, when the confidence level of the deep learning model is greater than or equal to the second preset confidence threshold and less than the first preset confidence threshold, the human body sway probability calculation strategy selects a weighted fusion of time-frequency analysis and deep learning model strategy, that is, the weighted fusion of the human body sway probability from time-frequency analysis and the human body sway probability output by the deep learning model is used to calculate the final human body sway probability.

[0017] As one implementation method, the formula for calculating the confidence score of a deep learning model is:

[0018] ;

[0019] in, The confidence level of the deep learning model; The probability of human body swaying output by the deep learning model; and It is a constant coefficient.

[0020] As one implementation method, the formula for calculating the probability of human body swaying using time-frequency analysis is:

[0021] ;

[0022] in, The probability of human body swaying in time-frequency analysis; The probability of ECG noise; The probability of human body swaying is determined by impedance.

[0023] As one implementation method, when the variance of the preprocessed transthoracic impedance signal is greater than a preset variance threshold, and the number of peak values ​​of the processed transthoracic impedance signal is greater than a preset peak value threshold, the probability of human body swaying determined by impedance is 1; otherwise, it is 0.

[0024] A second aspect of the present invention provides a defibrillator human body sway detection system that integrates electrocardiogram and transthoracic impedance.

[0025] In one or more embodiments, a defibrillator body sway detection system integrating electrocardiogram and transthoracic impedance includes:

[0026] The signal preprocessing module is used to simultaneously acquire and preprocess the raw electrocardiogram signal and transthoracic impedance signal.

[0027] The signal time-frequency analysis module is used to perform autocorrelation analysis on the preprocessed electrocardiogram signal to calculate the probability of electrocardiogram noise; to perform variance calculation and peak detection on the preprocessed transthoracic impedance signal to obtain the probability of human body swaying based on impedance judgment; and then, combined with the probability of electrocardiogram noise, to obtain the probability of human body swaying based on time-frequency analysis.

[0028] The feature deep learning module is used to extract the ECG features and impedance features of the preprocessed ECG signal and transthoracic impedance signal respectively, perform feature fusion, and then pass them through the deep learning model to obtain the probability of human body sway output by the deep learning model.

[0029] The calculation strategy matching module is used to calculate the confidence of the deep learning model based on the human body sway probability output by the deep learning model, and then adaptively match the human body sway probability calculation strategy to obtain the final human body sway probability; the human body sway probability calculation strategies include time-frequency analysis strategy, deep learning model strategy, and weighted fusion of time-frequency analysis and deep learning model strategy.

[0030] The swaying state determination module is used to compare the final probability of human body swaying with a preset judgment threshold to determine the human body swaying state, so as to determine whether to stop defibrillation rhythm analysis.

[0031] A third aspect of the present invention provides a computer-readable storage medium.

[0032] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps in the defibrillator body sway detection method fused with electrocardiogram and transthoracic impedance as described above.

[0033] A fourth aspect of the present invention provides an electronic device.

[0034] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the steps in the defibrillator body sway detection method fused with electrocardiogram and transthoracic impedance as described above.

[0035] Compared with the prior art, the beneficial effects of the present invention are:

[0036] This invention integrates dual-mode signals from electrocardiogram (ECG) and transthoracic impedance, effectively overcoming the limitations of single signal sources being susceptible to interference. It calculates the confidence level of the deep learning model based on the human body sway probability output by the deep learning model, employing an adaptive matching strategy for calculating the human body sway probability. This strategy includes a time-frequency analysis strategy, a deep learning model strategy, and a weighted fusion of time-frequency analysis and deep learning model strategies to obtain the final human body sway probability. It leverages the high-precision recognition capability of the deep learning model under high confidence levels while relying on stable and reliable traditional features when the model is uncertain, ensuring optimal decision-making across various signal quality conditions and significantly improving the accuracy and reliability of sway detection in complex clinical environments. Attached Figure Description

[0037] The accompanying drawings, which form part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments of the invention and their descriptions are used to explain the invention and do not constitute an improper limitation of the invention.

[0038] Figure 1 is a flowchart of the defibrillator human body sway detection method integrating electrocardiogram and transthoracic impedance according to an embodiment of the present invention.

[0039] Figure 2 is a schematic diagram of the defibrillator human body sway detection system that integrates electrocardiogram and transthoracic impedance according to an embodiment of the present invention.

[0040] Figure 3 is a schematic diagram of an electronic device according to an embodiment of the present invention. Detailed Implementation

[0041] The present invention will be further described below with reference to the accompanying drawings and embodiments.

[0042] It should be noted that the following detailed description is illustrative and intended to provide further explanation of the invention. Unless otherwise specified, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention pertains.

[0043] It should be noted that the terminology used herein is for the purpose of describing particular embodiments only and is not intended to limit the scope of exemplary embodiments according to the invention. As used herein, the singular form is intended to include the plural form as well, unless the context clearly indicates otherwise. Furthermore, it should be understood that when the terms "comprising" and / or "including" are used in this specification, they indicate the presence of features, steps, operations, devices, components, and / or combinations thereof.

[0044] Figure 1 shows a schematic diagram of the defibrillator body sway detection method integrating ECG and transthoracic impedance according to an embodiment of the present invention. According to Figure 1, the defibrillator body sway detection method integrating ECG and transthoracic impedance according to this embodiment may include the following steps S101 to S105.

[0045] The specific implementation process of steps S101 to S105 is as follows:

[0046] Step S101: Simultaneously acquire the raw electrocardiogram signal and transthoracic impedance signal and perform preprocessing.

[0047] Simultaneous acquisition of electrocardiogram (ECG) and transthoracic impedance signals yields raw ECG and transthoracic impedance signals, i.e., raw bimodal physiological data. Specifically, the simultaneous acquisition utilizes an ECG acquisition module and an impedance measurement module to synchronously acquire the corresponding signals. The ECG acquisition module acquires standard lead II ECG signals at a sampling rate of 500Hz, a single session duration of 2 seconds, corresponding to a signal length of 1000 sampling points. The impedance measurement module uses a two-electrode method to measure transthoracic impedance at a sampling rate of 500Hz, a single session duration of 2 seconds, corresponding to a signal length of 1000 sampling points. The simultaneous acquisition mechanism is implemented by a unified clock and data frame structure within the chip, ensuring strict synchronization of the ECG and impedance signals at the hardware level, providing a time-aligned data foundation for subsequent bimodal fusion analysis.

[0048] During preprocessing, for example, a 0.5-100Hz bandpass filter is used for the electrocardiogram signal, and a 1-100Hz bandpass filter is used for the impedance signal. The bandpass filtering is implemented using a third-order Butterworth filter.

[0049] It should be noted that other filtering methods may be used for signal processing in other embodiments, which will not be described in detail here.

[0050] Step S102: Perform autocorrelation analysis on the preprocessed ECG signal to calculate the probability of ECG noise; perform variance calculation and peak detection on the preprocessed transthoracic impedance signal to obtain the probability of human body swaying based on impedance judgment, and then combine it with the probability of ECG noise to obtain the probability of human body swaying based on time-frequency analysis.

[0051] In step S102, autocorrelation analysis is performed on the preprocessed ECG signal. The formula for calculating the probability of ECG noise is as follows:

[0052] ;

[0053] The periodicity quality of a signal is evaluated based on the autocorrelation function, which is calculated using the following formula:

[0054] ;

[0055] The formula for calculating the autocorrelation value with zero delay is:

[0056] ;

[0057] Among them, ACF peak intensity This is the ratio of the maximum peak value of the autocorrelation function within 0.5 seconds to the value at zero delay. The number of delay points; Maximum delay points; based on peak intensity Calculate the probability of ECG noise. .

[0058] The variance calculation formula is as follows during the process of variance calculation and peak detection of the preprocessed transthoracic impedance signal:

[0059] ;

[0060] in, For the i-th impedance sample value, Here, N represents the average impedance, and N is the number of sampling points.

[0061] The peak count K is detected using a sliding window (window size M = 50). First, the counter parameters are initialized. and Secondly, for each window Calculate window mean ,if ,but ,otherwise ;

[0062] If the counter parameter and ,but And reset the counter.

[0063] When the variance of the preprocessed transthoracic impedance signal is greater than a preset variance threshold (e.g., the preset variance threshold is 2, which can be set by those skilled in the art according to actual conditions), and the number of peak values ​​of the processed transthoracic impedance signal is greater than a preset peak value threshold (e.g., the preset peak value threshold is 10, which can be set by those skilled in the art according to actual conditions), the probability of human body swaying determined by impedance is 1 (i.e., ), otherwise 0 ( ).

[0064] The formula for calculating the probability of human body swaying in time-frequency analysis is:

[0065] ;

[0066] in, The probability of human body swaying in time-frequency analysis; The probability of ECG noise; The probability of human body swaying is determined by impedance.

[0067] Step S103: Extract the ECG features and impedance features of the preprocessed ECG signal and transthoracic impedance signal respectively, perform feature fusion, and then pass them through a deep learning model to obtain the probability of human body sway output by the deep learning model.

[0068] In some optional embodiments, a CNN network is used to extract ECG features from the ECG signal. The CNN network comprises three convolutional layers and an adaptive pooling layer, with the input ECG signal (length N = 1000). The three convolutional layers are: Layer 1: kernel size 3, padding 1, output channels 16, followed by batch normalization, ReLU activation, and max pooling (e.g., pooling size 2); Layer 2: kernel size 3, padding 1, output channels 32, followed by batch normalization, ReLU activation, and max pooling (pooling size 2); Layer 3: kernel size 3, padding 1, output channels 64, followed by batch normalization, ReLU activation, and adaptive average pooling; Fully connected layer: after flattening the pooling output, a 128-dimensional ECG depth feature vector is output as the ECG feature.

[0069] An LSTM network is used to extract impedance features from the impedance signal. The LSTM network is a bidirectional two-layer structure. The input impedance signal (data length N = 1000) is used, and the bidirectional two-layer LSTM is used. The hidden layer size is 64, the bidirectional structure (actual hidden state size is 128), the number of layers is 2, and the dropout rate is 0.3. The fully connected layer takes the 128-dimensional impedance depth feature vector output from the last time step of the LSTM as the impedance feature.

[0070] It should be noted that in other embodiments, other existing neural networks may also be used to extract the corresponding features of the preprocessed electrocardiogram signal and transthoracic impedance signal.

[0071] Deep learning models are used to represent the relationship between feature fusion and the probability of human body swaying. Deep learning models can be specifically configured according to the actual situation and their specific structure is not limited.

[0072] Step S104: Calculate the confidence level of the deep learning model based on the human body sway probability output by the deep learning model, and then adaptively match the human body sway probability calculation strategy to obtain the final human body sway probability; the human body sway probability calculation strategy includes time-frequency analysis strategy, deep learning model strategy, and weighted fusion of time-frequency analysis and deep learning model strategy.

[0073] In this embodiment, the formula for calculating the confidence score of the deep learning model is:

[0074] The formula for calculating the confidence score of a deep learning model is:

[0075] ;

[0076] in, The confidence level of the deep learning model; The probability of human body swaying output by the deep learning model; and For example, constant coefficients , , and The specific settings depend on the different types of deep learning models.

[0077] The confidence level of the deep learning model; This refers to the probability of human body swaying output by the deep learning model.

[0078] When the confidence level of the deep learning model is greater than or equal to the first preset confidence threshold (e.g., 0.8), the human body sway probability calculation strategy selects the deep learning model strategy, that is, the human body sway probability output by the deep learning model is used as the final human body sway probability. .

[0079] When the confidence level of the deep learning model is less than the second preset confidence threshold (e.g., 0.5), the time-frequency analysis strategy is selected for calculating the probability of human body swaying, that is, the probability of human body swaying obtained from the time-frequency analysis is used as the final probability of human body swaying. The second preset confidence threshold is less than the first preset confidence threshold.

[0080] When the confidence level of the deep learning model is greater than or equal to the second preset confidence threshold and less than the first preset confidence threshold, the human body sway probability calculation strategy selects the weighted fusion of time-frequency analysis and deep learning model strategy, that is, the weighted fusion of the human body sway probability from time-frequency analysis and the human body sway probability output by the deep learning model to calculate the final human body sway probability.

[0081]

[0082] in, This represents the probability of human body swaying in time-frequency analysis. and Let be the weighting coefficient, satisfying .

[0083] It should be noted that the values ​​of the second preset reliability threshold and the first preset reliability threshold can be selected by those skilled in the art according to the actual situation, and will not be described in detail here.

[0084] Step S105: Compare the final probability of human body sway with the preset judgment threshold T to determine the state of human body sway, so as to determine whether to stop defibrillation rhythm analysis.

[0085] when When, it is judged as a shaking state; when When a certain condition is met, it is determined to be in a state without shaking, for example, by determining a preset threshold. =0.5; where the preset judgment threshold can be selected by those skilled in the art according to the actual situation, and will not be described in detail here.

[0086] When a person is detected to be shaking, the analysis of defibrillable rhythms of the electrocardiogram signal is stopped and a voice alarm is issued to prompt the patient to stop shaking. The analysis continues to check whether the patient is shaking until the patient is detected to be non-shaking, at which point the normal defibrillable rhythm analysis procedure is performed.

[0087] This invention utilizes the calculated probability of swaying to trigger a safety control strategy. When swaying is detected, the system immediately pauses the defibrillable rhythm analysis and issues a warning. It then continues analyzing the patient's swaying until the swaying stops and the defibrillable rhythm analysis process begins. This fundamentally eliminates the significant safety risk of misjudgment and subsequent misdefibrillation due to patient movement. This safety strategy maximizes patient safety while ensuring the timeliness and continuity of the resuscitation process in a stable state.

[0088] As shown in Figure 2, the defibrillator human body sway detection system that integrates ECG and transthoracic impedance provided in this embodiment of the invention can be implemented in software. The defibrillator human body sway detection system that integrates ECG and transthoracic impedance includes the following software modules: signal preprocessing module 201, signal time-frequency analysis module 202, feature deep learning module 203, calculation strategy matching module 204, and sway state determination module 205.

[0089] The following is an introduction to the functions of each software module in the defibrillator body sway detection system that integrates electrocardiogram and transthoracic impedance:

[0090] The signal preprocessing module 201 is used to simultaneously acquire the raw electrocardiogram signal and transthoracic impedance signal and perform preprocessing.

[0091] The signal time-frequency analysis module 202 is used to perform autocorrelation analysis on the preprocessed electrocardiogram signal and calculate the probability of electrocardiogram noise; to perform variance calculation and peak detection on the preprocessed transthoracic impedance signal to obtain the probability of human body swaying based on impedance judgment, and then combine it with the probability of electrocardiogram noise to obtain the probability of human body swaying based on time-frequency analysis.

[0092] The feature deep learning module 203 is used to extract the electrocardiogram features and impedance features of the preprocessed electrocardiogram signal and transthoracic impedance signal respectively, perform feature fusion, and then pass them through the deep learning model to obtain the probability of human body sway output by the deep learning model.

[0093] The calculation strategy matching module 204 is used to calculate the confidence of the deep learning model based on the human body sway probability output by the deep learning model, and then adaptively match the human body sway probability calculation strategy to obtain the final human body sway probability; the human body sway probability calculation strategy includes time-frequency analysis strategy, deep learning model strategy, and weighted fusion of time-frequency analysis and deep learning model strategy.

[0094] The swaying state determination module 205 is used to compare the final human body swaying probability with a preset judgment threshold to determine the human body swaying state, so as to determine whether to stop defibrillation rhythm analysis.

[0095] It should be noted that each module in the defibrillator body sway detection system that integrates ECG and transthoracic impedance in this embodiment of the invention corresponds one-to-one with each step in the defibrillator body sway detection method that integrates ECG and transthoracic impedance in the above embodiment, and their specific implementation processes are the same, so they will not be repeated here.

[0096] The structure of the electronic device according to the embodiments of the present invention will be described in detail below. Figure 3 is a schematic diagram of the composition structure of the electronic device provided in the embodiments of the present invention. It can be understood that Figure 3 only shows an exemplary structure of the electronic device and not all of the structures. Some or all of the structures shown can be implemented as needed.

[0097] The electronic device provided in this embodiment of the invention includes at least one processor 301, a memory 302, a user interface 303, and at least one network interface 304. The various components of the defibrillator body sway detection system integrating ECG and transthoracic impedance are coupled together via a bus system 305. It is understood that the bus system 305 is used to realize communication between these components. In addition to a data bus, the bus system 305 also includes a power bus, a control bus, and a status signal bus. However, for clarity, all buses are labeled as bus system 305 in Figure 3.

[0098] The user interface 303 may include a monitor, keyboard, mouse, trackball, click wheel, buttons, touchpad, or touch screen.

[0099] It is understood that memory 302 can be volatile memory or non-volatile memory, or both. In this embodiment of the invention, memory 302 is capable of storing data to support the operation of the terminal. Examples of this data include any computer programs used to operate on the terminal, such as operating systems and applications. The operating system includes various system programs, such as framework layers, core library layers, driver layers, etc., used to implement various basic services and handle hardware-based tasks. Applications can include various applications.

[0100] In some embodiments, the defibrillator body sway detection system integrating ECG and transthoracic impedance provided in this invention can be implemented using a combination of hardware and software. As an example, the defibrillator body sway detection system integrating ECG and transthoracic impedance provided in this invention can be a processor in the form of a hardware decoding processor, programmed to execute the defibrillator body sway detection method integrating ECG and transthoracic impedance provided in this invention. For example, the processor in the form of a hardware decoding processor can employ one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components.

[0101] As an example, processor 301 can be an integrated circuit chip with signal processing capabilities, such as a general-purpose processor, a digital signal processor (DSP), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc., wherein the general-purpose processor can be a microprocessor or any conventional processor, etc.

[0102] As an example of the hardware implementation of the defibrillator body sway detection system integrating ECG and transthoracic impedance provided in this embodiment of the invention, the device provided in this embodiment of the invention can be directly executed by a processor 301 in the form of a hardware decoding processor. For example, it can be executed by one or more application-specific integrated circuits (ASICs), DSPs, programmable logic devices (PLDs), complex programmable logic devices (CPLDs), field-programmable gate arrays (FPGAs), or other electronic components to implement the defibrillator body sway detection method integrating ECG and transthoracic impedance provided in this embodiment of the invention.

[0103] In this embodiment of the invention, memory 302 is used to store various types of data to support the operation of the defibrillator body sway detection system that integrates ECG and transthoracic impedance, or to store program code for executing the method shown in FIG1. ​​Examples of such data include: any executable instructions for operating on the defibrillator body sway detection system that integrates ECG and transthoracic impedance, such as executable instructions, whereby the program implementing the defibrillator body sway detection method that integrates ECG and transthoracic impedance of this embodiment of the invention may be included in the executable instructions.

[0104] Specifically, according to embodiments of this application, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, embodiments of this application include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in FIG1. ​​In such embodiments, the computer program can be downloaded and installed from a network via a communication component, and / or installed from a removable medium. When the computer program is executed by a central processing unit, it performs various functions defined in the apparatus of this application.

[0105] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, as well as combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in one or more flowchart illustrations and / or one or more block diagrams.

[0106] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Various modifications and variations can be made to the present invention by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A method for detecting body sway in a defibrillator by integrating electrocardiogram and transthoracic impedance, characterized in that, include: Simultaneously acquire raw electrocardiogram signals and transthoracic impedance signals and perform preprocessing; Autocorrelation analysis was performed on the preprocessed ECG signal to calculate the probability of ECG noise; Variance calculation and peak detection are performed on the preprocessed transthoracic impedance signal to obtain the probability of human body swaying based on impedance judgment. Then, combined with the probability of electrocardiogram noise, the probability of human body swaying based on time-frequency analysis is obtained. The ECG and impedance features of the preprocessed ECG and transthoracic impedance signals are extracted and fused, and then processed by a deep learning model to obtain the human body sway probability output by the deep learning model. The confidence level of the deep learning model is calculated based on the human body sway probability output by the deep learning model, and then an adaptive human body sway probability calculation strategy is obtained to obtain the final human body sway probability. The human body sway probability calculation strategy includes a time-frequency analysis strategy, a deep learning model strategy, and a weighted fusion of time-frequency analysis and deep learning model strategy. The final human body sway probability is compared with a preset judgment threshold to determine the human body sway state and to determine whether to stop defibrillation rhythm analysis. Among them, when the confidence level of the deep learning model is greater than or equal to the second preset confidence threshold and less than the first preset confidence threshold, the human body sway probability calculation strategy selects the weighted fusion of time-frequency analysis and deep learning model strategy, that is, the weighted fusion of the human body sway probability from time-frequency analysis and the human body sway probability output by the deep learning model is used to calculate the final human body sway probability. ； The probability of human body swaying in time-frequency analysis; The probability of human body swaying output by the deep learning model; and Let be the weighting coefficient, satisfying 。 2. The defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance as described in claim 1, characterized in that, When the confidence level of the deep learning model is greater than or equal to the first preset confidence threshold, the human body sway probability calculation strategy selects the deep learning model strategy, that is, the human body sway probability output by the deep learning model is used as the final human body sway probability.

3. The defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance as described in claim 2, characterized in that, When the confidence level of the deep learning model is less than the second preset confidence threshold, the human body sway probability calculation strategy selects the time-frequency analysis strategy, that is, the human body sway probability of time-frequency analysis is used as the final human body sway probability; wherein, the second preset confidence threshold is less than the first preset confidence threshold.

4. The defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance as described in claim 1, characterized in that, The formula for calculating the confidence score of a deep learning model is: ;in, The confidence level of the deep learning model; The probability of human body swaying output by the deep learning model; and It is a constant coefficient.

5. The defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance as described in claim 1, characterized in that, The formula for calculating the probability of human body swaying in time-frequency analysis is: ;in, The probability of human body swaying in time-frequency analysis; The probability of ECG noise; The probability of human body swaying is determined by impedance.

6. The defibrillator body sway detection method integrating electrocardiogram and transthoracic impedance as described in claim 1, characterized in that, When the variance of the preprocessed transthoracic impedance signal is greater than the preset variance threshold, and the number of peak values ​​of the processed transthoracic impedance signal is greater than the preset peak value threshold, the probability of human body swaying determined by impedance is 1; otherwise, it is 0.

7. A defibrillator body sway detection system integrating electrocardiogram and transthoracic impedance, characterized in that, The defibrillator human sway detection method based on the fusion of ECG and transthoracic impedance as described in any one of claims 1-6 includes: a signal preprocessing module, which is used to simultaneously acquire and preprocess the original ECG signal and transthoracic impedance signal; a signal time-frequency analysis module, which is used to perform autocorrelation analysis on the preprocessed ECG signal to calculate the probability of ECG noise; to perform variance calculation and peak detection on the preprocessed transthoracic impedance signal to obtain the probability of human sway based on impedance judgment, and then combine it with the probability of ECG noise to obtain the probability of human sway based on time-frequency analysis; and a feature deep learning module, which is used to extract the ECG signals of the preprocessed ECG signal and the transthoracic impedance signal respectively. The system analyzes and fuses human body sway characteristics and impedance characteristics, then processes them through a deep learning model to obtain the probability of human body sway output by the deep learning model. A strategy matching module calculates the confidence level of the deep learning model based on the probability of human body sway output by the deep learning model, and then adaptively matches the human body sway probability calculation strategy to obtain the final probability of human body sway. The human body sway probability calculation strategies include time-frequency analysis strategy, deep learning model strategy, and weighted fusion of time-frequency analysis and deep learning model strategy. A sway state determination module compares the final probability of human body sway with a preset judgment threshold to determine the human body sway state, thereby determining whether to stop defibrillation rhythm analysis.

8. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps in the defibrillator human body sway detection method that integrates electrocardiogram and transthoracic impedance as described in any one of claims 1-6.

9. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the steps in the defibrillator human body sway detection method that integrates electrocardiogram and transthoracic impedance as described in any one of claims 1-6.

Images