Fabric electrode-based electrical stimulation interference filtering method, device, terminal and medium

Abstract

This invention discloses a method, device, terminal, and medium for filtering electrical stimulation interference based on fabric electrodes. The method includes the following steps: acquiring a first electromyography (EMG) signal; identifying a second EMG signal after selective blanking processing based on the first EMG signal; performing CEEMDAN decomposition on the second EMG signal to obtain multiple intrinsic pattern functions (IPFs), and identifying a reconstructed IPF associated with the second EMG signal among the multiple IPFs; and reconstructing the signal based on the reconstructed IPFs to obtain a filtered signal after artifact filtering. Therefore, this invention can effectively eliminate motion artifacts, interference noise, and functional electrical stimulation artifacts in EMG signals acquired based on fabric electrodes by combining selective blanking processing and signal reconstruction methods.

Description

Technical Field

[0001] This invention relates to the field of signal processing technology, and in particular to a method, apparatus, terminal device, and computer storage medium for filtering electrical stimulation interference based on fabric electrodes. Background Technology

[0002] The electrodes of FES (Functional Electrical Stimulation) devices serve as the human-machine interface for electromyography (EMG) signal acquisition and FES electrical stimulation output, playing a crucial role in the entire system. Compared with traditional gel electrodes, fabric electrodes have better breathability, flexibility, and foldability, making them ideal for long-term monitoring of EMG signals. However, fabric electrodes do not have close contact with the human body, and during use, they are prone to horizontal relative friction and vertical relative compression with the skin. This results in the acquired EMG signals being accompanied by significant motion artifacts and interference noise. Furthermore, when EMG signals are interfered with by functional electrical stimulation pulses, the stimulation pulses simultaneously activate motor units, and multiple similar MUAPs (motor unit action potentials) superimpose on the skin surface at the same time to form stimulation-induced muscle response M waves.

[0003] Existing techniques for filtering out stimulation-induced muscle response M-waves, eliminating motion artifacts and interference noise, such as comb filters, adaptive matched filters, autoregressive AR models, wavelet transforms, or heterogeneous decomposition, all require certain prior knowledge. When the amplitude and frequency of functional electrical stimulation change dynamically, artifacts may not be filtered out, or the filtering of artifacts may also eliminate the voluntary electromyographic signal.

[0004] In summary, existing fabric electrode artifact filtering methods cannot effectively reduce fabric electrode motion artifacts and extract weak electromyographic signals of human voluntary intent. Summary of the Invention

[0005] The main objective of this invention is to provide a method, terminal device, and computer storage medium for filtering electrical stimulation interference based on fabric electrodes. The aim is to effectively eliminate motion artifacts, interference noise, and functional electrical stimulation artifacts in the electromyographic signals acquired based on fabric electrodes by combining selective blanking processing and signal reconstruction methods.

[0006] To achieve the above objectives, the present invention provides a method for filtering electrical stimulation interference based on fabric electrodes, the method comprising the following steps:

[0007] Acquire a first electromyographic signal, and confirm the second electromyographic signal after selective blanking based on the first electromyographic signal;

[0008] The second electromyographic signal is subjected to CEEMDAN decomposition to obtain multiple original pattern functions, and among the multiple original pattern functions, the reconstructed original pattern function related to the second electromyographic signal is identified;

[0009] The signal is reconstructed based on the reconstructed proof mode function to obtain the filtered signal after artifact filtering.

[0010] Optionally, the step of confirming the selectively hidden second electromyographic signal based on the first electromyographic signal includes:

[0011] Identify the time period during which the SA wave of the first electromyographic signal appears;

[0012] Based on the time period, selective blanking is performed on the first electromyographic signal to obtain a second electromyographic signal after selective blanking.

[0013] Optionally, the step of identifying the reconstructed evidence pattern function related to the first electromyographic signal among the plurality of evidence pattern functions includes:

[0014] Confirm the correlation coefficients between each of the aforementioned pattern functions and the second electromyographic signal;

[0015] The reconstructed theorem pattern function associated with the second electromyographic signal is confirmed based on the correlation coefficient.

[0016] Optionally, the step of confirming the reconstructed intrinsic pattern function related to the second electromyographic signal based on the correlation coefficient includes:

[0017] Confirm whether the correlation coefficient is greater than the preset correlation coefficient threshold:

[0018] If it is confirmed that the correlation coefficient is greater than the correlation coefficient threshold, then the proof pattern function with a correlation coefficient greater than the correlation coefficient threshold is confirmed as the reconstructed proof pattern function.

[0019] Optionally, the reconstructed proof mode function includes: a residual reconstructed proof mode function, and the step of reconstructing the signal based on the reconstructed proof mode function to obtain the filtered signal after artifact filtering includes:

[0020] In the reconstructed proof mode function, a noisy target reconstructed proof mode function is identified;

[0021] The target reconstructed proof mode function is subjected to noise reduction processing to obtain the noise-reduced target reconstructed proof mode function.

[0022] The target reconstructed proof mode function after noise reduction and the remaining reconstructed proof mode function are used to reconstruct the signal to obtain the filtered signal after artifact filtering.

[0023] Optionally, the step of identifying the noisy target reconstruction proof pattern function in the reconstructed proof pattern function includes:

[0024] Confirm the sample entropy of the reconstructed proof mode function;

[0025] The target reconstruction proof pattern function is confirmed in the reconstruction proof pattern function based on the sample entropy.

[0026] Optionally, the step of confirming the target reconstructed proof pattern function in the reconstructed proof pattern function based on the sample entropy includes:

[0027] Confirm whether the sample quotient is greater than the preset sample entropy threshold;

[0028] If it is confirmed that the sample entropy is greater than the sample entropy threshold, then in the reconstructed proof pattern function, the reconstructed proof pattern function whose sample entropy is greater than the sample entropy threshold is the target reconstructed proof pattern function.

[0029] Furthermore, to achieve the above objectives, the present invention also provides a stimulus artifact interference filtering device, characterized in that the stimulus artifact interference filtering device comprises:

[0030] The acquisition module is used to acquire a first electromyographic signal and to confirm a second electromyographic signal after selective blanking processing based on the first electromyographic signal.

[0031] The confirmation module is used to perform CEEMDAN decomposition on the second electromyographic signal to obtain multiple original pattern functions, and to confirm the reconstructed original pattern function related to the second electromyographic signal among the multiple original pattern functions.

[0032] The reconstruction module is used to reconstruct the signal based on the reconstructed proof mode function to obtain the filtered signal after artifact filtering.

[0033] In addition, to achieve the above objectives, the present invention also provides a terminal device, the terminal device comprising: a memory, a processor, and a fabric electrode-based electrical stimulation interference filtering program stored in the memory and executable on the processor, wherein when the fabric electrode-based electrical stimulation interference filtering program is executed by the processor, it implements the steps of the fabric electrode-based electrical stimulation interference filtering method as described above.

[0034] In addition, to achieve the above objectives, the present invention also provides a computer storage medium storing a fabric electrode-based electrical stimulation interference filtering program, wherein when the fabric electrode-based electrical stimulation interference filtering program is executed by a processor, the steps of the fabric electrode-based electrical stimulation interference filtering method described above are implemented.

[0035] This invention proposes a method, apparatus, terminal device, and computer-readable storage medium for filtering electrical stimulation interference based on fabric electrodes. The method includes the following steps: acquiring a first electromyography (EMG) signal; identifying a second EMG signal after selective blanking processing based on the first EMG signal; performing CEEMDAN decomposition on the second EMG signal to obtain multiple original pattern functions; and identifying a reconstructed original pattern function related to the second EMG signal among the multiple original pattern functions; and reconstructing the signal based on the reconstructed original pattern function to obtain a filtered signal after artifact filtering.

[0036] The technical solution of the present invention is applied to an FES device. A first electromyography (EMG) signal is obtained through the fabric electrodes of the FES device. A second EMG signal after selective blanking is identified based on the first EMG signal. Then, CEEMDAN decomposition is performed on the second EMG signal to obtain multiple original pattern functions. Among the multiple original pattern functions, a reconstructed original pattern function related to the second EMG signal is identified. Finally, signal reconstruction is performed on the reconstructed original pattern function to obtain a filtered signal after artifact filtering.

[0037] Compared to traditional methods, this invention acquires electromyography (EMG) signals from fabric electrodes, firstly performs selective blanking on these signals, and then performs CEEMDAN decomposition and signal reconstruction on the selectively blanked EMG signals to obtain a filtered signal after artifact filtering. Thus, this invention achieves the processing of EMG signals acquired from fabric electrodes by combining selective blanking and signal reconstruction methods, thereby effectively eliminating motion artifacts, interference noise, and functional electrical stimulation artifacts in the fabric electrode EMG signals. Attached Figure Description

[0038] Figure 1 This is a schematic diagram of the hardware operation of the terminal device involved in the embodiment of the present invention;

[0039] Figure 2 This is a schematic flowchart of an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0040] Figure 3 This is a schematic diagram of FES electromyography signals involved in an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0041] Figure 4 This is a schematic diagram of FES electromyographic signal after selective blanking processing, as described in an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0042] Figure 5This is a schematic diagram of the decomposed IMF involved in an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0043] Figure 6 This is a schematic diagram of the IMF component sample entropy involved in an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0044] Figure 7 This is a schematic diagram showing the correlation coefficient between the IMF component and the electromyographic signal in an embodiment of an electrical stimulation interference filtering method based on fabric electrodes according to the present invention.

[0045] Figure 8 This is a schematic diagram of the structural relationship of an electrical stimulation interference filtering system based on fabric electrodes according to the present invention.

[0046] The objectives, features, and advantages of this invention will be further explained in conjunction with the embodiments and with reference to the accompanying drawings. Detailed Implementation

[0047] It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

[0048] like Figure 1 As shown, Figure 1 This is a schematic diagram of the hardware operating environment of the terminal device involved in the embodiment of the present invention.

[0049] It should be noted that, Figure 1 This can be a structural diagram of the hardware operating environment of the terminal device. In this embodiment of the invention, the terminal device can be a device that integrates filtering of electrical stimulation interference based on fabric electrodes. Specifically, the terminal device can be a mobile terminal, a data storage and control terminal, a PC, or a portable computer, etc.

[0050] like Figure 1 As shown, the terminal device may include: a processor 1001, such as a CPU; a network interface 1004; a user interface 1003; a memory 1005; and a communication bus 1002. The communication bus 1002 is used to enable communication between these components. The user interface 1003 may include a display screen and an input unit such as a keyboard. Optionally, the user interface 1003 may also include a standard wired interface or a wireless interface. The network interface 1004 may optionally include a standard wired interface or a wireless interface (such as a Wi-Fi interface). The memory 1005 may be non-volatile memory (such as Flash memory), high-speed RAM, or stable memory (such as disk storage). Optionally, the memory 1005 may also be a storage device independent of the aforementioned processor 1001.

[0051] Those skilled in the art will understand that Figure 1 The terminal device structure shown does not constitute a limitation on the terminal device and may include more or fewer components than shown, or combine certain components, or have different component arrangements.

[0052] like Figure 1 As shown, the memory 1005, which serves as a computer storage medium, may include an operating system, a network communication module, a user interface module, and a fabric electrode-based electrical stimulation interference filtering program. The operating system is a program that manages and controls the hardware and software resources of the sample terminal device, supporting the operation of the fabric electrode-based electrical stimulation interference filtering program and other software or programs.

[0053] exist Figure 1 In the terminal device shown, the user interface 1003 is mainly used for data communication with various terminals; the network interface 1004 is mainly used for connecting to the backend server and communicating with the backend server; and the processor 1001 can be used to call the electrical stimulation interference filtering program based on fabric electrodes in the memory 1005.

[0054] Based on the aforementioned terminal device, various embodiments of the present invention's method for filtering electrical stimulation interference based on fabric electrodes are proposed. These embodiments of the present invention's method for filtering electrical stimulation interference based on fabric electrodes are described below.

[0055] Please refer to Figure 2 , Figure 2 This is a schematic flowchart of the first embodiment of the electrical stimulation interference filtering method based on fabric electrodes of the present invention. In the first embodiment of the electrical stimulation interference filtering method based on fabric electrodes of the present invention, the method includes:

[0056] Step S10: Obtain the first electromyography (EMG) signal, and confirm the second EMG signal after selective blanking based on the first EMG signal;

[0057] In this embodiment, when the terminal device filters out stimulus artifacts, it acquires a first electromyographic signal collected by a fabric electrode, and then confirms a second electromyographic signal after selective blanking based on the first electromyographic signal.

[0058] For example, in this embodiment, when the terminal device filters out stimulation artifacts, it acquires a first electromyographic signal y collected by a fabric electrode, and then, based on the first electromyographic signal y, after confirming selective blanking processing, filters out the second electromyographic signal x of SA (stimulation artifact) and M-wave (compound muscle action potential).

[0059] It should be noted that, in this embodiment, under the interference of the FES (functional electrical stimulation) device, the electromyographic signals include SA waves, M waves, and voluntary EMG (electromyography) signals. The amplitude of the SA wave segment is usually 50-200 times that of the EMG signal. The extracted voluntary EMG signals can be used for human intention recognition, such as estimating the continuous range of motion of joints and identifying discrete limb movement modes. They can also be used to analyze physiological information such as the user's muscle state, such as muscle fatigue.

[0060] Optionally, in some feasible embodiments, the step S10 of "confirming the second electromyographic signal after selective blanking based on the first electromyographic signal" may include the following steps:

[0061] Step S101: Confirm the time period during which the SA wave of the first electromyographic signal appears;

[0062] In this embodiment, when the terminal device filters out stimulus artifacts, it identifies the time period during which the SA wave of the first electromyographic signal appears.

[0063] For example, such as Figure 3 The schematic diagram of FES electromyography signal shown in this embodiment indicates that when the terminal device filters out stimulation artifacts, it identifies the time period during which the SA wave of the first electromyography signal appears, which is the duration of the blanking.

[0064] It should be noted that in this embodiment, the first electromyographic signal x is the electromyographic signal acquired by the FES device, and the second electromyographic signal y is the electromyographic signal after selectively blanking the first electromyographic signal x. The segment amplitude of the SA wave and the stimulation pulse appear simultaneously. The start time of the SA wave is first identified as follows, by calculating the zero-crossing point of its second derivative. Calculate time t, where y is the first electromyographic signal, and then use the first derivative as the threshold. Where thr is the decision threshold, thr = 10·std(dy / dt), where std is the standard deviation, to select the start time of the positive SA mutation. If t satisfies this threshold formula, then t is marked as the start time of the selective blanking method. The blanking duration is defined as from y(t) to y(t+n*pw), where pw is the stimulus pulse width, and n is based on the charge diffusion time window law. Specifically, n can be selected from 5 to 15, at which point the effect is better.

[0065] Step S102: Selectively blanking the first electromyographic signal according to the time period to obtain the blanked second electromyographic signal.

[0066] In this embodiment, after the terminal device confirms the time period of the SA wave occurrence of the first electromyographic signal, it performs selective blanking processing on the first electromyographic signal according to the time period of the SA wave occurrence to obtain the blanked second electromyographic signal.

[0067] It should be noted that, in this embodiment, as Figure 4 The diagram shows the FES electromyography signal after selective blanking. The terminal device acquires the first electromyography signal y from the fabric electrode, confirms the blanking duration of the first electromyography signal y, and then sets all data within the time period from y(t) to y(t+n*pw) to zero to replace the peak of the SA segment of the electrical stimulation artifact. That is, selective blanking is performed on the data within the time period from y(t) to y(t+n*pw) to obtain the blanked second electromyography signal x. It should be understood that in the blanked second electromyography signal x, the SA wave and most of the M wave are completely blocked. The second electromyography signal x contains some residual M wave and voluntary electromyography signal.

[0068] Step S20: Perform CEEMDAN decomposition on the second electromyography signal to obtain multiple original pattern functions, and identify the reconstructed original pattern function related to the second electromyography signal among the multiple original pattern functions;

[0069] In this embodiment, the terminal device acquires a first electromyography (EMG) signal collected by a fabric electrode, then confirms a second EMG signal after selective blanking based on the first EMG signal, performs CEEMDAN decomposition on the second EMG signal to obtain multiple procedural pattern functions, and identifies a reconstructed procedural pattern function related to the second EMG signal among the multiple procedural pattern functions.

[0070] For example, in this embodiment, the terminal device acquires a first electromyographic signal y collected by a fabric electrode, then confirms a second electromyographic signal x after selective blanking processing based on the first electromyographic signal y, performs CEEMDAN decomposition on the second electromyographic signal x to obtain multiple intrinsic mode functions (IMFs), and identifies the reconstructed IMF that is highly correlated with the second electromyographic signal x and requires signal reconstruction among the multiple IMFs.

[0071] It should be noted that in this embodiment, adaptive noise-complete ensemble empirical mode decomposition (CEEMDAN) is performed on the second electromyography signal x. CEEMDAN is an adaptive noise-assisted EMD technique that improves upon EMD. It borrows the idea of ​​adding Gaussian noise and canceling noise through multiple superpositions and averaging from the EEMD method, thus restoring the completeness of EMD. The CEEMDAN method is used to decompose the signal x after selective blanking, and the blanked FES-EMG signal is defined as follows: t = 1, 2, ..., M, where Let n(t) be the ideal noise-free signal, n(t) be the noise following the N(0,1) distribution, and σ be the variance of the noise.

[0072] For example, the specific decomposition steps are as follows: adding Gaussian white noise n of different amplitudes to the signal x(t) k (t), to obtain several signals x k (t)=x(t)+σ k n k (t), and then use the EMD method to analyze x. k (t) is decomposed to obtain the first IMF, i.e., the first-order IMF, and then its average value is calculated:

[0073]

[0074] Then according to Calculate the first residual component: Let Ej(·) be the j-th IMF component after EMD decomposition of the signal. Then the second IMF component, i.e., the second-order IMF component, is:

[0075]

[0076] And so on, the L remaining components are calculated as follows: The (L+1)th IMF is:

[0077]

[0078] The decomposition process continues until the remaining components become monotonic functions, meaning the EMD decomposition conditions are no longer met. At this point, the decomposition process ends, and the FES-EMG signal is finally represented as:

[0079]

[0080] For example, such as Figure 5The diagram shows the decomposed IMF components. For example, the first electromyographic signal x is decomposed into 14 IMF components: IMF1 (first decomposition of the first-order IMF component), IMF2 (second decomposition of the second-order IMF component), IMF3 (third decomposition of the third-order IMF component), IMF4 (fourth decomposition of the fourth-order IMF component), IMF5 (fifth decomposition of the fifth-order IMF component), IMF6 (sixth decomposition of the sixth-order IMF component), IMF7 (seventh decomposition of the seventh-order IMF component), IMF8 (eighth decomposition of the eighth-order IMF component), IMF9 (ninth decomposition of the ninth-order IMF component), IMF10 (tenth decomposition of the tenth-order IMF component), IMF11 (eleventh decomposition of the eleventh-order IMF component), IMF12 (twelfth decomposition of the twelfth-order IMF component), IMF13 (thirteenth decomposition of the thirteenth-order IMF component), and IMF14 (fourteenth decomposition of the fourteenth-order IMF component).

[0081] Optionally, in some feasible embodiments, the step of "identifying the reconstructed evidence pattern function associated with the first electromyographic signal among the plurality of evidence pattern functions" in step S20 may include the following steps:

[0082] Step S201: Confirm the correlation coefficient between each of the plurality of the aforementioned pattern functions and the second electromyographic signal;

[0083] In this embodiment, the terminal device acquires a first electromyography (EMG) signal collected by fabric electrodes, then confirms a second EMG signal after selective blanking processing based on the first EMG signal, and then confirms the correlation coefficients between multiple empirical mode functions and the second EMG signal.

[0084] For example, such as Figure 7 The diagram showing the correlation coefficients between the IMF components and the electromyographic signal illustrates that, in this embodiment, after the terminal device performs CEEMDAN decomposition on the second electromyographic signal, it uses a signal correlation calculation method to calculate the correlation between each order of IMF and the second electromyographic signal, that is, to confirm the correlation coefficients between each order of IMF and the second electromyographic signal.

[0085] Specifically, the formula for obtaining the correlation coefficients between each order of the intrinsic mode function and the second electromyographic signal is as follows:

[0086]

[0087] Where Cov(X,Y) is the covariance of X and Y, Var[X] is the variance of X, Var[Y] is the variance of Y, r represents the correlation coefficient, X is the intrinsic mode function of each order, and Y is the second electromyographic signal.

[0088] For example, in this embodiment, the correlation coefficient between IMF1 and the second electromyographic signal x is 0.0093, the correlation coefficient between IMF2 and the second electromyographic signal x is 0.0016, the correlation coefficient between IMF3 and the second electromyographic signal x is 0.0344, the correlation coefficient between IMF4 and the second electromyographic signal x is 0.1674, the correlation coefficient between IMF5 and the second electromyographic signal x is 0.3852, the correlation coefficient between IMF6 and the second electromyographic signal x is 0.5036, and the correlation coefficient between IMF7 and the second electromyographic signal x is 0.49. 48. The correlation coefficients between IMF8 and the second electromyographic signal x were 0.5238, 0.5148, 0.4471, 0.4184, 0.2560, -0.0189, and -0.0083, respectively.

[0089] Step S202: Confirm the reconstructed thematic mode function associated with the second electromyographic signal based on the correlation coefficient.

[0090] In this embodiment, after the terminal device confirms the correlation coefficients between each of the multiple proof pattern functions and the second electromyographic signal, it further confirms the reconstructed proof pattern function related to the second electromyographic signal based on the correlation coefficients.

[0091] Optionally, in some feasible embodiments, step S202 may include the following steps:

[0092] Step S2021: Confirm whether the correlation coefficient is greater than the preset correlation coefficient threshold:

[0093] In this embodiment, after the terminal device confirms the correlation coefficients between each of the multiple proof mode functions and the second electromyographic signal, it further confirms whether the correlation coefficients are greater than a preset correlation coefficient threshold.

[0094] For example, in this embodiment, after the terminal device confirms the correlation coefficients of each of the multiple proof mode functions of each order with the second electromyographic signal x, if the preset correlation coefficient threshold is confirmed to be 0.3, it further confirms whether the correlation coefficient is greater than 0.3.

[0095] Step S2022: If it is confirmed that the correlation coefficient is greater than the correlation coefficient threshold, then the proof pattern function with a correlation coefficient greater than the correlation coefficient threshold is confirmed as the reconstructed proof pattern function.

[0096] In this embodiment, after the terminal device confirms whether the correlation coefficient is greater than the preset correlation coefficient threshold, if it confirms that the correlation coefficient is greater than the correlation coefficient threshold, then the proof mode function with the correlation coefficient greater than the correlation coefficient threshold is confirmed as the reconstructed proof mode function.

[0097] For example, in this embodiment, after the terminal device confirms whether the correlation coefficient is greater than the preset correlation coefficient threshold of 0.3, if it confirms that the correlation coefficients between IMF5 and the second electromyographic signal x are 0.3852, 0.5036, 0.4948, 0.5238, 0.5148, 0.4471, and 0.4184, respectively, are greater than the correlation coefficient threshold of 0.3, then the original pattern functions with correlation coefficients greater than the correlation coefficient threshold—IMF5, IMF6, IMF7, IMF8, IMF9, IMF10, and IMF11—are the reconstructed original pattern functions.

[0098] Step S30: Reconstruct the signal for the reconstructed proof mode function to obtain the filtered signal after artifact filtering.

[0099] In this embodiment, the terminal device performs CEEMDAN decomposition on the second electromyography signal to obtain multiple original pattern functions. After identifying the reconstructed original pattern function related to the second electromyography signal among the multiple original pattern functions, the device further reconstructs the signal based on the reconstructed original pattern function to obtain the filtered signal after artifact filtering.

[0100] Optionally, in some feasible embodiments, the reconstructed proof mode function includes: the remaining reconstructed proof mode function, and step S30 may include the following steps:

[0101] Step S301: Identify the noisy target reconstruction proof mode function in the reconstructed proof proof mode function;

[0102] In this embodiment, the terminal device performs CEEMDAN decomposition on the second electromyography signal to obtain multiple original pattern functions. After identifying the reconstructed original pattern function related to the second electromyography signal among the multiple original pattern functions, the device further identifies the target reconstructed original pattern function containing noise among the reconstructed original pattern functions.

[0103] For example, in this embodiment, the terminal device performs CEEMDAN decomposition on the second electromyography signal x to obtain 14 IMF components. After confirming that IMF5, IMF6, IMF7, IMF8, IMF9, IMF10 and IMF11 related to the second electromyography signal x are the reconstructed intrinsic mode functions, the noisy IMF5, IMF6, IMF7 and IMF8 are further confirmed as the target reconstructed intrinsic mode functions in the reconstructed intrinsic mode functions.

[0104] Optionally, in some feasible embodiments, step S301 may include the following steps:

[0105] Step S3011: Confirm the sample entropy of the reconstructed proof mode function;

[0106] In this embodiment, the terminal device performs CEEMDAN decomposition on the first electromyography signal to obtain multiple original pattern functions. After identifying the reconstructed original pattern function related to the first electromyography signal among the multiple original pattern functions, the sample entropy of the reconstructed original pattern function is further identified.

[0107] For example, in this embodiment, assuming x(i) are the IMF components after CEEMDAN decomposition, i = i:N, the sample entropy calculation process is as follows: construct an m-dimensional vector X according to x(i) in sequence. i ,Right now:

[0108] X i =[x(i) x(i+1)…x(i+m-1)]

[0109] definition For X i With X j The distance between two vectors is defined as:

[0110]

[0111] Given a similarity tolerance r, statistically... The number is defined as C i And the ratio of this to the total number of vectors Nm is denoted as

[0112]

[0113] Then calculate its average value and denot it as B. m (r):

[0114]

[0115] Update the dimension by adding one dimension, m = m + 1, and repeat the above to obtain B. m+1 (r);

[0116] Therefore, the sample entropy is:

[0117] SampEN = -ln[B m+1 (r) / B m (r)]

[0118] It should be noted that in this embodiment, the sample entropy reflects the complexity of the time series signal. The larger the sample entropy, the more complex the signal, indicating that it contains more noise. Conversely, the smaller the sample entropy, the higher the autocorrelation of the signal and the less noise it contains.

[0119] For example, such as Figure 6 The diagram showing the sample entropy of the IMF components illustrates that, in this embodiment, the terminal device confirms that the sample entropy of the reconstructed proof mode function IMF5 is 0.7116, the sample entropy of IMF6 is 0.5440, the sample entropy of IMF7 is 0.4066, the sample entropy of IMF8 is 0.3394, the sample entropy of IMF9 is 0.3156, the sample entropy of IMF10 is 0.1997, and the sample entropy of IMF11 is 0.0784.

[0120] Step S3012: Confirm the target reconstructed proof pattern function in the reconstructed proof pattern function based on the sample entropy.

[0121] In this embodiment, after the terminal device confirms the sample entropy of the reconstructed proof pattern function, it further confirms the target reconstructed proof pattern function in the reconstructed proof pattern function based on the sample entropy.

[0122] For example, after the terminal device confirms that the sample entropy of the reconstructed certificate mode function IMF5 is 0.7116, the sample entropy of IMF6 is 0.5440, the sample entropy of IMF7 is 0.4066, the sample entropy of IMF8 is 0.3394, the sample entropy of IMF9 is 0.3156, the sample entropy of IMF10 is 0.1997, and the sample entropy of IMF11 is 0.0784, it further confirms the target reconstructed certificate mode function in the reconstructed certificate mode function based on the sample entropy.

[0123] Optionally, in some feasible embodiments, step S3012 may further include the following steps:

[0124] Step A: Confirm whether the sample quotient is greater than the preset sample quotient threshold;

[0125] In this embodiment, after the terminal device confirms the sample entropy of the reconstructed proof mode function, it confirms whether the sample quotient is greater than a preset sample quotient threshold.

[0126] For example, in this embodiment, after the terminal device confirms the sample entropy of the reconstructed proof mode function, if the preset sample quotient threshold is 0.32, it confirms whether the sample entropy of IMF5 (0.7116), IMF6 (0.5440), IMF7 (0.4066), IMF8 (0.3394), IMF9 (0.3156), IMF10 (0.1997), and IMF11 (0.0784) are greater than the preset sample quotient threshold of 0.32.

[0127] Step B: If it is confirmed that the sample entropy is greater than the sample entropy threshold, then in the reconstructed proof pattern function, the reconstructed proof pattern function whose sample entropy is greater than the sample entropy threshold is the target reconstructed proof pattern function.

[0128] In this embodiment, after the terminal device confirms whether the sample entropy is greater than the preset sample entropy threshold, if it confirms that the sample entropy is greater than the sample entropy threshold, then in the reconstructed proof mode function, the reconstructed proof mode function that confirms the sample entropy is greater than the sample entropy threshold is the target reconstructed proof mode function.

[0129] For example, in this embodiment, after the terminal device confirms whether the sample entropy of IMF5 (0.7116), IMF6 (0.5440), IMF7 (0.4066), IMF8 (0.3394), IMF9 (0.3156), IMF10 (0.1997), and IMF11 (0.0784) are greater than a preset sample quotient threshold of 0.32, if it is confirmed that the sample entropy of IMF5, IMF6, IMF7, and IMF8 is greater than the sample entropy threshold of 0.32, then in reconstructing the proof pattern functions IMF5, IMF6, IMF7, IMF8, IMF9, IMF10, and IMF11, IMF5, IMF6, IMF7, and IMF8 are identified as the target proof pattern functions for reconstruction.

[0130] Step S302: Perform noise reduction processing on the target reconstructed proof mode function to obtain the noise-reduced target reconstructed proof mode function;

[0131] In this embodiment, after the terminal device identifies the target reconstructed proof pattern function containing noise in the reconstructed proof pattern function, it performs noise reduction processing on the target reconstructed proof pattern function to obtain the noise-reduced target reconstructed proof pattern function.

[0132] For example, in this embodiment, after the terminal device identifies the noisy IMF5, IMF6, IMF7 and IMF8 as the target reconstruction proof mode functions in the reconstruction proof mode function, it performs noise reduction processing on IMF5, IMF6, IMF7 and IMF8 through bandpass filtering with a cutoff frequency of 20-500Hz to obtain noise-reduced IMF5, IMF6, IMF7 and IMF8.

[0133] Step S303: Perform signal reconstruction on the target reconstructed proof mode function after noise reduction and the remaining reconstructed proof mode function to obtain the filtered signal after artifact filtering.

[0134] In this embodiment, the terminal device performs noise reduction processing on the target reconstructed proof pattern function to obtain the noise-reduced target reconstructed proof pattern function. Then, it performs signal reconstruction on the noise-reduced target reconstructed proof pattern function and the remaining reconstructed proof pattern functions to obtain the filtered signal after artifact filtering.

[0135] It should be noted that, in this embodiment, the remaining proof pattern functions are the reconstructed proof pattern functions other than the target reconstructed proof pattern function in the reconstructed proof pattern functions. That is, among the reconstructed proof pattern functions of IMF5, IMF6, IMF7, IMF8, IMF9, IMF10 and IMF11, excluding the target reconstructed proof pattern functions IMF5, IMF6, IMF7 and IMF8, are the remaining proof pattern functions.

[0136] For example, in this embodiment, after the terminal device identifies noisy IMF5, IMF6, IMF7 and IMF8 as the target reconstructed proof pattern function in the reconstructed proof pattern function, it performs noise reduction processing on the target reconstructed proof pattern function to obtain noise-reduced IMF5, IMF6, IMF7 and IMF8. Then, it performs signal reconstruction on the noise-reduced IMF5, IMF6, IMF7 and IMF8 and IMF9, IMF10 and IMF11 to obtain the filtered signal after artifact filtering.

[0137] Additionally, it should be noted that in this embodiment, the signal-to-noise ratio (SNR) and normalized root mean square error (NRMSE) are selected as the quantitative evaluation metrics for filtering effectiveness. The formulas for calculating SNR and NRMSE are as follows:

[0138]

[0139]

[0140]

[0141]

[0142] Among them, SNR x SNR is the signal-to-noise ratio before filtering. y Let s(t) represent the signal-to-noise ratio after filtering, x(t) represent the voluntary electromyography signal, y(t) represent the input signal before filtering, y(t) represent the output signal after filtering, T represent the length of the entire data segment, and sd() represent the standard deviation. It should be noted that the output signal y(t) after filtering is different from the second electromyography signal y. Table 1 below shows the signal-to-noise ratio and normalized root mean square error statistics before and after signal filtering.

[0143]

[0144] Table 1

[0145] Thus, in this embodiment, when the terminal device filters artifacts from the fabric electrode, it acquires the first electromyography (EMG) signal after selective blanking of the fabric electrode. Then, after acquiring the first EMG signal after selective blanking of the fabric electrode, the terminal device performs CEEMDAN decomposition on the first EMG signal to obtain multiple intrinsic pattern functions, and identifies the reconstructed intrinsic pattern function related to the first EMG signal among these multiple intrinsic pattern functions. Finally, after performing CEEMDAN decomposition on the first EMG signal to obtain multiple intrinsic pattern functions, and identifying the reconstructed intrinsic pattern function related to the first EMG signal among these multiple intrinsic pattern functions, the terminal device further reconstructs the signal based on the reconstructed intrinsic pattern function to obtain the filtered signal after artifact filtering.

[0146] This invention acquires electromyography (EMG) signals from fabric electrodes, firstly performs selective blanking on the EMG signals, and then performs CEEMDAN decomposition and signal reconstruction on the selectively blanked EMG signals to obtain a filtered signal after artifact filtering. Thus, this invention achieves the processing of EMG signals acquired from fabric electrodes by combining selective blanking and signal reconstruction methods, thereby effectively eliminating motion artifacts, interference noise, and functional electrical stimulation artifacts in the fabric electrode EMG signals.

[0147] In addition, please refer to Figure 8 The present invention also proposes an electrical stimulation interference filtering device based on fabric electrodes. The electrical stimulation interference filtering device based on fabric electrodes of the present invention includes:

[0148] Acquisition module 10 is used to acquire a first electromyographic signal and confirm a second electromyographic signal after selective blanking processing based on the first electromyographic signal.

[0149] The confirmation module 20 is used to perform CEEMDAN decomposition on the second electromyography signal to obtain multiple original pattern functions, and to confirm the reconstructed original pattern function related to the second electromyography signal among the multiple original pattern functions.

[0150] The reconstruction module 30 is used to reconstruct the signal based on the reconstructed proof mode function to obtain the filtered signal after artifact filtering.

[0151] Optionally, module 10 includes:

[0152] The first confirmation unit is used to confirm the time period during which the SA wave of the first electromyographic signal appears;

[0153] The blanking processing unit is used to selectively blank the first electromyographic signal according to the time period to obtain a second electromyographic signal after selective blanking processing.

[0154] Optionally, the confirmation module 20 includes:

[0155] The second confirmation unit is used to confirm the correlation coefficient between each of the plurality of the original pattern functions and the second electromyographic signal;

[0156] The third confirmation unit is used to confirm the reconstructed thematic mode function related to the second electromyographic signal based on the correlation coefficient.

[0157] Optionally, the third confirmation unit includes:

[0158] The first confirmation subunit is used to confirm whether the correlation coefficient is greater than a preset correlation coefficient threshold.

[0159] The second confirmation subunit is used to confirm, if the correlation coefficient is greater than the correlation coefficient threshold, that the proof pattern function with a correlation coefficient greater than the correlation coefficient threshold is a reconstructed proof pattern function.

[0160] Optionally, the reconstructed proof mode function includes: a remaining reconstructed proof mode function, and a reconstruction module 30, including:

[0161] The fourth confirmation unit is used to confirm the noisy target reconstruction proof mode function in the reconstructed proof mode function;

[0162] The noise reduction processing unit is used to perform noise reduction processing on the target reconstructed proof mode function to obtain the noise-reduced target reconstructed proof mode function.

[0163] The signal reconstruction unit is used to reconstruct the target reconstructed intrinsic mode function and the remaining reconstructed intrinsic mode function after noise reduction to obtain the filtered signal after artifact filtering.

[0164] Optionally, the fourth confirmation unit includes:

[0165] The third confirmation subunit is used to confirm the sample entropy of the reconstructed proof mode function;

[0166] The fourth confirmation subunit is used to confirm the target reconstructed proof pattern function in the reconstructed proof pattern function based on the sample entropy.

[0167] Optionally, the fourth confirmation subunit includes:

[0168] The first confirmation unit is used to confirm whether the sample quotient is greater than a preset sample entropy threshold.

[0169] The second confirmation unit is used to, if it is confirmed that the sample entropy is greater than the sample entropy threshold, then in the reconstructed proof pattern function, confirm that the reconstructed proof pattern function with sample entropy greater than the sample entropy threshold is the target reconstructed proof pattern function.

[0170] Furthermore, this embodiment of the invention also proposes a terminal device, which includes: a memory, a processor, and a fabric electrode-based electrical stimulation interference filtering program stored in the memory and executable on the processor. When the fabric electrode-based electrical stimulation interference filtering program is executed by the processor, it implements the steps of the fabric electrode-based electrical stimulation interference filtering method as described above.

[0171] The steps implemented when the fabric electrode-based electrical stimulation interference filtering program running on the processor is executed can be referred to in various embodiments of the fabric electrode-based electrical stimulation interference filtering method of the present invention, and will not be repeated here.

[0172] Furthermore, this embodiment of the invention also proposes a computer storage medium for use in a computer. The computer storage medium can be a non-volatile computer-readable storage medium. The computer storage medium stores a program for filtering electrical stimulation interference based on fabric electrodes. When the program for filtering electrical stimulation interference based on fabric electrodes is executed by a processor, it implements the steps of the method for filtering electrical stimulation interference based on fabric electrodes as described above.

[0173] The steps implemented when the fabric electrode-based electrical stimulation interference filtering program running on the processor is executed can be referred to in various embodiments of the fabric electrode-based electrical stimulation interference filtering method of the present invention, and will not be repeated here.

[0174] It should be noted that, in this document, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or system that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such a process, method, article, or system. Unless otherwise specified, an element defined by the phrase "comprising one..." does not exclude the presence of other identical elements in the process, method, article, or system that includes that element.

[0175] The sequence numbers of the above embodiments of the present invention are for descriptive purposes only and do not represent the superiority or inferiority of the embodiments.

[0176] Through the above description of the embodiments, those skilled in the art can clearly understand that the methods of the above embodiments can be implemented by means of software plus necessary general-purpose hardware platforms. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product is stored in a computer storage medium (such as Flash memory, ROM / RAM, magnetic disk, optical disk), and includes several instructions to cause a controller in a terminal device (which may be a mobile phone, computer, server, or network device, etc.) to control the data read and write operations of the storage medium to execute the methods described in the various embodiments of the present invention.

[0177] The above are merely preferred embodiments of the present invention and do not limit the scope of the patent. Any equivalent structural or procedural transformations made based on the description and drawings of the present invention, or direct or indirect applications in other related technical fields, are similarly included within the scope of patent protection of the present invention.

Claims

1. A method for filtering electrical stimulation interference based on fabric electrodes, characterized in that, Applied to FES devices, the method for filtering electrical stimulation interference based on fabric electrodes includes the following steps: Acquire a first electromyography (EMG) signal collected by the fabric electrodes of the FES device, and confirm a second EMG signal after selective blanking processing based on the first EMG signal. The step of confirming the second EMG signal after selective blanking processing based on the first EMG signal includes: confirming the time period of the SA wave in the first EMG signal; performing selective blanking processing on the first EMG signal based on the time period to obtain the second EMG signal after selective blanking processing, so as to eliminate functional electrical stimulation artifacts in the first EMG signal. The second electromyographic signal is subjected to CEEMDAN decomposition to obtain multiple original pattern functions, and the reconstructed original pattern function related to the second electromyographic signal is identified among the multiple original pattern functions. Signal reconstruction is performed on the reconstructed proof mode function to obtain a filtered signal after artifact filtering. The signal reconstruction process for the reconstructed proof mode function to obtain the filtered signal after artifact filtering includes: The reconstructed evidence pattern function includes a residual reconstructed evidence pattern function; in the reconstructed evidence pattern function, a target reconstructed evidence pattern function containing noise is identified based on the sample entropy being greater than a preset sample entropy threshold; the target reconstructed evidence pattern function is subjected to noise reduction processing based on a preset cutoff frequency, and the noise-reduced target reconstructed evidence pattern function and the residual reconstructed evidence pattern function are used for signal reconstruction to obtain the filtered signal after artifact filtering, so as to further eliminate motion artifacts and interference noise in the first electromyography signal.

2. The method for filtering electrical stimulation interference based on fabric electrodes as described in claim 1, characterized in that, The step of identifying the reconstructed evidence pattern function related to the second electromyographic signal among the plurality of evidence pattern functions includes: Confirm the correlation coefficients between each of the aforementioned pattern functions and the second electromyographic signal; The reconstructed theorem pattern function associated with the second electromyographic signal is confirmed based on the correlation coefficient.

3. The method for filtering electrical stimulation interference based on fabric electrodes as described in claim 2, characterized in that, The step of confirming the reconstructed intrinsic pattern function related to the second electromyographic signal based on the correlation coefficient includes: Confirm whether the correlation coefficient is greater than the preset correlation coefficient threshold: If it is confirmed that the correlation coefficient is greater than the correlation coefficient threshold, then the proof pattern function with a correlation coefficient greater than the correlation coefficient threshold is confirmed as the reconstructed proof pattern function.

4. The method for filtering electrical stimulation interference based on fabric electrodes as described in claim 1, characterized in that, The step of identifying a noisy target reconstruction pattern function based on a sample entropy greater than a preset sample entropy threshold in the reconstructed proof pattern function includes: Confirm the sample entropy of the reconstructed proof mode function; The target reconstruction proof pattern function is confirmed in the reconstruction proof pattern function based on the sample entropy.

5. A device for filtering electrical stimulation interference based on fabric electrodes, characterized in that, The fabric electrode-based electrical stimulation interference filtering device, applied to FES devices, includes: The acquisition module is used to acquire the first electromyographic signal collected by the fabric electrode of the FES device, and to confirm the second electromyographic signal after selective blanking processing based on the first electromyographic signal. The acquisition module is further configured to confirm the time period of the SA wave occurrence of the first electromyography signal; and to perform selective blanking processing on the first electromyography signal according to the time period to obtain a second electromyography signal after selective blanking processing, so as to eliminate functional electrical stimulation artifacts in the first electromyography signal. The confirmation module is used to perform CEEMDAN decomposition on the second electromyography signal to obtain multiple original pattern functions, and to confirm the reconstructed original pattern function related to the second electromyography signal among the multiple original pattern functions. The reconstruction module is used to reconstruct the signal based on the reconstructed proof mode function to obtain the filtered signal after artifact filtering. The reconstructed proof mode function includes the remaining reconstructed proof mode function. The reconstruction is further used to identify a noisy target reconstruction pattern function based on the sample entropy being greater than a preset sample entropy threshold in the reconstructed pattern function; to perform noise reduction processing on the target reconstruction pattern function based on a preset cutoff frequency; and to perform signal reconstruction between the noise-reduced target reconstruction pattern function and the remaining reconstruction pattern functions to obtain the filtered signal after artifact filtering, so as to further eliminate motion artifacts and interference noise in the first electromyography signal.

6. A terminal device, characterized in that, The terminal device includes: a memory, a processor, and a fabric electrode-based electrical stimulation interference filtering program stored in the memory and executable on the processor, wherein when the fabric electrode-based electrical stimulation interference filtering program is executed by the processor, it implements the steps of the fabric electrode-based electrical stimulation interference filtering method as described in any one of claims 1 to 4.

7. A computer storage medium, characterized in that, The computer storage medium stores a program for filtering electrical stimulation interference based on fabric electrodes, which, when executed by a processor, implements the steps of the method for filtering electrical stimulation interference based on fabric electrodes as described in any one of claims 1 to 4.

Images