Electrical stimulation rehabilitation training system and method for neurosurgical nursing

Abstract

The invention provides an electrical stimulation rehabilitation training system and method for neurosurgical nursing. The method comprises the following steps: collecting a plurality of electromyographic signals of an uninjured side part of a target patient; selecting one electromyographic signal as a selected electromyographic signal, determining a plurality of motion states of the uninjured side part of the target patient according to the density of time-frequency points in each time-frequency window of the selected electromyographic signal, and decomposing corresponding sub-signal segments of the selected electromyographic signal in each motion state into a plurality of electromyographic components, continuing to determine a plurality of electromyographic components of the remaining electromyographic signals; determining distribution time sequences of different motion units when the target patient moves according to all the myoelectricity components in the same motion state; and generating a plurality of electrical stimulation pulse signals based on the cepstrum entropy of all the distribution time sequences, and applying electrical stimulation to the affected side part of the target patient through all the electrical stimulation pulse signals. By adopting the scheme of the application, the synchronism of the motion intention of the patient and the affected side stimulation can be improved.

Description

Technical Field

[0001] The present application relates to the field of medical rehabilitation technology, and more specifically, to an electrical stimulation rehabilitation training system and method for neurosurgery care. Background Art

[0002] Neurosurgical nursing is a crucial component of rehabilitation for patients with neurological disorders such as stroke and spinal cord injury. Its core goal is to restore patients' motor function and neural control. Electrical stimulation rehabilitation training, as a non-invasive treatment, stimulates the affected muscles or nerves by simulating nerve impulses and has been widely used to promote neuroplasticity and restore motor function.

[0003] Electrical stimulation rehabilitation training methods are primarily based on the principle of Hebbian learning. They stimulate the affected side through bioelectrical signals from the healthy side's movements, creating a movement illusion and promoting neurological compensation. In the existing art, electrical stimulation rehabilitation training is typically implemented through two methods: one is to analyze healthy side movements based on a fixed threshold and stimulate the affected side's muscles. This method lacks consideration for individual patient differences, resulting in a mismatch between the electrical stimulation signals on the affected side and the patient. The other is to model the healthy side's electromyographic signals. For example, the article "Adaptive Activation Feature Extraction Algorithm for Surface Electromyography" detects motion signals using extreme point space. However, this modeling method is computationally complex and lacks real-time performance. All of these methods can cause a mismatch between the patient's movement intention and the stimulation on the affected side (method one causes a mismatch between stimulation intensity and frequency, and method two causes a mismatch between stimulation timing). This weakens the neurophysiological basis for closed-loop reconstruction of motor function and, in turn, affects the activation of neurological compensation on the affected side. Therefore, improving the synchronization between the patient's movement intention and stimulation on the affected side has become a challenge facing the industry. Summary of the Invention

[0004] The present application provides an electrical stimulation rehabilitation training system and method for neurosurgery nursing, which can improve the synchronization between the patient's movement intention and the stimulation of the affected side.

[0005] In a first aspect, the present application provides an electrical stimulation rehabilitation training method for neurosurgery nursing, comprising: Multiple sampling points are set on the healthy side of the target patient, and the electromyographic signals at each sampling point are collected; Selecting an electromyographic signal as a selected electromyographic signal, dividing a time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows, and determining muscle state characteristics of the target patient at each moment according to the density of time-frequency points in each time-frequency window; Identifying multiple motion states of the healthy side of the target patient based on all muscle state features, decomposing the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components, and continuing to determine multiple electromyographic components of the remaining electromyographic signal; Determine the firing time series of different motor units during the target patient's movement based on the trend coefficients of all myoelectric components under the same movement state; A plurality of electrical stimulation pulse signals are generated based on the cepstrum entropy of all the emission time series, and then electrical stimulation is applied to the affected side of the target patient through all the electrical stimulation pulse signals.

[0006] In some embodiments, dividing the time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows specifically includes: determining a time-frequency diagram of the selected electromyographic signal; Dividing the time axis of the time-frequency graph according to the time interval units to obtain a plurality of time windows; The frequency axis of each time window is divided according to a preset frequency passband range to obtain multiple time-frequency windows.

[0007] In some embodiments, determining the muscle state characteristics of the target patient at each moment according to the density of the time-frequency points in each time-frequency window specifically includes: Determine the density of time-frequency points in each time-frequency window; Determine the time-frequency distribution characteristics within each time window according to the density of time-frequency points in all time-frequency windows; Initialize the upper limit and lower limit of the time-frequency distribution; All time-frequency distribution features are dynamically normalized according to the time-frequency distribution upper limit and the time-frequency distribution lower limit, and all normalized time-frequency distribution features are used as muscle state features of the target patient at each moment.

[0008] In some embodiments, identifying multiple motion states of the target patient's healthy side based on all muscle state features specifically includes: Preset motion state threshold; All muscle state features are sequentially compared with the motion state thresholds to obtain multiple motion states of the healthy side of the target patient.

[0009] In some embodiments, decomposing the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components specifically includes: dividing the selected electromyographic signal into a plurality of sub-signal segments according to all motion states; For each sub-signal segment, each sub-signal segment is decomposed into multiple myoelectric components according to the envelope of each sub-signal segment.

[0010] In some embodiments, determining the firing time series of different motor units when the target patient exercises based on the trend coefficients of all myoelectric components under the same motion state specifically includes: Divide all motion states into multiple motion state clusters under the same motion state; selecting a movement state cluster as a selected movement state cluster, and determining trend coefficients of all myoelectric components under each movement state in the selected movement state cluster; determining a firing time subsequence of each motor unit according to all trend coefficients, and continuing to determine a firing time subsequence of each motor unit according to the remaining selected motion state clusters; The firing time sequence of each motor unit during the target patient's movement is reconstructed based on the firing time subsequence of each motor unit.

[0011] In some embodiments, generating multiple electrical stimulation pulse signals based on the cepstrum entropy of all firing time series specifically includes: Determine the cepstral entropy of each emission time series; determining a plurality of electrical stimulation parameters according to all cepstral entropies; A plurality of electrical stimulation signals are generated according to all electrical stimulation parameters.

[0012] In a second aspect, the present application provides an electrical stimulation rehabilitation training system for neurosurgery nursing, comprising: An acquisition module is used to set multiple sampling points on the healthy side of the target patient and collect the electromyographic signals at each sampling point; a processing module, configured to select an electromyographic signal as a selected electromyographic signal, divide a time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows, and determine muscle state characteristics of the target patient at each moment according to the density of time-frequency points in each time-frequency window; The processing module is further configured to identify multiple motion states of the healthy side of the target patient based on all muscle state features, decompose the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components, and continue to determine multiple electromyographic components of the remaining electromyographic signal; The processing module is further configured to determine the firing time series of different motor units when the target patient exercises based on the trend coefficients of all myoelectric components under the same motion state; The execution module is used to generate multiple electrical stimulation pulse signals based on the cepstrum entropy of all the emission time series, and then apply electrical stimulation to the affected side of the target patient through all the electrical stimulation pulse signals.

[0013] In a third aspect, the present application provides a computer device comprising a memory and a processor, wherein the memory stores a code, and the processor is configured to obtain the code and execute the above-mentioned electrical stimulation rehabilitation training method for neurosurgical care.

[0014] In a fourth aspect, the present application provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the above-mentioned electrical stimulation rehabilitation training method for neurosurgical care.

[0015] The technical solutions provided by the embodiments disclosed in this application have the following beneficial effects: In the electrical stimulation rehabilitation training system and method for neurosurgical nursing provided by the present application, first, multiple sampling points are set at the healthy side of the target patient, and the electromyographic signals at each sampling point are collected; one electromyographic signal is selected as the selected electromyographic signal, and the time-frequency diagram of the selected electromyographic signal is divided into multiple time-frequency windows, and the muscle state characteristics of the target patient at each moment are determined according to the density of the time-frequency points in each time-frequency window; based on all the muscle state characteristics, multiple motion states of the healthy side of the target patient are identified, and then the sub-signal segments corresponding to the selected electromyographic signal in each motion state are decomposed into multiple electromyographic components, and the multiple electromyographic components of the remaining electromyographic signals are further determined; based on the trend coefficients of all electromyographic components under the same motion state, the emission time series of different motor units of the target patient when exercising are determined; based on the inverse spectral entropy of all emission time series, multiple electrical stimulation pulse signals are generated, and then electrical stimulation is applied to the affected side of the target patient through all the electrical stimulation pulse signals.

[0016] It can be seen from this that the present application collects the electromyographic signals of different muscles on the healthy side of the patient, and adaptively divides the movement state of the target patient's healthy side muscles by transforming the density of the time-frequency points of the electromyographic signals in different time-frequency windows, so as to quickly identify the movement state of the patient's healthy side muscles, thereby dividing the electromyographic signals of various muscles into sub-signal segments under different movement states. Subsequently, the electromyographic signals (i.e., the emission time series) emitted by the same motor unit are screened out through the trend of each sub-signal segment, and finally electrical stimulation is applied to the corresponding position of the patient's affected side according to the emission time series of each motor unit, thereby ensuring that the movement of the patient's healthy side is synchronized to the patient's affected side in real time for neural activation. In summary, the scheme of the present application can improve the synchronization between the patient's movement intention and the stimulation of the affected side. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 is an exemplary flow chart of an electrical stimulation rehabilitation training method for neurosurgery care according to some embodiments of the present application; Figure 2 is an exemplary flow chart of dividing time-frequency windows according to some embodiments of the present application; Figure 3 is an exemplary flow chart of generating an electrical stimulation pulse signal according to some embodiments of the present application; Figure 4is a schematic structural diagram of an electrical stimulation rehabilitation training system for neurosurgery nursing according to some embodiments of the present application; Figure 5 This is a structural diagram of a computer device for implementing an electrical stimulation rehabilitation training method for neurosurgical care according to some embodiments of the present application. DETAILED DESCRIPTION

[0018] In order to better understand the technical solution of the present application, the technical solution of the present application will be described in detail below with reference to the accompanying drawings and specific implementation methods.

[0019] refer to Figure 1 , which is an exemplary flow chart of an electrical stimulation rehabilitation training method for neurosurgery care according to some embodiments of the present application. The electrical stimulation rehabilitation training method 100 for neurosurgery care mainly includes the following steps: In step 101, a plurality of sampling points are set on the healthy side of a target patient, and electromyographic signals at each sampling point are collected.

[0020] In a specific implementation, setting multiple sampling points on the healthy side of the target patient can be achieved in the following manner, namely: marking the muscle boundaries on the healthy side of the target patient using the surface anatomical marking positioning method in the prior art, and then setting a sampling point at the center of the area surrounded by each muscle boundary. In other embodiments, multiple sampling points can also be set on the healthy side of the target patient by other methods, which are not limited here.

[0021] In specific implementation, the collection of electromyographic signals at each sampling point can be achieved in the following manner: first, a flexible electrode array is attached at each sampling point, and the electromyographic signals at each sampling point are collected through the flexible electrode array at each sampling point. The sampling frequency of each flexible electrode array can be set according to actual needs. For example, in this application, the sampling frequency is preset to 2000 Hz.

[0022] In step 102, an electromyographic signal is selected as a selected electromyographic signal, a time-frequency diagram of the selected electromyographic signal is divided into a plurality of time-frequency windows, and the muscle state characteristics of the target patient at each moment are determined according to the density of the time-frequency points in each time-frequency window.

[0023] In some embodiments, reference Figure 2 This figure is an exemplary flow chart of dividing the time-frequency window according to some embodiments of the present application. In the present application, dividing the time-frequency graph of the selected electromyographic signal into multiple time-frequency windows can be achieved by using the following steps: In step 1021, a time-frequency diagram of the selected electromyographic signal is determined; In step 1022, the time axis of the time-frequency graph is divided according to the time interval units to obtain a plurality of time windows; In step 1023, the frequency axis of each time window is divided according to a preset frequency passband range to obtain multiple time-frequency windows.

[0024] In specific implementation, the time-frequency diagram of the selected electromyographic signal can be determined in the following manner, namely: the time-frequency diagram of the selected electromyographic signal is determined by the short-time Fourier transform in the prior art, wherein the window length in the short-time Fourier transform process can be preset to 4000. In other embodiments, the time-frequency diagram of the selected electromyographic signal can also be determined by other time-frequency analysis algorithms, which is not limited here.

[0025] It should be noted that the time interval unit in this application is a parameter preset according to actual needs. The larger the time interval unit, the more features in the selected electromyographic signal can be captured, but the calculation takes a long time and cannot meet the real-time requirements. The smaller the time interval unit, the better the real-time requirements can be, but the fewer features in the selected electromyographic signal can be captured. Therefore, this application adaptively sets the time interval unit according to the sampling frequency of the selected electromyographic signal and the characteristics of the time-frequency diagram. For example, the ratio of the window length in the process of determining the time-frequency diagram to the sampling frequency of the selected electromyographic signal can be used as the time interval unit.

[0026] In a specific implementation, the time axis of the time-frequency diagram is divided according to the time interval unit to obtain multiple time windows, which can be implemented by sampling in the following manner: the time axis of the time-frequency diagram is divided into multiple intervals of length n, and the divided intervals are all used as time windows, where n is the time interval unit.

[0027] In specific implementation, the frequency axis of each time window is divided according to the preset frequency passband range, and multiple time-frequency windows can be obtained by sampling in the following manner, namely: for each time window, the frequency axis of each time window is divided into multiple intervals of length m, and the divided intervals are all used as time-frequency windows, wherein m is the preset frequency passband range, and the frequency passband range is a parameter preset according to actual needs. For example, in this application, the frequency passband range can be preset to 70Hz.

[0028] In some embodiments, determining the muscle state characteristics of the target patient at each moment based on the density of the time-frequency points in each time-frequency window can be achieved by using the following steps: Determine the density of time-frequency points in each time-frequency window; Determine the time-frequency distribution characteristics within each time window according to the density of time-frequency points in all time-frequency windows; Initialize the upper limit and lower limit of the time-frequency distribution; All time-frequency distribution features are dynamically normalized according to the time-frequency distribution upper limit and the time-frequency distribution lower limit, and all normalized time-frequency distribution features are used as muscle state features of the target patient at each moment.

[0029] In specific implementation, the density of the time-frequency points in each time-frequency window can be determined in the following manner, namely: calculating the power of each time-frequency point in each time-frequency window, and taking the sum of the powers of all time-frequency points in each time-frequency window as the density of the time-frequency points in each time-frequency window.

[0030] It should be noted that the density of the time-frequency points in this application is the sum of the powers of all time-frequency points in a unit time-frequency window.

[0031] In specific implementation, the following method can be used to determine the time-frequency distribution characteristics in each time window based on the density of time-frequency points in all time-frequency windows: first, an initial value is defined to provide a benchmark for the subsequent identification of time-frequency distribution characteristics. The density of time-frequency points in all time-frequency windows in the first time window and the density of time-frequency points in all time-frequency windows in the second time window can be averaged, and the obtained average value is used as the initial value to provide a benchmark for subsequent time-frequency distribution characteristics. Then, starting from the third time window, the power of each time-frequency point in each time window is compared with the initial value in turn, and the sum of the powers of all time-frequency points greater than the initial value is used as the time-frequency distribution characteristic in each time window.

[0032] It should be noted that the time-frequency distribution characteristics in this application are parameters used to reflect the changes in the time-frequency point density of the electromyographic signal in each time period.

[0033] In specific implementation, the initialization of the upper limit and lower limit of the time-frequency distribution can be achieved in the following way: in order to prevent errors introduced by the sampling of the electromyographic signal, the data in the first time window can be discarded, and the density of the time-frequency points in the second time window can be doubled to compensate for the impact of low frequency resolution during short-time Fourier transform, that is, twice the density of the time-frequency points in the second time window is used as the upper and lower limits of the density.

[0034] In specific implementation, all time-frequency distribution features are dynamically normalized according to the time-frequency distribution upper limit and the time-frequency distribution lower limit, and all normalized time-frequency distribution features are used as the muscle state features of the target patient at each moment. This can be achieved in the following way, namely: in order to prevent errors introduced by sampling of electromyographic signals, the time-frequency distribution features of the first time window are discarded, and since the time-frequency distribution features of the second time window are used to initialize the time-frequency distribution upper limit and the time-frequency distribution lower limit, the time-frequency distribution features of the second time window are discarded, that is, first, all time windows are arranged in chronological order, and then the time-frequency distribution features of the first time window and the time-frequency distribution features of the second time window are discarded, and the normalized results are output as -1, then, dynamic normalization is performed starting from the third time window, that is, first the time-frequency distribution feature A of the third time window is compared with the size of the time-frequency distribution upper limit and the time-frequency distribution lower limit. If A is between the time-frequency distribution upper limit and the time-frequency distribution lower limit, then A is normalized to the interval of [-1,1] through the time-frequency distribution upper limit and the time-frequency distribution lower limit. If A is greater than the time-frequency distribution upper limit, the time-frequency distribution upper limit is updated to A, and the normalized result is output as 1. If A is less than the time-frequency distribution lower limit, the time-frequency distribution lower limit is updated to A, and the normalized result is output as -1. Repeat the above dynamic normalization steps to normalize the time-frequency distribution features of subsequent time windows in turn, and use all normalized time-frequency distribution features as the muscle state features at each moment.

[0035] It should be noted that when the human body moves, the electromyographic signal changes significantly. Specifically, when the muscle contracts, the energy of the electromyographic signal is large, and when the muscle relaxes, the energy of the electromyographic signal is small. Therefore, the change of the time-frequency points (that is, in the step of determining the time-frequency distribution characteristics, the sum of the power of the time-frequency points greater than the initial value is used as the time-frequency distribution characteristics) can be used to quickly identify the patient's movement state, and the normalized time-frequency distribution characteristics, that is, the muscle state characteristics, are used for identification, thereby further improving the real-time performance of movement state recognition.

[0036] In step 103, multiple movement states of the healthy side of the target patient are identified based on all muscle state characteristics, and then the sub-signal segments corresponding to the selected electromyographic signal in each movement state are decomposed into multiple electromyographic components, and the multiple electromyographic components of the remaining electromyographic signal are further determined.

[0037] In some embodiments, identifying multiple motion states of the target patient's healthy side based on all muscle state features can be achieved by using the following steps: Preset motion state threshold; All muscle state features are sequentially compared with the motion state thresholds to obtain multiple motion states of the healthy side of the target patient.

[0038] It should be noted that when the muscle is relaxed, the number and power of frequency points in the electromyographic signal are small, and the muscle state characteristics obtained after normalization are all non-positive values. When the muscle contracts, the number and power of frequency points in the electromyographic signal increase significantly, and the muscle state characteristics obtained after normalization are all positive values. Therefore, the motion state threshold can be preset to 0.

[0039] In specific implementation, all muscle state features are compared with the motion state threshold in turn, and then multiple motion states of the healthy side of the target patient are obtained. This can be achieved in the following way, namely: all muscle state features are compared with the motion state threshold in turn, and the motion states of the time periods corresponding to all muscle state features greater than the motion state threshold are marked as muscle contraction, and the motion states of the time periods corresponding to all muscle state features less than or equal to the motion state threshold are marked as muscle relaxation.

[0040] It should be noted that the movement state in this application refers to the activity state of the target patient's muscles, and the movement state includes two states: muscle contraction and muscle relaxation.

[0041] In some embodiments, decomposing the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components can be achieved by using the following steps: dividing the selected electromyographic signal into a plurality of sub-signal segments according to all motion states; For each sub-signal segment, each sub-signal segment is decomposed into multiple myoelectric components according to the envelope of each sub-signal segment.

[0042] In specific implementation, the selected electromyographic signal can be divided into multiple sub-signal segments according to all movement states in the following way, namely: the data segments corresponding to the selected electromyographic signal in the time period of each movement state are cut out, and all the cut-out data segments are used as sub-signal segments corresponding to each movement state.

[0043] It should be noted that the neutron signal segment in this application is the signal segment corresponding to the electromyographic signal in a specific motion state.

[0044] In specific implementation, decomposing each sub-signal segment into multiple electromyographic components according to the envelope of each sub-signal segment can be achieved in the following manner, namely: first, extracting the envelope of each sub-signal segment by Hilbert transform in the prior art, and then, according to the empirical mode decomposition in the prior art, decomposing each sub-signal segment into multiple component signals according to the envelope of each sub-signal segment, and taking all the component signals decomposed from each sub-signal segment as the myoelectric components of each sub-signal segment.

[0045] It should be noted that the myoelectric component in this application refers to the component of the myoelectric signal in different modes.

[0046] In step 104 , the firing time series of different motor units when the target patient is exercising are determined based on the trend coefficients of all myoelectric components under the same motion state.

[0047] In some embodiments, determining the firing time series of different motor units during the target patient's movement based on the trend coefficients of all myoelectric components under the same movement state can be achieved by using the following steps: Divide all motion states into multiple motion state clusters under the same motion state; selecting a movement state cluster as a selected movement state cluster, and determining trend coefficients of all myoelectric components under each movement state in the selected movement state cluster; determining a firing time subsequence of each motor unit according to all trend coefficients, and continuing to determine a firing time subsequence of each motor unit according to the remaining selected motion state clusters; The firing time sequence of each motor unit during the target patient's movement is reconstructed based on the firing time subsequence of each motor unit.

[0048] In specific implementation, all motion states are divided into multiple motion state clusters under the same motion state, which can be achieved in the following manner, namely: obtaining the starting time of the corresponding time period of each motion state, calculating the difference between the starting times of the corresponding time periods of each two motion states in which the motion state is muscle relaxation, and taking the set formed by the motion states whose difference is less than the preset time difference threshold as the motion state cluster, calculating the difference between the starting times of the corresponding time periods of each two motion states in which the motion state is muscle contraction, and taking the set formed by the motion states whose difference is less than the preset time difference threshold as the motion state cluster, wherein the time difference threshold can be preset according to actual needs. For example, in this application, the time difference threshold is preset to 0.1 second.

[0049] It should be noted that, in the present application, a motion state cluster is a collection of the same motion states within a specific time period.

[0050] In specific implementation, determining the trend coefficient of all myoelectric components under each motion state in the selected motion state cluster can be achieved in the following manner, namely: first, calculating the Hurst index of each myoelectric component under each motion state in the selected motion state cluster, and using the Hurst index of each myoelectric component as the trend coefficient of all myoelectric components under each motion state in the selected motion state cluster.

[0051] It should be noted that the trend coefficient in this application is a parameter indicating the long-term dependence of the myoelectric component.

[0052] In specific implementation, the following method can be used to determine the firing time subsequence of each motor unit based on all trend coefficients, namely: first, calculate the difference between every two trend coefficients in the same motion state cluster, and screen out the electromyographic components whose difference is less than the preset trend difference threshold as the component cluster; for each component cluster, calculate the energy of each electromyographic component in the component cluster, and take the sampling point corresponding to the electromyographic component with the largest energy as the motor unit corresponding to the component cluster; and then take all the electromyographic components in each component cluster as the firing time subsequence of the motor unit corresponding to each component cluster.

[0053] It should be noted that, in this application, the release time subsequence is a component sequence of one mode of the release time sequence.

[0054] In specific implementation, reconstructing the emission time sequence of each motor unit during the target patient's movement based on the emission time subsequence of each motor unit can be achieved in the following way: for each motor unit, the sequence distribution obtained by adding all the emission time subsequences of each motor unit is used as the emission time sequence of each motor unit.

[0055] It should be noted that the firing time series in this application is a sequence that describes the moments when a muscle fiber group of a specific muscle triggers action potentials during muscle activity, arranged in chronological order.

[0056] In step 105 , an electrical stimulation pulse signal is generated based on the cepstrum entropy of all the emission time series, and then electrical stimulation is applied to the affected side of the target patient through the electrical stimulation pulse signal.

[0057] In some embodiments, reference Figure 3 This figure is an exemplary flow chart of generating electrical stimulation pulse signals according to some embodiments of the present application. In the present application, generating multiple electrical stimulation pulse signals based on the cepstrum entropy of all emission time series can be implemented by the following steps: In step 1051, the cepstral entropy of each emission time series is determined; In step 1052, a plurality of electrical stimulation parameters are determined according to all the cepstral entropies; In step 1053 , a plurality of electrical stimulation signals are generated according to all the electrical stimulation parameters.

[0058] In specific implementation, the cepstrum entropy of each emission time series can be determined in the following way, namely: first, perform Fourier transform on each emission time series to obtain the spectrum of each emission time series, then take the natural logarithm of each spectrum, and then perform inverse Fourier transform on each spectrum after taking the natural logarithm, and use the distribution of each sequence after inverse Fourier transform as the cepstrum of each emission time series. Finally, calculate the information entropy of each cepstrum, and use the information entropy of each cepstrum as the cepstrum entropy of each emission time series.

[0059] It should be noted that, in this application, the cepstral entropy is a parameter representing the complexity of the emission time series.

[0060] In specific implementation, multiple electrical stimulation parameters can be determined based on all cepstral entropies in the following manner, namely: first, all cepstral entropies are normalized to the range of zero to one. The higher the cepstral entropy, the more complex the patient's movement, and low-frequency and high-intensity electrical stimulation is required to match the movement. The lower the cepstral entropy, the simpler the patient's movement, and high-frequency and low-intensity stimulation is used to quickly activate the muscle. Therefore, the normalized cepstral entropy can be used to reconstruct the electrical stimulation signal. For example, a release time series is selected as the selected release time series, and the difference between the maximum and minimum values ​​in the selected release time series is compared with the selected release time series. The normalized cepstral entropy corresponding to the time series is multiplied, and the sum of the obtained value and the minimum value in the selected release time series is used as the pulse amplitude of the selected release time series. The difference between the maximum frequency and the minimum frequency in the selected release time series is multiplied by the normalized cepstral entropy corresponding to the selected release time series. The multiplied value is then subtracted from the maximum frequency of the selected release time series. The difference is used as the pulse frequency of the selected release time series, and both the pulse amplitude and the pulse frequency are used as the electrical stimulation parameters of the selected release time series. The electrical stimulation parameters of the remaining selected release time series are continued to be determined.

[0061] It should be noted that the electrical stimulation parameters in this application are parameters used to regulate the electrical stimulation signals.

[0062] In specific implementation, generating multiple electrical stimulation signals according to all electrical stimulation parameters can be achieved in the following manner, namely: selecting a motor unit as the selected motor unit, obtaining the electrical stimulation parameters of the emission time series of the selected motor unit, generating a pulse signal according to the pulse amplitude and pulse frequency in the electrical stimulation parameters, the pulse width can be preset to any value between 200 and 300 microseconds according to clinical experience, and using the generated pulse signal as the electrical stimulation signal of the selected motor unit, and continuing to generate electrical stimulation signals for the remaining motor units.

[0063] It should be noted that the electrical stimulation pulse signal in the present application is a pulse electrical signal used to apply electrical stimulation to a patient.

[0064] In specific implementation, applying electrical stimulation to the affected side of the target patient through all electrical stimulation pulse signals can be achieved in the following manner, namely: setting multiple stimulation points on the affected side of the target patient, and attaching electrical stimulation electrodes at each stimulation point, wherein each stimulation point corresponds one-to-one to a sampling point on the healthy side of the target patient, and then inputting the electrical stimulation pulse signal of the motor unit corresponding to each sampling point into the electrical stimulation electrode at the stimulation point corresponding to each sampling point, and applying electrical stimulation to the affected side of the target patient through the electrical stimulation electrode.

[0065] In addition, in another aspect of the present application, in some embodiments, the present application provides an electrical stimulation rehabilitation training system for neurosurgery nursing, referring to Figure 4 , which is a schematic diagram of the structure of an electrical stimulation rehabilitation training system for neurosurgery nursing according to some embodiments of the present application. The electrical stimulation rehabilitation training system 400 for neurosurgery nursing includes: an acquisition module 401, a processing module 402, and an execution module 403, which are described as follows: Acquisition module 401, in this application, the acquisition module 401 is mainly used to set multiple sampling points on the healthy side of the target patient and collect the electromyographic signals at each sampling point; Processing module 402, in this application, is mainly used to select an electromyographic signal as a selected electromyographic signal, divide the time-frequency graph of the selected electromyographic signal into multiple time-frequency windows, and determine the muscle state characteristics of the target patient at each time according to the density of the time-frequency points in each time-frequency window; It should be noted that the processing module 402 in the present application is further configured to identify multiple motion states of the healthy side of the target patient based on all muscle state features, and then decompose the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components, and continue to determine multiple electromyographic components of the remaining electromyographic signal; It should be noted that the processing module 402 in the present application is also used to determine the firing time series of different motor units when the target patient exercises based on the trend coefficients of all myoelectric components under the same motion state; The execution module 403 in this application is mainly used to generate multiple electrical stimulation pulse signals based on the cepstrum entropy of all emission time series, and then apply electrical stimulation to the affected side of the target patient through all the electrical stimulation pulse signals.

[0066] In addition, the present application also provides a computer device, which includes a memory and a processor, wherein the memory stores code, and the processor is configured to obtain the code and execute the above-mentioned electrical stimulation rehabilitation training method for neurosurgical care.

[0067] In some embodiments, reference Figure 5 , which is a schematic diagram of the structure of a computer device for implementing an electrical stimulation rehabilitation training method for neurosurgery nursing according to some embodiments of the present application. The electrical stimulation rehabilitation training method for neurosurgery nursing in the above embodiment can be Figure 5 The computer device 500 shown in FIG. 5 is implemented as shown in FIG. 5 . The computer device 500 includes at least one processor 501 , a communication bus 502 , a memory 503 , and at least one communication interface 504 .

[0068] The processor 501 may be a general-purpose central processing unit (CPU) or an application-specific integrated circuit (ASIC).

[0069] The communication bus 502 may be used to transmit information between the aforementioned components.

[0070] The memory 503 may be a read-only memory (ROM) or other type of static storage device capable of storing static information and instructions, a random access memory (RAM) or other type of dynamic storage device capable of storing information and instructions, an electrically erasable programmable read-only memory (EEPROM), a compact disc read-only memory (CDROM) or other optical disc storage, an optical disc storage (including compact discs, laser discs, optical discs, digital versatile discs, Blu-ray discs, etc.), a magnetic disk or other magnetic storage device, or any other medium capable of carrying or storing desired program code in the form of instructions or data structures and accessible by a computer, but is not limited thereto. The memory 503 may be independent and connected to the processor 501 via the communication bus 502. The memory 503 may also be integrated with the processor 501.

[0071] Memory 503 is used to store program code for executing the solution of the present application, and is controlled by processor 501 for execution. Processor 501 is used to execute the program code stored in memory 503. The program code may include one or more software modules. The electrical stimulation rehabilitation training method for neurosurgical nursing in the above embodiment can be implemented by processor 501 and one or more software modules in the program code in memory 503.

[0072] The communication interface 504 uses any device such as a transceiver to communicate with other devices or communication networks, such as Ethernet, radio access network (RAN), wireless local area network (WLAN), etc.

[0073] In a specific implementation, as an example, a computer device may include multiple processors, each of which may be a single-core (singleCPU) processor or a multi-core (multiCPU) processor. The processor herein may refer to one or more devices, circuits, and / or processing cores for processing data (e.g., computer program instructions).

[0074] The aforementioned computer device can be a general-purpose computer device or a dedicated computer device. In a specific implementation, the computer device can be a desktop computer, a portable computer, a network server, a personal digital assistant (PDA), a mobile phone, a tablet computer, a wireless terminal device, a communication device, or an embedded device. The embodiments of this application do not limit the type of computer device.

[0075] In addition, the present application also provides a computer-readable storage medium, which stores a computer program. When the computer program is executed by a processor, it implements the above-mentioned electrical stimulation rehabilitation training method for neurosurgical care.

[0076] In summary, in the electrical stimulation rehabilitation training system and method for neurosurgical nursing disclosed in the embodiments of the present application, first, multiple sampling points are set on the healthy side of the target patient, and the electromyographic signals at each sampling point are collected; one electromyographic signal is selected as the selected electromyographic signal, and the time-frequency diagram of the selected electromyographic signal is divided into multiple time-frequency windows, and the muscle state characteristics of the target patient at each moment are determined according to the density of the time-frequency points in each time-frequency window; based on all the muscle state characteristics, multiple motion states of the healthy side of the target patient are identified, and then the sub-signal segments corresponding to the selected electromyographic signal in each motion state are decomposed into multiple electromyographic components, and the multiple electromyographic components of the remaining electromyographic signals are further determined; the emission time series of different motion units of the target patient when exercising is determined according to the trend coefficients of all electromyographic components under the same motion state; multiple electrical stimulation pulse signals are generated based on the inverse entropy of all emission time series, and then electrical stimulation is applied to the affected side of the target patient through all the electrical stimulation pulse signals.

[0077] It can be seen from this that the present application collects the electromyographic signals of different muscles on the healthy side of the patient, and adaptively divides the movement state of the target patient's healthy side muscles by transforming the density of the time-frequency points of the electromyographic signals in different time-frequency windows, so as to quickly identify the movement state of the patient's healthy side muscles, thereby dividing the electromyographic signals of various muscles into sub-signal segments under different movement states. Subsequently, the electromyographic signals (i.e., the emission time series) emitted by the same motor unit are screened out through the trend of each sub-signal segment, and finally electrical stimulation is applied to the corresponding position of the patient's affected side according to the emission time series of each motor unit, thereby ensuring that the movement of the patient's healthy side is synchronized to the patient's affected side in real time for neural activation. In summary, the scheme of the present application can improve the synchronization between the patient's movement intention and the stimulation of the affected side.

[0078] Although the preferred embodiments of the present application have been described, those skilled in the art may make additional changes and modifications to these embodiments once they have learned the basic creative concept. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments and all changes and modifications that fall within the scope of the present application.

[0079] Obviously, those skilled in the art may make various changes and modifications to this application without departing from the spirit and scope of this application. Thus, if these modifications and variations of this application fall within the scope of the claims of this application and their equivalents, this application is intended to include these modifications and variations.

Claims

1. An electrical stimulation rehabilitation training method for neurosurgery nursing, characterized in that: include: Multiple sampling points are set on the healthy side of the target patient, and the electromyographic signals at each sampling point are collected; Selecting an electromyographic signal as a selected electromyographic signal, dividing a time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows, and determining muscle state characteristics of the target patient at each moment according to the density of time-frequency points in each time-frequency window; Identifying multiple motion states of the healthy side of the target patient based on all muscle state features, decomposing the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components, and continuing to determine multiple electromyographic components of the remaining electromyographic signal; Determine the firing time series of different motor units during the target patient's movement based on the trend coefficients of all myoelectric components under the same movement state; A plurality of electrical stimulation pulse signals are generated based on the cepstrum entropy of all the emission time series, and then electrical stimulation is applied to the affected side of the target patient through all the electrical stimulation pulse signals.

2. The method according to claim 1, wherein Dividing the time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows specifically includes: determining a time-frequency diagram of the selected electromyographic signal; Dividing the time axis of the time-frequency graph according to the time interval units to obtain a plurality of time windows; The frequency axis of each time window is divided according to a preset frequency passband range to obtain multiple time-frequency windows.

3. The method according to claim 1, wherein The muscle state characteristics of the target patient at each moment are determined based on the density of the time-frequency points in each time-frequency window, specifically including: Determine the density of time-frequency points in each time-frequency window; Determine the time-frequency distribution characteristics in each time window according to the density of time-frequency points in all time-frequency windows; Initialize the upper limit and lower limit of the time-frequency distribution; All time-frequency distribution features are dynamically normalized according to the time-frequency distribution upper limit and the time-frequency distribution lower limit, and all normalized time-frequency distribution features are used as muscle state features of the target patient at each moment.

4. The method according to claim 1, wherein Based on all muscle state characteristics, multiple motion states of the target patient's healthy side are identified, including: Preset motion state threshold; All muscle state features are sequentially compared with the motion state thresholds to obtain multiple motion states of the healthy side of the target patient.

5. The method according to claim 1, wherein Decomposing the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components specifically includes: dividing the selected electromyographic signal into a plurality of sub-signal segments according to all motion states; For each sub-signal segment, each sub-signal segment is decomposed into multiple myoelectric components according to the envelope of each sub-signal segment.

6. The method according to claim 1, wherein The specific method for determining the firing time series of different motor units during the target patient's movement based on the trend coefficients of all myoelectric components under the same movement state includes: Divide all motion states into multiple motion state clusters under the same motion state; selecting a movement state cluster as a selected movement state cluster, and determining trend coefficients of all myoelectric components under each movement state in the selected movement state cluster; determining a firing time subsequence of each motor unit according to all trend coefficients, and continuing to determine a firing time subsequence of each motor unit according to the remaining selected motion state clusters; The firing time sequence of each motor unit during the target patient's movement is reconstructed based on the firing time subsequence of each motor unit.

7. The method according to claim 1, wherein The generation of multiple electrical stimulation pulse signals based on the cepstrum entropy of all the firing time series specifically includes: Determine the cepstral entropy of each emission time series; determining a plurality of electrical stimulation parameters according to all cepstral entropies; A plurality of electrical stimulation signals are generated according to all electrical stimulation parameters.

8. An electrical stimulation rehabilitation training system for neurosurgery nursing, characterized in that: include: An acquisition module is used to set multiple sampling points on the healthy side of the target patient and collect the electromyographic signals at each sampling point; a processing module, configured to select an electromyographic signal as a selected electromyographic signal, divide a time-frequency graph of the selected electromyographic signal into a plurality of time-frequency windows, and determine muscle state characteristics of the target patient at each moment according to the density of time-frequency points in each time-frequency window; The processing module is further configured to identify multiple motion states of the healthy side of the target patient based on all muscle state features, decompose the sub-signal segments corresponding to the selected electromyographic signal in each motion state into multiple electromyographic components, and continue to determine multiple electromyographic components of the remaining electromyographic signal; The processing module is further configured to determine the firing time series of different motor units when the target patient exercises based on the trend coefficients of all myoelectric components under the same motion state; The execution module is used to generate multiple electrical stimulation pulse signals based on the cepstrum entropy of all the emission time series, and then apply electrical stimulation to the affected side of the target patient through all the electrical stimulation pulse signals.

9. A computer device, characterized in that: The computer device includes a memory and a processor, the memory stores codes, and the processor is configured to obtain the codes and execute the electrical stimulation rehabilitation training method for neurosurgical care according to any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the electrical stimulation rehabilitation training method for neurosurgery care according to any one of claims 1 to 7 is implemented.

Images