Myoelectricity data processing method, fatigue determination method, device and equipment

Abstract

The invention relates to the technical field of muscle detection, and discloses a myoelectricity data processing method, a fatigue determination method, a device and equipment, and the myoelectricity data processing method comprises the steps: collecting target power generation signals and target myoelectricity signals corresponding to a plurality of target muscles when a target user does each preset cervical vertebra motion according to a preset power generation rule at the same time; processing each target power generation signal to obtain a corresponding power generation signal feature and continuous power generation time; processing each target electromyographic signal to obtain a corresponding overall linear feature, a dynamic linear feature and a nonlinear feature; and processing the target electromyographic signals of the plurality of target muscles to obtain collaborative features of the plurality of target muscles. According to the method and the device, various different types of features related to the neck muscle fatigue can be obtained, so that a more accurate neck muscle fatigue result can be obtained subsequently according to the features.

Description

Technical Field

[0001] The present application relates to the technical field of muscle detection, and in particular to a method for processing electromyographic data, a method for determining fatigue, a device and equipment. Background Art

[0002] The cervical spine usually moves or maintains posture under the joint action of multiple cervical muscles, and cervical muscle fatigue is one of the recognized main causes of common and high-incidence joint diseases such as neck pain and cervical disc degeneration. Changes in cervical muscle activation levels and collaboration forms when fatigued will directly affect the biomechanical environment of the cervical spine, and are an important theoretical basis for studying the pathogenesis and therapeutic intervention of diseases such as neck pain and cervical disc degeneration. However, the analysis of cervical muscle fatigue in the existing technology usually only focuses on the functional characteristics of a single muscle, which will cause the analysis of cervical muscle fatigue to be imperfect and one-sided. Summary of the invention

[0003] In view of this, the embodiments of the present application provide an electromyographic data processing method, a fatigue determination method, an apparatus and a device, which can achieve the acquisition of various types of features related to cervical muscle fatigue, and then obtain more accurate cervical muscle fatigue results based on these features.

[0004] In a first aspect, an embodiment of the present application provides a method for processing electromyographic data, comprising:

[0005] At the same time, target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to the preset force rules are collected; wherein the target force signals and the target electromyographic signals correspond to each other in time;

[0006] Processing each of the target force signals to obtain corresponding force signal characteristics and continuous force time;

[0007] Processing each of the target electromyographic signals to obtain corresponding overall linear features, dynamic linear features and nonlinear features; wherein the overall linear features and the dynamic linear features are the time domain and frequency domain features of the electromyographic signals of the target muscles; and the nonlinear features are the complexity features of the muscle activity mapping of the target muscles;

[0008] Processing the target electromyographic signals of the plurality of target muscles to obtain synergistic features of the plurality of target muscles;

[0009] Among them, the force signal characteristics, the continuous force time, the overall linear characteristics, the dynamic linear characteristics, the nonlinear characteristics and the collaborative characteristics are used to jointly determine the cervical muscle fatigue results.

[0010] In some embodiments, the simultaneously collecting target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to the preset force rule includes:

[0011] Collecting corresponding force signals and corresponding electromyographic signals of a plurality of target muscles when the target user performs at least one preset cervical spine movement according to maximum force and one continuous force for a preset number of times, and generating corresponding force signal graphs and electromyographic signal graphs according to the force signals and the electromyographic signals;

[0012] Selecting a force starting point, a target force starting point, a target force ending point and a force ending point from each of the force signal graphs;

[0013] Matching the time of the electromyographic signal starting point in each electromyographic signal graph with the force starting point in the corresponding force signal graph, and determining the target electromyographic signal starting point, target electromyographic signal ending point and electromyographic signal ending point corresponding to the electromyographic signal according to the force signal sampling rate and the electromyographic signal sampling rate;

[0014] The force signal between the target force starting point and the target force ending point corresponding to the maximum force is used as a reference target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the maximum force is used as a reference target electromyographic signal;

[0015] The force signal between the target force starting point and the target force ending point corresponding to the continuous force is used as the target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the continuous force is used as the target electromyographic signal.

[0016] In some embodiments, the processing of each target force signal to obtain corresponding force signal characteristics and continuous force time includes:

[0017] Dividing the target force signal into multiple segments according to a preset equal division rule;

[0018] Calculating the mean and standard deviation of each segment of the target force signal respectively, and obtaining the coefficient of variation of the target force signal of the current segment according to the mean and standard deviation;

[0019] The continuous force exertion time of the target user when corresponding to the preset cervical spine movement is obtained according to the force signal diagram.

[0020] In some embodiments, processing each of the target electromyographic signals to obtain a corresponding overall linear characteristic includes:

[0021] Calculate the reference root mean square value, reference median frequency and reference average power frequency of the reference target electromyographic signal corresponding to each maximum force;

[0022] Calculating an average value of the reference root mean square value, an average value of the reference average median frequency, and an average value of the reference average power frequency according to the reference root mean square value, the reference median frequency, and the reference average power frequency corresponding to each of the maximum forces;

[0023] Calculate the myoelectric root mean square value, myoelectric median frequency and myoelectric average power frequency of the target myoelectric signal of the continuous force;

[0024] The myoelectric root mean square value, the myoelectric median frequency and the myoelectric average power frequency are normalized to obtain the overall linear characteristic.

[0025] In some embodiments, processing each of the target electromyographic signals to obtain a corresponding dynamic linear feature includes:

[0026] According to the preset sliding window rules, the sliding window root mean square value, sliding window median frequency and sliding window average power frequency of each target muscle in each sliding window are obtained and normalized;

[0027] According to the normalized sliding window root mean square value, the sliding window median frequency and the sliding window average power frequency of each sliding window, they are arranged in chronological order to obtain a root mean square value dynamic change graph, a median frequency dynamic change graph and an average power frequency dynamic change graph of each target muscle over time;

[0028] Performing linear fitting on the sliding window root mean square value, the sliding window median frequency and the sliding window average power frequency of each sliding window after normalization, respectively, to obtain the root mean square value slope, median frequency slope and average power frequency slope of each target muscle;

[0029] According to the electromyographic signals of each target muscle in multiple preset time periods, the time period root mean square value, time period median frequency and time period average power frequency corresponding to each preset time period are obtained, and normalized.

[0030] In some embodiments, processing each of the target electromyographic signals to obtain corresponding nonlinear features includes:

[0031] Dividing the target electromyographic signal of each target muscle into multiple segments of electromyographic signals;

[0032] The sample entropy of each segment of electromyographic signal, the percentage certainty of recursive quantitative analysis and the average self-conversion probability are obtained according to the preset calculation formula.

[0033] In some embodiments, the processing of the target electromyographic signals of the plurality of target muscles to obtain the synergistic features of the plurality of target muscles comprises:

[0034] Processing the target electromyographic signals corresponding to the target muscles by non-negative matrix decomposition until the calculated variance ratio is greater than a preset value, so as to obtain a current basis matrix and coefficient matrix;

[0035] Based on the base matrix and the coefficient matrix, the temporal similarity and the collaborative structural similarity of the target user relative to the reference matrix are calculated in combination with the cosine similarity.

[0036] In a second aspect, an embodiment of the present application provides a method for determining cervical muscle fatigue, which inputs the force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and collaborative characteristics obtained by the above-mentioned electromyographic data processing method into a pre-trained cervical muscle fatigue network model to output cervical muscle fatigue results; or

[0037] The force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and synergistic characteristics obtained by the above-mentioned electromyographic data processing method are combined with preset statistical data to determine the cervical muscle fatigue results.

[0038] In a third aspect, an embodiment of the present application provides an electromyographic data processing device, comprising:

[0039] A collection module, used for simultaneously collecting target force signals and target electromyographic signals corresponding to multiple target muscles when a target user performs each preset cervical spine movement according to a preset force rule; wherein the target force signals and the target electromyographic signals correspond to each other in time; and the target muscles are cervical spine-related muscles;

[0040] A processing module, used for processing each of the target force signals to obtain corresponding force signal characteristics and continuous force time;

[0041] The processing module is further used to process each of the target electromyographic signals to obtain corresponding overall linear characteristics, dynamic linear characteristics and nonlinear characteristics; wherein the overall linear characteristics and the dynamic linear characteristics are the time domain and frequency domain characteristics of the electromyographic signals of the target muscles; and the nonlinear characteristics are the characteristics of the muscle activity complexity of the target muscles;

[0042] The processing module is further used to process the target electromyographic signals of the plurality of target muscles to obtain the synergistic features of the plurality of target muscles;

[0043] Among them, the force signal characteristics, the continuous force time, the overall linear characteristics, the dynamic linear characteristics, the nonlinear characteristics and the collaborative characteristics are used to jointly determine the cervical muscle fatigue results.

[0044] In a fourth aspect, an embodiment of the present application provides a terminal device, comprising a processor and a memory, wherein the memory stores a computer program, and the processor is used to execute the computer program to implement the above-mentioned electromyography data processing method or the above-mentioned cervical muscle fatigue determination method.

[0045] The embodiments of the present application have the following beneficial effects: the present application simultaneously collects the surface electromyography signals of the target user, thereby inducing cervical muscle fatigue of the target user by isometric contraction, synchronously collects the force signal and the multi-channel surface electromyography signals during the fatigue process, obtains the force signal characteristics and the continuous force time according to the force signal, and obtains the overall linear characteristics, dynamic linear characteristics and nonlinear characteristics as well as the synergistic characteristics between each muscle according to the electromyography signal. The present application obtains multi-faceted and multi-dimensional muscle fatigue-related characteristics, and can be more accurate in the subsequent analysis of cervical muscle fatigue of the target user. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings required for use in the embodiments will be briefly introduced below. It should be understood that the following drawings only show certain embodiments of the present application and therefore should not be regarded as limiting the scope. For ordinary technicians in this field, other related drawings can be obtained based on these drawings without paying creative work.

[0047] Figure 1 A first flow chart of the electromyographic data processing method according to an embodiment of the present application is shown;

[0048] Figure 2 A second flow chart of the method for processing electromyographic data according to an embodiment of the present application is shown;

[0049] Figure 3 A force signal diagram of an embodiment of the present application is shown;

[0050] Figure 4 A first electromyographic signal diagram of an embodiment of the present application is shown;

[0051] Figure 5 A second electromyographic signal diagram of an embodiment of the present application is shown;

[0052] Figure 6 A third flow chart of the electromyographic data processing method according to an embodiment of the present application is shown;

[0053] Figure 7A fourth flow chart of the electromyographic data processing method according to an embodiment of the present application is shown;

[0054] Figure 8 A fifth flow chart of the method for processing electromyographic data according to an embodiment of the present application is shown;

[0055] Fig. 9 A sixth flow chart of the electromyographic data processing method according to an embodiment of the present application is shown;

[0056] Fig.10 A structural schematic diagram of the electromyographic data processing device according to an embodiment of the present application is shown. DETAILED DESCRIPTION

[0057] The technical solutions in the embodiments of the present application will be described clearly and completely below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments.

[0058] The components of the embodiments of the present application generally described and shown in the drawings herein may be arranged and designed in various configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.

[0059] Hereinafter, the terms "including", "having" and their cognates that can be used in various embodiments of the present application are intended only to indicate specific features, numbers, steps, operations, elements, components or a combination of the foregoing items, and should not be understood as first excluding the existence of one or more other features, numbers, steps, operations, elements, components or a combination of the foregoing items or increasing the possibility of one or more features, numbers, steps, operations, elements, components or a combination of the foregoing items. In addition, the terms "first", "second", "third" and the like are only used to distinguish descriptions and cannot be understood as indicating or implying relative importance.

[0060] Unless otherwise defined, all terms (including technical terms and scientific terms) used herein have the same meanings as those generally understood by those skilled in the art to which the various embodiments of the present application belong. The terms (such as those defined in generally used dictionaries) will be interpreted as having the same meanings as the contextual meanings in the relevant technical field and will not be interpreted as having idealized meanings or overly formal meanings unless clearly defined in the various embodiments of the present application.

[0061] In conjunction with the accompanying drawings, some embodiments of the present application are described in detail below. In the absence of conflict, the following embodiments and features in the embodiments can be combined with each other.

[0062] In the prior art, the functional characteristic analysis of cervical muscle fatigue usually focuses on a single muscle, which is difficult to intuitively reflect the functional changes and collaborative abnormalities of multiple cervical muscles in the process of joint activation fatigue, and the statistical significance of the changes in muscle activation characteristics is not obvious, and it is difficult to quantitatively analyze the abnormal indicators with significant differences, which is not conducive to a comprehensive analysis of the biomechanical characteristics of cervical muscle fatigue. Based on this, the present application proposes a myoelectric data processing method, a fatigue determination method, a device and an equipment, by using multiple channels to simultaneously collect the myoelectric signals and force signals of multiple target muscles when the target user performs a preset cervical action each time, and after obtaining the myoelectric signals and force signals, the myoelectric signals and force signals are processed to obtain the force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and collaborative characteristics and other multi-faceted and multi-type characteristics, and then through the acquisition of all-round force characteristics and myoelectric characteristics, it can meet the subsequent accurate analysis of cervical muscle fatigue. The following is a description of the electromyographic data processing method in conjunction with some specific embodiments.

[0063] Figure 1 A schematic flow chart of the electromyographic data processing method of an embodiment of the present application is shown. Exemplarily, the electromyographic data processing method includes the following steps:

[0064] Step S100, simultaneously collecting target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to the preset force rule.

[0065] Among them, the target force signal and the target electromyographic signal correspond in time; the target muscles are the cervical vertebrae-associated muscles.

[0066] Exemplarily, the preset force rule stipulates that the target user performs a preset cervical spine movement according to the maximum force requirement and the continuous force requirement, wherein the continuous force is usually 50% of the maximum force. In this embodiment, the target force signals and target electromyographic signals corresponding to multiple target muscles when the same target user performs each preset cervical spine movement according to the maximum force requirement are collected three times. By collecting the force signals and electromyographic signals under the maximum force conditions three times, random errors can be reduced, and the subsequent normalization of the continuous force signal can be more accurate.

[0067] The simultaneous collection mentioned in this step means that when the target user performs a preset cervical spine movement, in order to make the subsequently obtained electromyographic characteristics and force characteristics more comprehensive, it is necessary to collect the force signals and electromyographic signals of multiple target muscles. Therefore, in this embodiment, the force signals and electromyographic signals of multiple target muscles are collected simultaneously through multiple channels when the target user performs the preset cervical spine movement. It can be understood that the number of collection channels is equal to the number of target muscles. Through this multi-channel simultaneous collection method, the signals of multiple target muscles under the same circumstances can be collected.

[0068] Among them, the preset cervical spine movements include but are not limited to craniocervical extension CCE, craniocervical flexion CCF, cervicothoracic extension CTE and cervicothoracic flexion CTF, and the target muscles include but are not limited to the left / right sternocleidomastoid muscle, left / right upper trapezius muscle, left / right semispinalis capitis muscle, left / right splenius capitis muscle, etc.

[0069] In some embodiments, Figure 2 The step S100 shown includes steps S110 to S150:

[0070] Step S110, collecting corresponding force signals and corresponding electromyographic signals of multiple target muscles when the target user performs at least one preset cervical spine movement according to maximum force and one continuous force for a preset number of times, and generating corresponding force signal graphs and electromyographic signal graphs based on the force signals and electromyographic signals.

[0071] In this step, it is necessary to collect the corresponding force signals of multiple target muscles when the target user performs each preset cervical spine movement under the maximum force for a preset number of times through multiple channels. It can be understood that the target user obtains a set of force signals corresponding to each target muscle each time each preset cervical spine movement is collected; in addition, in this step, it is also necessary to collect the corresponding force signals of multiple target muscles when the target user performs each preset cervical spine movement under the maximum force for a preset number of times through multiple channels. Among them, in order to make the subsequent normalized electromyographic characteristics data more accurate, it is necessary to collect force signals and electromyographic signals of the target user under multiple maximum force conditions for the same preset cervical spine movement in this embodiment, and the preset number of times is usually 3 times.

[0072] In addition, in this step, it is necessary to collect the corresponding force signals of multiple target muscles when the target user performs each preset cervical spine movement under continuous force, and the corresponding electromyographic signals of multiple target muscles when the target user performs each preset cervical spine movement under continuous force through multiple channels.

[0073] Import the force signals obtained under each maximum force condition and the force signals under continuous force condition into the Matlab platform to obtain the following Figure 3The force signal diagram corresponding to each group of force signals shown in FIG. 1 is imported into the Matlab platform to obtain the force signal under each maximum force condition and the electromyographic signal under continuous force condition. Figure 4 The electromyographic signal graph corresponding to each group of electromyographic signals shown.

[0074] Step S120, selecting a force starting point, a target force starting point, a target force ending point and a force ending point from each force signal diagram.

[0075] Step S130, match the time of the electromyographic signal starting point in each electromyographic signal graph with the force starting point in the corresponding force signal graph, and determine the target electromyographic signal starting point, target electromyographic signal ending point and electromyographic signal ending point of the corresponding electromyographic signal according to the force signal sampling rate and the electromyographic signal sampling rate.

[0076] Step S140, taking the force signal between the target force starting point and the target force ending point corresponding to the maximum force as the reference target force signal, and taking the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the maximum force as the reference target electromyographic signal.

[0077] Step S150, taking the force signal between the target force start point and the target force end point corresponding to the continuous force as the target force signal, and taking the electromyographic signal between the target electromyographic signal start point and the target electromyographic signal end point corresponding to the continuous force as the target electromyographic signal.

[0078] Exemplarily, since the electromyographic signal and the force signal are collected separately, there will inevitably be a time difference in the collection of the two signals. Therefore, in order to ensure the accuracy of the target electromyographic signal intercepted subsequently, it is necessary to first match the two signals in time.

[0079] According to the characteristics of the force signal, Figure 3 From the force signal diagram in , we can draw four points: the transition from rest to force, the force reaching the target force, the target force starting to decrease, and the transition from force to rest. These four points can be respectively recorded as the force start point (force start, FS), the target force start point (target start, TS), the target force end point (target end, TE), and the force end point (force end, FE). Figure 3 Points a, b, c and d in the figure correspond to the force start point, target force start point, target force end point and force end point respectively. In this embodiment, the force signal and the electromyographic signal are matched by the inflection point, where the "inflection point" is an obvious fluctuation when the electromyographic signal changes from the resting state to the force or when the force ends and turns to the resting state. It can be easily found after the electromyographic signal is visualized, such as Figure 4The electromyographic signal shown in the figure, Figure 4 Point e and point f are the inflection points. The "inflection point" of the electromyographic signal is made to correspond to the force start point and (or) force end point of the force measurement signal, thereby eliminating the time difference caused by signal acquisition. That is, the relative position relationship of the four points, namely the force start point, the target force start point, the target force end point and the force end point in the above-mentioned force signal, is applied to the electromyographic signal, and the target electromyographic signal corresponding to the target force signal is intercepted with the "inflection point" as a reference.

[0080] Due to the different sampling rates of the force signal and the electromyographic signal, the length of the time period between every two data points of the force signal and the electromyographic signal is also inconsistent. It is necessary to calculate the value of the above relative position relationship on the electromyographic signal based on the sampling rate multiple relationship. The multiple is the value obtained by dividing the sampling frequency of the electromyographic signal by the sampling frequency of the force signal. For example, if the acquisition frequency of the force signal is 12.5Hz and the acquisition frequency of the electromyographic signal is 1500Hz, then the multiple M=1500 / 12.5=120, that is, the time period between two adjacent points of the force signal corresponds to the time period between 121 points in the electromyographic signal (120 electromyographic signal time periods). Based on this, the relative positions between the four points of the electromyographic signal can be calculated.

[0081] Since the starting point of the power generation signal can be determined from the figure, the inflection point when the electromyographic signal starts can also be determined, then the three points of the electromyographic signal relative to the target power start point, target power end point and power end point can be determined according to the following formula: sEMG n -sEMG FS =(F n -F FS )×x; in the formula, n is TS, TE, FE, and x is a multiple of the sampling frequency of the electromyographic signal and the force signal. After obtaining the relative positions of the force starting point, target force starting point, target force ending point, and force ending point of the electromyographic signal corresponding to the force signal, the sEMG FS to sEMG FE The force window (force) and sEMG TS to sEMG TE The target window is defined between the two windows, and the relative positions of the two windows remain unchanged (e.g. Figure 4 ), in the electromyographic signal graph, slide the force window until the start of the force window corresponds to the "inflection point" of the electromyographic signal (make the start and end points of the force window at the "inflection point" in the graph, such as Figure 3 ), then the target electromyographic signal can be determined, and the target electromyographic signal is Figure 5 sEMG TS to sEMG TE The signal between Figure 5 light grey part in the middle).

[0082] The above-mentioned acquisition of target electromyographic signals and target force signals is universal for signals of maximum force and signals of continuous force. In this embodiment, the force signal between the target force starting point and the target force ending point corresponding to the maximum force is used as a reference target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the maximum force is used as a reference target electromyographic signal; the force signal between the target force starting point and the target force ending point corresponding to continuous force is used as the target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to continuous force is used as the target electromyographic signal.

[0083] Step S200, processing each target force signal to obtain corresponding force signal characteristics and continuous force time.

[0084] When the muscle is continuously exerting force, the force will fluctuate around the target force (i.e., the force specified by the continuous force). In this embodiment, the fluctuation of force is judged by the mean, standard deviation, and coefficient of variation, thereby reflecting the stability of muscle control. In addition, changes in muscle fatigue characteristics will affect the total time of continuous muscle force. In order to analyze the overall fatigue performance of neck-related muscles in different preset spinal movements, the continuous force time of each movement of the target user needs to be determined in this embodiment.

[0085] For example, Figure 6 As shown, step S200 includes steps S210 to S230:

[0086] Step S210, dividing the target force signal into multiple segments according to a preset equal division rule.

[0087] Step S220, respectively calculating the mean and standard deviation of each segment of the target force signal, and obtaining the coefficient of variation of the current segment of the target force signal according to the mean and standard deviation;

[0088] Step S230, obtaining the continuous force application time of the target user when the target user performs the corresponding preset cervical spine movement according to the force application signal diagram.

[0089] It can be understood that in this embodiment, the target force signal in the continuous force process when the target user performs each preset cervical spine movement is divided into multiple uniform segments according to the time sequence, and the force signal in each segment must meet the preset length to ensure that data analysis can be performed. Then the mean and standard deviation of each force signal are calculated respectively, and then the coefficient of difference is calculated according to the mean and standard deviation corresponding to each force signal, where the formula for calculating the coefficient of difference is: CV = A / B × 100%, where CV is the coefficient of difference, A is the standard deviation, and B is the mean.

[0090] Step S300, processing each target electromyographic signal to obtain corresponding overall linear characteristics, dynamic linear characteristics and nonlinear characteristics.

[0091] Among them, the overall linear features and dynamic linear features are the time domain and frequency domain features of the target muscle electromyographic signal; the nonlinear features are the complex features of the muscle activity of the mapped target muscle. The linear features include but are not limited to the root mean square value (RMS) of the time domain features, the median frequency (MF) and the mean power frequency (MPF) of the frequency domain features, and the nonlinear features include but are not limited to the sample entropy (SampEn), the percentage certainty of the recursive quantization analysis (RQA), the average self-transition probability, the average clustering coefficient, and modularity.

[0092] For example, Figure 7 As shown, processing each target electromyographic signal to obtain the corresponding overall linear feature includes steps S310 to S340:

[0093] Step S310, calculating the reference root mean square value, reference median frequency and reference average power frequency of the reference target electromyographic signal corresponding to each maximum force.

[0094] Step S320, calculating the average value of the reference root mean square value, the average value of the reference average median frequency and the average value of the reference average power frequency according to the reference root mean square value, the reference median frequency and the reference average power frequency corresponding to each maximum force.

[0095] Step S330, calculating the myoelectric root mean square value, myoelectric median frequency and myoelectric average power frequency of the target myoelectric signal of continuous force.

[0096] Step S340, normalizing the EMG root mean square value, the EMG median frequency and the EMG average power frequency to obtain an overall linear characteristic.

[0097] Due to the differences in muscle content and muscle characteristics among users, the signal amplitude and frequency characteristics in the electromyographic signals of each target user are also different. Therefore, in this embodiment, the time and frequency domain characteristics during the maximum force are used as the denominator for normalization to eliminate individual differences, so that the analyzed characteristics are all expressed as "% maximum characteristics". Therefore, in this embodiment, when obtaining the overall linear characteristics, it is first necessary to determine the parameters used for normalization. First, the root mean square value, median frequency and average power frequency at each maximum force are calculated as the reference root mean square value, reference median frequency and reference average power frequency, respectively. In order to ensure the accuracy of the characteristics, this embodiment requires multiple electromyographic signals of the target user under maximum force, and calculates the average value of the multiple maximum force reference root mean square values, reference median frequencies and reference average power frequencies, so as to obtain the average value of the reference root mean square value, the average value of the reference average median frequency and the average value of the reference average power frequency. Then the EMG root mean square value, EMG median frequency and EMG average power frequency of the target EMG signal under continuous force are calculated; when normalizing the EMG root mean square value, EMG median frequency and EMG average power frequency of the target EMG signal, the corresponding EMG root mean square value, EMG median frequency and EMG average power frequency in the continuous force are normalized with the average value of the reference root mean square value, the average value of the reference average median frequency and the average value of the reference average power frequency as denominators, respectively, to obtain the overall linear characteristics. It can be understood that the overall linear characteristics are the normalized characteristics of the EMG root mean square value, the EMG median frequency and the EMG average power frequency.

[0098] For example, Figure 8 As shown, each target electromyographic signal is processed to obtain a corresponding dynamic linear feature, including steps S350 to S380:

[0099] Step S350, according to the preset sliding window rule, the sliding window root mean square value, the sliding window median frequency and the sliding window average power frequency of each target muscle in each sliding window are obtained, and after normalization processing, the normalized sliding window root mean square value, the normalized sliding window median frequency and the normalized sliding window average power frequency after normalization processing in each sliding window are obtained.

[0100] Step S360, according to the normalized sliding window root mean square value, the normalized sliding window median frequency and the normalized sliding window average power frequency of each sliding window, they are arranged in chronological order to obtain a dynamic change graph of the root mean square value, the median frequency and the average power frequency of each target muscle over time.

[0101] It can be understood that in order to intuitively represent the changes in muscle time domain and frequency domain characteristics during continuous exertion in the present embodiment, the idea of ​​sliding window is adopted, and the sliding window method with a design window length of 500 and a moving step length of 250 is used to calculate the root mean square value (RMS), median frequency (MF), and mean power frequency (MPF) of the electromyographic signal of each channel (each target muscle) in the window, and normalize it to obtain the normalized sliding window root mean square value, normalized sliding window median frequency, and normalized sliding window average power frequency after normalization in each section of sliding window of each target muscle; After the eigenvalues ​​normalized in each section of sliding window of each target muscle obtained, it is marked in the figure in chronological order, and the changes of electromyographic signal characteristics over time can be intuitively reflected. Wherein, the corresponding sliding window root mean square value, the sliding window median frequency, and the sliding window average power frequency can be normalized according to the average value of the corresponding reference root mean square value, the average value of the reference average median frequency, and the average value of the reference average power frequency.

[0102] Step S370, linear fitting is performed on the normalized sliding window root mean square value, the normalized sliding window median frequency and the normalized sliding window average power frequency of each sliding window, respectively, to obtain the root mean square value slope, median frequency slope and average power frequency slope of each target muscle.

[0103] Muscle fatigue can be reflected as a linear feature that changes over time (such as an increase in time domain features and a decrease in frequency domain features). Therefore, the slope obtained by linear fitting of time domain and frequency domain features can be used to reflect the speed of overall muscle changes. In order to analyze the overall speed differences in fatigue evolution of target users, the normalized sliding window root mean square value, normalized sliding window median frequency, and normalized sliding window average power frequency obtained after the above normalization are fitted by a first-order linear fit on the original time series or normalized time series (displaying the time series in percentages, starting at 0% and ending at 100%), and the root mean square slope, median frequency slope, and average power frequency slope corresponding to each target muscle can be obtained.

[0104] Step S380, based on the electromyographic signals of each target muscle in multiple preset time periods, the time period root mean square value, time period median frequency and time period average power frequency corresponding to each preset time period are obtained, and after normalization processing, the normalized time period root mean square value, normalized time period median frequency and normalized time period average power frequency of each preset time period are obtained.

[0105] In addition to the above-mentioned trend of changes and dynamic changes of the overall characteristics of each muscle over time, in order to ensure the comprehensiveness of the subsequent analysis of the cervical muscles, this embodiment also needs to quantify the muscle characteristic trend. In this embodiment, the signal of a specific time position is extracted from the target electromyographic signal of continuous force in a preset manner, and the difference between these time position signals is analyzed to quantify the difference in the evolution trend. Specifically, the electromyographic signal of continuous force is divided into n stages, and a data length of 1s (1500 data) is taken after the start and before the end of the signal, and a data length of 0.5s (750 data) is taken before and after the middle n-1 segmentation points to form a 1s data window, and the root mean square value (RMS), median frequency (MF), and mean power frequency (MPF) of these n+1 1s signal segments are calculated respectively, and then the root mean square value, median frequency of each time period, and average power frequency of each time period are normalized to obtain the normalized time period root mean square value, normalized time period median frequency, and normalized time period average power frequency of each preset time period.

[0106] In some embodiments, each target electromyographic signal is processed to obtain corresponding nonlinear features, including: dividing the target electromyographic signal of each target muscle into multiple segments of electromyographic signals; obtaining the sample entropy, percentage certainty of recursive quantization analysis and average self-transition probability of each segment of the electromyographic signal according to a preset calculation formula.

[0107] The nonlinear characteristics of EMG signals refer to the nonlinear behaviors or complex patterns exhibited in EMG signals. These nonlinear characteristics reflect the complexity of the EMG signal segment (the higher the complexity, the smaller the above eigenvalues, and vice versa). In order to analyze the differences in the complexity of EMG signals during the continuous exertion of different target users, the EMG signals of each target muscle are divided into n equal segments, and the sample entropy (SampEn), the percentage certainty of the recursive quantization analysis (RQA), the average self-conversion probability, etc. of each segment of the EMG signal are calculated according to the formula of each feature.

[0108] In this embodiment, the sample entropy can be calculated using the following formula:

[0109] In the formula, since the EMG signal is divided into n uniform segments when calculating the sample entropy, N is the number of data points in each EMG signal segment, m is the length of the template vector (the template vector refers to the subsequence extracted from the time series data) (taken as 2), r is the similarity capacity of each EMG signal segment (usually taken as 0.2 times the standard deviation of the EMG signal), and A i It takes 1 in similar cases and 0 in opposite cases. A is the total number of matches of the i-th template vector with length m+1 data points, and B i and B is the number of matches for the i-th template vector of length m data points.

[0110] After calculating the sample entropy using the above formula, the smaller the value of sample entropy, the more regular the signal, and the larger the value of sample entropy, the more complex the signal.

[0111] In this embodiment, the regularity of the electromyographic signal is determined by the percentage certainty of the recursive quantitative analysis, wherein the percentage of the recursive quantitative analysis can be obtained by the following formula:

[0112] When calculating the percentage of recursive quantification analysis, x i and x j The recursive matrix R is constructed for the states of two different points of each EMG signal. ij :

[0113] r is the similarity capacity of each EMG signal segment (usually 0.2 times the signal standard deviation), and all consecutive diagonal segments (length ≥ l) are found in the diagonal of the matrix. min is a continuous line segment, l min Take 2), P(l) is the number of line segments parallel to the main diagonal and of length l.

[0114] In order to evaluate the stability of the muscle in a certain characteristic state (such as low fatigue, medium fatigue or high fatigue), this embodiment also needs to calculate the self-conversion probability of the electromyographic signal. The formula for calculating the self-conversion probability is:

[0115]

[0116] In this embodiment, each segment of the electromyographic signal is divided into x discrete intervals according to the signal amplitude (i.e., the amplitude from sEMG min to sEMG max is divided into x parts), each interval is defined as a state, each point of the original signal is converted into the corresponding discrete interval sequence number (1 to x) to obtain a state sequence, and the number of times the signal transfers from state i to state j is counted and expressed as N ij Indicates that N ij Normalization is performed to obtain the transition probability matrix N (the sum of each row of the N matrix is ​​1), then the self-conversion probability is the diagonal element of the transition matrix (N 11 、N 22 、N 33 、···、N xx ) is the mean of the sum of .

[0117] Step S400: Process the target electromyographic signals of the multiple target muscles to obtain the synergistic features of the multiple target muscles.

[0118] When the cervical spine moves, there will be a synergistic effect between the neck muscles. This pattern may cause the same fatigue evolution pattern between the synergistic muscles. Therefore, in this embodiment, non-negative matrix decomposition is used to obtain similar change trends (synergy) of certain muscles and their participation in the synergy, thereby quantifying the synergy pattern between muscles.

[0119] In some embodiments, Fig. 9 As shown, step S400 includes:

[0120] Step S410, processing multiple target electromyographic signals corresponding to multiple target muscles by non-negative matrix decomposition until the calculated variance ratio is greater than a preset value, so as to obtain the current basis matrix and coefficient matrix.

[0121] Step S420 : Based on the base matrix and the coefficient matrix, the temporal similarity and the collaborative structural similarity of the target user relative to the reference matrix are calculated in combination with the cosine similarity.

[0122] Demonstratively, the electromyographic signal of continuous force is imported into matlab in matrix form and the absolute value is taken to obtain a non-negative matrix of data length × 8, and then the matrix decomposition order is gradually increased from the 1st order until the calculated VAF (variance ratio) is greater than 0.95, and a base matrix W and a coefficient matrix H are obtained. The coefficient matrix is ​​the number of synergies in the action and the contribution of each muscle in the synergy. Among them, the reference matrix is ​​the base matrix and coefficient matrix of a determined healthy user. In this embodiment, the cosine similarity is used to calculate the similarity of the time change of muscle synergy (base matrix W) of each target user relative to the reference matrix and the similarity of the overall synergy structure (coefficient matrix H).

[0123] The present application simultaneously collects the surface electromyography signals of the target user through multiple channels, so as to induce cervical muscle fatigue of the target user by isometric contraction, synchronously collects the force signal and multi-channel surface electromyography signals during the fatigue process, and obtains the force signal characteristics and continuous force time according to the force signal, and obtains the overall linear characteristics, dynamic linear characteristics and nonlinear characteristics as well as the synergistic characteristics between each muscle according to the electromyography signal. The present application obtains multi-faceted and multi-dimensional muscle fatigue-related characteristics, so that the subsequent analysis of cervical muscle fatigue of the target user can be more accurate.

[0124] The present application also provides a method for determining cervical muscle fatigue, which outputs cervical muscle fatigue results by inputting the force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and synergistic characteristics obtained by the myoelectric data processing method of the above embodiment into a pre-trained cervical muscle fatigue network model;

[0125] When obtaining the output cervical muscle fatigue results by pre-training the cervical muscle fatigue network model, it is necessary to first obtain the cervical muscle fatigue network model through training. The training data used in the training includes the electromyographic signal data and force signal data of the cervical muscles of multiple healthy subjects, and the electromyographic signal data and force signal data of the cervical muscles of multiple subjects with various cervical muscle degeneration. The electromyographic signals and force signals in the training data are processed in the manner described in the above embodiment. The basic model for training to obtain the cervical muscle fatigue network model can be a neural network model in the prior art.

[0126] The present application also proposes a method for determining cervical muscle fatigue. The force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and synergistic characteristics obtained by the electromyography data processing method of the above embodiment of this embodiment are combined with preset statistical data to determine the cervical muscle fatigue results.

[0127] If the cervical muscle fatigue result of the target user is determined based on preset statistical data, the relevant characteristics of the electromyographic signals and force signals of multiple healthy subjects and subjects with various cervical degeneration are obtained in the same manner as in the above embodiment, and then the differences in characteristics between healthy people and subjects with different cervical degeneration are analyzed statistically. These muscle fatigue characteristics with statistical differences constitute risk factors that affect cervical degeneration. For the subject, if the above-mentioned risk factor characteristics extracted during the fatigue experiment are inconsistent with those of the healthy group, it indicates that the muscle fatigue characteristics of the subject are abnormal, and the subject can be considered to be at risk of cervical degeneration. By comparing with the healthy group as the standard, the risk level of the subject can be warned, and various cervical muscle fatigue statistical data can be obtained. After obtaining the muscle characteristics of the target user, the cervical muscle fatigue result of the target user can be determined according to the obtained cervical muscle fatigue statistical data.

[0128] Fig.10 A schematic diagram of the structure of the electromyographic data processing device of an embodiment of the present application is shown. Exemplarily, the electromyographic data processing device includes:

[0129] The acquisition module 100 is used to simultaneously collect target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to preset force rules; wherein the target force signals and target electromyographic signals correspond in time; and the target muscles are cervical spine-associated muscles.

[0130] The processing module 200 is used to process each target force signal to obtain corresponding force signal characteristics and continuous force time.

[0131] The processing module 200 is also used to process each target electromyographic signal to obtain corresponding overall linear characteristics, dynamic linear characteristics and nonlinear characteristics; wherein the overall linear characteristics and dynamic linear characteristics are the time domain and frequency domain characteristics of the target muscle electromyographic signal; and the nonlinear characteristics are the muscle activity complexity characteristics that map the target muscle.

[0132] The processing module 200 is further used to process the target electromyographic signals of the multiple target muscles to obtain the synergistic features of the multiple target muscles.

[0133] It can be understood that the device of this embodiment corresponds to the electromyographic data processing method of the above embodiment, and the options in the above embodiment are also applicable to this embodiment, so they will not be described repeatedly here.

[0134] The present application also provides a terminal device. Exemplarily, the terminal device includes a processor and a memory, wherein the memory stores a computer program, and the processor runs the computer program to enable the terminal device to execute the above-mentioned electromyography data processing method or the above-mentioned cervical muscle fatigue determination method.

[0135] Among them, the processor can be an integrated circuit chip with signal processing capabilities. The processor can be a general-purpose processor, including a central processing unit (CPU), a graphics processing unit (GPU) and a network processor (NP), a digital signal processor (DSP), an application-specific integrated circuit (ASIC), a field programmable gate array (FPGA) or at least one of other programmable logic devices, discrete gates or transistor logic devices, and discrete hardware components. The general-purpose processor can be a microprocessor or the processor can also be any conventional processor, etc., which can implement or execute the disclosed methods, steps and logic block diagrams in the embodiments of the present application.

[0136] The memory may be, but is not limited to, a random access memory (RAM), a read only memory (ROM), a programmable read-only memory (PROM), an erasable programmable read-only memory (EPROM), an electrically erasable programmable read-only memory (EEPROM), etc. The memory is used to store a computer program, and the processor may execute the computer program accordingly after receiving an execution instruction.

[0137] The present application also provides a computer-readable storage medium for storing the computer program used in the above-mentioned terminal device. For example, the computer-readable storage medium may include, but is not limited to, various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0138] In several embodiments provided in the present application, it should be understood that the disclosed devices and methods can also be implemented in other ways. The device embodiments described above are merely schematic. For example, the flowcharts and structure diagrams in the accompanying drawings show the possible architecture, functions and operations of the devices, methods and computer program products according to multiple embodiments of the present application. In this regard, each box in the flowchart or block diagram can represent a module, a program segment or a part of a code, and the module, a program segment or a part of a code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in an alternative implementation, the functions marked in the box can also occur in a different order from the order marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the structure diagram and / or the flow diagram, and the combination of boxes in the structure diagram and / or the flow diagram, can be implemented with a dedicated hardware-based system that performs a specified function or action, or can be implemented with a combination of dedicated hardware and computer instructions.

[0139] In addition, the functional modules or units in the various embodiments of the present application may be integrated together to form an independent part, or each module may exist separately, or two or more modules may be integrated to form an independent part.

[0140] If the functions are implemented in the form of software function modules and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present application, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product, which is stored in a storage medium and includes several instructions for a computer device (which can be a smart phone, a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present application.

[0141] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.

Claims

1. A method for processing electromyographic data, characterized in that: include: At the same time, target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to the preset force rules are collected; wherein the target force signals and the target electromyographic signals correspond to each other in time; Processing each of the target force signals to obtain corresponding force signal characteristics and continuous force time; Processing each of the target electromyographic signals to obtain corresponding overall linear features, dynamic linear features and nonlinear features; wherein the overall linear features and the dynamic linear features are the time domain and frequency domain features of the electromyographic signals of the target muscles; and the nonlinear features are the complexity features of the muscle activity mapping of the target muscles; Processing the target electromyographic signals of the plurality of target muscles to obtain synergistic features of the plurality of target muscles; Among them, the force signal characteristics, the continuous force time, the overall linear characteristics, the dynamic linear characteristics, the nonlinear characteristics and the collaborative characteristics are used to jointly determine the cervical muscle fatigue results.

2. The method for processing electromyographic data according to claim 1, characterized in that: The simultaneously collecting target force signals and target electromyographic signals corresponding to multiple target muscles when the target user performs each preset cervical spine movement according to the preset force rule includes: Collecting corresponding force signals and corresponding electromyographic signals of a plurality of target muscles when the target user performs at least one preset cervical spine movement according to maximum force and one continuous force for a preset number of times, and generating corresponding force signal graphs and electromyographic signal graphs according to the force signals and the electromyographic signals; Selecting a force starting point, a target force starting point, a target force ending point and a force ending point from each of the force signal graphs; Matching the time of the electromyographic signal starting point in each electromyographic signal graph with the force starting point in the corresponding force signal graph, and determining the target electromyographic signal starting point, target electromyographic signal ending point and electromyographic signal ending point corresponding to the electromyographic signal according to the force signal sampling rate and the electromyographic signal sampling rate; The force signal between the target force starting point and the target force ending point corresponding to the maximum force is used as a reference target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the maximum force is used as a reference target electromyographic signal; The force signal between the target force starting point and the target force ending point corresponding to the continuous force is used as the target force signal, and the electromyographic signal between the target electromyographic signal starting point and the target electromyographic signal ending point corresponding to the continuous force is used as the target electromyographic signal.

3. The method for processing electromyographic data according to claim 2, characterized in that: The processing of each target force signal to obtain corresponding force signal characteristics and continuous force time includes: Dividing the target force signal into multiple segments according to a preset equal division rule; Calculating the mean and standard deviation of each segment of the target force signal respectively, and obtaining the coefficient of variation of the target force signal of the current segment according to the mean and standard deviation; The continuous force exertion time of the target user when corresponding to the preset cervical spine movement is obtained according to the force signal diagram.

4. The method for processing electromyographic data according to claim 2, characterized in that: The step of processing each of the target electromyographic signals to obtain a corresponding overall linear characteristic includes: Calculate the reference root mean square value, reference median frequency and reference average power frequency of the reference target electromyographic signal corresponding to each maximum force; Calculating an average value of reference root mean square values, an average value of reference average median frequencies, and an average value of reference average power frequencies according to the reference root mean square values, the reference median frequencies, and the reference average power frequencies corresponding to each of the maximum forces; Calculate the myoelectric root mean square value, myoelectric median frequency and myoelectric average power frequency of the target myoelectric signal of the continuous force; The myoelectric root mean square value, the myoelectric median frequency and the myoelectric average power frequency are normalized to obtain the overall linear characteristic.

5. The method for processing electromyographic data according to claim 2, characterized in that: The step of processing each of the target electromyographic signals to obtain a corresponding dynamic linear feature includes: According to the preset sliding window rules, the sliding window root mean square value, sliding window median frequency and sliding window average power frequency of each target muscle in each sliding window are obtained and normalized; According to the normalized sliding window root mean square value, the sliding window median frequency and the sliding window average power frequency of each sliding window, they are arranged in chronological order to obtain a root mean square value dynamic change graph, a median frequency dynamic change graph and an average power frequency dynamic change graph of each target muscle over time; Performing linear fitting on the sliding window root mean square value, the sliding window median frequency and the sliding window average power frequency of each sliding window after normalization, respectively, to obtain the root mean square value slope, median frequency slope and average power frequency slope of each target muscle; According to the electromyographic signals of each target muscle in multiple preset time periods, the time period root mean square value, time period median frequency and time period average power frequency corresponding to each preset time period are obtained, and normalized.

6. The method for processing electromyographic data according to claim 2, characterized in that: The step of processing each of the target electromyographic signals to obtain corresponding nonlinear features includes: Dividing the target electromyographic signal of each target muscle into multiple segments of electromyographic signals; The sample entropy of each segment of electromyographic signal, the percentage certainty of recursive quantitative analysis and the average self-conversion probability are obtained according to the preset calculation formula.

7. The method for processing electromyographic data according to claim 1, characterized in that: The step of processing the target electromyographic signals of the plurality of target muscles to obtain the synergistic features of the plurality of target muscles comprises: Processing the target electromyographic signals corresponding to the target muscles by non-negative matrix decomposition until the calculated variance ratio is greater than a preset value, so as to obtain a current basis matrix and coefficient matrix; Based on the base matrix and the coefficient matrix, the temporal similarity and the collaborative structural similarity of the target user relative to the reference matrix are calculated in combination with the cosine similarity.

8. A method for determining cervical muscle fatigue, characterized in that: Input the force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and synergistic characteristics obtained by the electromyographic signal processing method described in any one of claims 1 to 7 into a pre-trained cervical muscle fatigue network model to output cervical muscle fatigue results; or The force signal characteristics, continuous force time, overall linear characteristics, dynamic linear characteristics, nonlinear characteristics and synergistic characteristics obtained by the electromyographic signal processing method according to any one of claims 1-7 are combined with preset statistical data to determine the cervical muscle fatigue results.

9. A myoelectric data processing device, characterized in that: include: A collection module, used for simultaneously collecting target force signals and target electromyographic signals corresponding to multiple target muscles when a target user performs each preset cervical spine movement according to a preset force rule; wherein the target force signals and the target electromyographic signals correspond to each other in time; and the target muscles are cervical spine-related muscles; A processing module, used for processing each of the target force signals to obtain corresponding force signal characteristics and continuous force time; The processing module is further used to process each of the target electromyographic signals to obtain corresponding overall linear characteristics, dynamic linear characteristics and nonlinear characteristics; wherein the overall linear characteristics and the dynamic linear characteristics are the time domain and frequency domain characteristics of the electromyographic signals of the target muscles; and the nonlinear characteristics are the characteristics of the muscle activity complexity of the target muscles; The processing module is further used to process the target electromyographic signals of the plurality of target muscles to obtain the synergistic features of the plurality of target muscles; Among them, the force signal characteristics, the continuous force time, the overall linear characteristics, the dynamic linear characteristics, the nonlinear characteristics and the collaborative characteristics are used to jointly determine the cervical muscle fatigue results.

10. A terminal device, characterized in that: The terminal device includes a processor and a memory, the memory stores a computer program, and the processor is used to execute the computer program to implement the electromyographic data processing method described in any one of claims 1 to 7 or the cervical muscle fatigue determination method described in claim 8.

Images