Method for detecting muscle mass of fracture patient based on electromyographic signal analysis

Abstract

The application provides a fracture patient muscle mass detection method based on electromyographic signal analysis, and relates to the technical field of medical treatment. The method comprises the following steps: acquiring electromyographic signals of a target muscle and its adjacent muscles of a patient; when the correlation coefficients of the target muscle and the adjacent muscle signals and their envelope lines meet preset conditions, judging whether the target muscle is interfered by the adjacent muscles based on the signal difference degree; when the adjacent muscles interfere, extracting independent frequency component sequences of the target muscle and the adjacent muscle signals, removing interference components based on the time domain correlation coefficients and the frequency overlap degree of each independent frequency component group, and obtaining a real target muscle signal; determining the muscle recovery state of the patient based on the amplitude variation trend of the real target muscle signal sequence of the patient in a target period, and the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment in the signal envelope line of the target sub-sequence of the sequence, and formulating a rehabilitation training scheme accordingly. The application can more accurately determine the muscle recovery state of the patient.

Description

Technical Field

[0001] This application relates to the field of medical technology, specifically to a method for detecting muscle quality in fracture patients based on electromyography signal analysis. Background Technology

[0002] Fracture patients often experience local muscle atrophy in the fracture area during the recovery process. Accurately monitoring changes in muscle strength is a key prerequisite for encouraging patients to actively engage in targeted rehabilitation training and ensuring effective rehabilitation.

[0003] In related technologies, surface electromyography (sEMG) is commonly used to detect muscle quality in fracture patients. The principle is to collect surface electromyographic signals of the target muscle and compare the intensity of the electrical signal response of the muscle at different stages to assess the patient's actual muscle strength and functional status, providing a reference for monitoring the rehabilitation process.

[0004] However, when using the above methods to detect muscle mass in patients, changes in muscle movement patterns are common due to local muscle atrophy after fractures, causing the activity patterns of the target muscles to deviate from the normal state. Furthermore, when collecting electromyographic signals from the surface of the target muscle, the electrical activity of adjacent muscles can be conducted to the detection electrodes of the target muscle through media such as body fluids and skin, forming electromyographic crosstalk. This makes it impossible for the collected electromyographic signals from the surface of the target muscle to truly reflect its actual activation state and functional level, resulting in a significant reduction in the accuracy of the test results. Consequently, it is difficult to accurately determine the true decline in the patient's muscle strength and cannot provide a reliable basis for the scientific formulation of rehabilitation training programs. Summary of the Invention

[0005] To address the problem that existing technologies fail to accurately reflect the actual activation state and functional level of the target muscle surface through the acquisition of electromyographic signals, leading to significantly reduced accuracy of test results and making it difficult to accurately determine the true decline in muscle strength in patients, thus hindering the scientific development of rehabilitation training programs, this application provides a method for detecting muscle quality in fracture patients based on electromyographic signal analysis. The specific technical solution adopted is as follows:

[0006] Acquire the target electromyographic signal of the target muscle and the adjacent electromyographic signals of the adjacent muscles of the patient;

[0007] Determine the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signal and their envelope. When the correlation coefficient meets a preset condition, determine whether there is adjacent muscle interference in the target muscle based on the degree of signal difference between the target electromyographic signal and the adjacent electromyographic signal.

[0008] When the target muscle is interfered with by neighboring muscles, the independent frequency component sequences of the target electromyography signal and the neighboring electromyography signal are extracted. The interference components are removed based on the time-domain correlation coefficient and frequency overlap of each independent frequency component group to obtain the real target muscle signal.

[0009] The amplitude change trend is determined based on the signal amplitude sequence corresponding to the target sequence formed by the real target muscle signals of the patient within the target time period.

[0010] Based on the sequence signal envelope of the target subsequence in the target sequence, identify the contraction state sequence segment and the relaxation state sequence segment, and determine the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment;

[0011] The patient's muscle recovery status is determined based on the amplitude change trend and the amplitude difference, and a rehabilitation training plan is formulated based on the muscle recovery status.

[0012] For example, determining the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signal and their envelope includes: calculating the Pearson coefficient between the target electromyographic signal and the adjacent electromyographic signal, denoted as the signal correlation coefficient; extracting the first upper envelope of the target electromyographic signal and the second upper envelope of the adjacent electromyographic signal, and calculating the Pearson coefficient between the first upper envelope and the second upper envelope, denoted as the envelope correlation coefficient.

[0013] For example, the preset condition is that the target muscle and its adjacent muscles are not functionally co-moving muscle groups; the step of determining whether the target muscle has adjacent muscle interference based on the signal difference between the target electromyography (EMG) signal and the adjacent EMG signal when the correlation coefficient meets the preset condition includes: when the signal correlation coefficient is less than a first preset threshold and the envelope correlation coefficient is greater than a second preset threshold, determining that the target muscle and its adjacent muscles are functionally co-moving muscle groups; otherwise, determining that the target muscle and its adjacent muscles meet the preset condition; when the correlation coefficient meets the preset condition, acquiring the first peak sequence of the corresponding target EMG signal and the second peak sequence of the adjacent EMG signal, and determining the signal difference based on the first peak sequence and the second peak sequence; determining the degree of adjacent muscle interference based on the signal correlation coefficient and the degree of signal difference, and determining that the target muscle has adjacent muscle interference when the degree of adjacent muscle interference is greater than a third preset threshold.

[0014] For example, determining the degree of signal difference based on the first peak sequence and the second peak sequence includes: for each first peak in the first peak sequence, determining the second peak in the second peak sequence that is closest to the first peak in time, calculating the peak difference between the first peak and the second peak and its corresponding time difference; and determining the degree of signal difference based on the peak difference and the time difference between each first peak and the second peak.

[0015] For example, when the target muscle is subject to interference from neighboring muscles, extracting independent frequency component sequences of the target electromyography (EMG) signal and the neighboring EMG signals, and removing interference components based on the time-domain correlation coefficient and frequency overlap degree of each independent frequency component group to obtain the true target muscle signal includes: when the target muscle is subject to interference from neighboring muscles, extracting a first independent frequency component sequence of the target EMG signal and a second independent frequency component sequence of the neighboring EMG signals based on an independent component analysis algorithm; calculating the Pearson coefficient between any first independent frequency component in the first independent frequency component sequence and any second independent frequency component in the second independent frequency component sequence, denoted as the time-domain correlation coefficient; performing a Fourier transform on the arbitrary first independent frequency component and the arbitrary second independent frequency component to obtain a first spectral distribution corresponding to the arbitrary first independent frequency component and a second spectral distribution corresponding to the arbitrary second independent frequency component, calculating the overlap area between the first spectral distribution and the second spectral distribution, denoted as the frequency overlap degree; and removing the interference components based on the time-domain correlation coefficient and the frequency overlap degree to obtain the true target muscle signal.

[0016] For example, the step of removing the interference component based on the time-domain correlation coefficient and the frequency overlap to obtain the real target muscle signal includes: determining the signal similarity between any first independent frequency component and any second independent frequency component based on the time-domain correlation coefficient and the frequency overlap; determining the arbitrary first independent frequency component as the interference component when the signal similarity is greater than a fourth preset threshold; removing the first independent frequency components determined to be the interference components from the first independent frequency component sequence to obtain the real target muscle signal.

[0017] For example, determining the amplitude change trend based on the signal amplitude sequence corresponding to the target sequence formed by the real target muscle signals of the patient within the target time period includes: obtaining the real target muscle signals of multiple consecutive days corresponding to the target time period to form the target sequence; calculating the average value of the real target muscle signals of each day in the target sequence and recording it as the real signal amplitude of that day; calculating the difference between the real signal amplitudes of all adjacent two days and averaging them to obtain the amplitude change trend.

[0018] For example, the step of identifying contraction state sequence segments and relaxation state sequence segments based on the sequence signal envelope of the target sub-sequence in the target sequence, and determining the amplitude difference between the contraction state sequence segments and the relaxation state sequence segments, includes: determining a target sub-sequence in the target sequence; the target sub-sequence is a sub-sequence composed of the real target muscle signal that is closest to the current time in the target sequence; obtaining the sequence signal envelope of the target sub-sequence; traversing the sequence signal envelope, identifying sequence segments whose amplitude is continuously greater than a fifth preset threshold and whose duration reaches a preset duration as the contraction state sequence segment, and identifying sequence segments whose amplitude is continuously less than the fifth preset threshold and whose duration reaches the preset duration as the relaxation state sequence segment; for adjacent contraction state sequence segments and relaxation state sequence segments, calculating the difference between the maximum contraction amplitude in the contraction state sequence segment and the minimum relaxation amplitude in the relaxation state sequence segment, and averaging the difference, which is recorded as the amplitude difference.

[0019] For example, determining the patient's muscle recovery status based on the amplitude change trend and the amplitude difference includes: calculating the product of the amplitude change trend and the amplitude difference and performing normalization processing to obtain the muscle recovery status.

[0020] For example, the step of developing a rehabilitation training program based on the muscle recovery status includes: determining the patient's recovery level based on the muscle recovery status, and determining the rehabilitation training program based on the recovery level.

[0021] This application may have some or all of the following beneficial effects:

[0022] In the muscle quality detection method for fracture patients based on electromyography (EMG) signal analysis provided in this application, the correlation coefficient between the target EMG signal and adjacent EMG signals and their envelopes is determined. When the correlation coefficient meets preset conditions, the presence of adjacent muscle interference in the target muscle is judged based on the degree of signal difference between the target EMG signal and adjacent EMG signals. This can accurately identify adjacent muscle interference and avoid misjudging functional synergistic movements as interference. When adjacent muscle interference is determined to exist, the independent frequency component sequences of the target muscle and adjacent EMG signals are extracted. By combining the time-domain correlation coefficient and frequency overlap of each independent frequency component group, the interference components are accurately located and removed. This can effectively isolate the influence of adjacent EMG activity on the target muscle signal and obtain a signal that truly reflects the activation state of the target muscle. Based on the amplitude change trend of the real target muscle signal sequence within the target time period and the amplitude difference between the contraction and relaxation state sequences of the target subsequence in the sequence, the muscle recovery status is assessed. This can achieve accurate quantification of muscle recovery, thereby providing a basis for the formulation of rehabilitation training programs and providing scientifically adapted training guidance for the recovery of muscle function in fracture patients.

[0023] It should be understood that the above general description and the following detailed description are exemplary and explanatory only, and do not limit this application. Attached Figure Description

[0024] To more clearly illustrate the technical solutions and advantages in the embodiments of this application or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0025] Figure 1 A flowchart is shown of a method for detecting muscle mass in fracture patients based on electromyography signal analysis according to an exemplary embodiment of this application. Detailed Implementation

[0026] To further illustrate the technical means and effects adopted by this application to achieve the intended purpose of the invention, the following, in conjunction with the accompanying drawings and preferred embodiments, details the specific implementation, structure, features, and effects of the method for detecting muscle quality in fracture patients based on electromyography signal analysis proposed in this application. In the following description, different "one embodiment" or "another embodiment" do not necessarily refer to the same embodiment. Furthermore, specific features, structures, or characteristics in one or more embodiments can be combined in any suitable form.

[0027] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains.

[0028] The following section, in conjunction with the accompanying drawings, details the specific scheme of the muscle quality detection method for fracture patients based on electromyography signal analysis provided in this application.

[0029] Please see Figure 1 It illustrates a flowchart of a method for detecting muscle quality in fracture patients based on electromyography signal analysis, according to an embodiment of this application. Figure 1 As shown, this method for detecting muscle quality in fracture patients based on electromyography signal analysis specifically includes the following steps:

[0030] S110: Acquire the target electromyographic signal of the target muscle and the adjacent electromyographic signals of the adjacent muscles of the patient;

[0031] S120: Determine the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signal and their envelope. When the correlation coefficient meets the preset conditions, determine whether there is interference from the adjacent muscle based on the degree of signal difference between the target electromyographic signal and the adjacent electromyographic signal.

[0032] S130: When there is interference from neighboring muscles in the target muscle, extract the independent frequency component sequences of the target electromyography signal and the neighboring electromyography signal, remove the interference components based on the time-domain correlation coefficient and frequency overlap of each independent frequency component group, and obtain the real target muscle signal.

[0033] S140: Determine the amplitude change trend based on the signal amplitude sequence corresponding to the target sequence formed by the patient's real target muscle signals within the target time period;

[0034] S150: Identify contraction state sequence segments and relaxation state sequence segments based on the sequence signal envelope of the target subsequence in the target sequence, and determine the amplitude difference between the contraction state sequence segments and the relaxation state sequence segments.

[0035] S160: Determine the patient's muscle recovery status based on the amplitude change trend and amplitude difference, and formulate a rehabilitation training plan based on the muscle recovery status.

[0036] The following is a detailed explanation of each step in the above-mentioned method for detecting muscle quality in fracture patients based on electromyography signal analysis:

[0037] In step S110, the target electromyographic signal of the patient's target muscle and the adjacent electromyographic signals of its neighboring muscles are acquired.

[0038] In this embodiment of the application, the above-mentioned patients refer to people who have suffered fractures and are in the process of post-injury rehabilitation, and who need to be evaluated for muscle quality and functional status in the fracture area through electromyography signal analysis.

[0039] In this embodiment of the application, the target muscle refers to the core muscle directly related to the patient's fracture area, which is the core muscle whose muscle quality and functional status need to be assessed, and is the core object of muscle quality detection.

[0040] In the embodiments of this application, the aforementioned adjacent muscle refers to a muscle that is anatomically close to the target muscle and whose electrical activity may be conducted to the target muscle detection electrode through body fluids or skin, thereby interfering with the target muscle's electrical signal.

[0041] In this embodiment, electromyography (EMG) signals refer to the weak potential changes generated by the activation of muscle fibers by motor neurons during muscle contraction or relaxation, reflecting the degree of muscle activation, force output, and functional state. Exemplarily, the aforementioned EMG signals can be surface EMG signals acquired non-invasively via skin surface electrodes for recording muscle activity.

[0042] Furthermore, the aforementioned target electromyographic signal refers to the weak potential change signal generated by the activation of muscle fibers by motor neurons during the contraction or relaxation of the target muscle, which is collected by non-invasive surface electrodes and is used to reflect the activation degree, force output and functional state of the target muscle; the aforementioned adjacent electromyographic signal refers to the weak potential change signal generated by adjacent muscles during the contraction or relaxation of the adjacent muscles, which is used to determine whether there is signal interference from adjacent muscles to the target muscle by comparing and analyzing it with the target electromyographic signal.

[0043] For example, the acquisition of the target electromyographic (EMG) signal of the patient's target muscle and the EMG signals of adjacent muscles can be achieved as follows: A non-invasive bipolar surface electrode is selected and placed at the center of the target muscle belly and the center of the adjacent muscle belly in the fracture area of ​​the patient, with its axis parallel to the muscle fiber direction; the amplitude, speed, and joint angle of muscle movements are standardized when the patient's signals are collected to achieve standardized movements; the EMG signals of the target muscle and adjacent muscles are simultaneously acquired using a surface EMG signal acquisition device, and the original waveform data of the target EMG signal and adjacent EMG signals, as well as auxiliary data such as the patient's actual output force, are recorded daily to form a systematic EMG activity archive; wherein, the sampling rate of the aforementioned surface EMG signal acquisition device is ≥1000Hz, the bandpass filter is 20–450Hz, and a 50 / 60Hz notch filter is enabled when necessary, and the acquired EMG signals must cover the entire movement cycle of the aforementioned standardized movements.

[0044] In step S120, the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signal and their envelope is determined. When the correlation coefficient meets the preset conditions, the presence of adjacent muscle interference in the target muscle is determined based on the degree of signal difference between the target electromyographic signal and the adjacent electromyographic signal.

[0045] In this embodiment, the envelope is a signal curve obtained by extracting the upper envelope of the target electromyographic signal and adjacent electromyographic signals. It can smooth the high-frequency fluctuations in the original electromyographic signal caused by the randomness of muscle fiber discharge, filter out local small amplitude changes, and retain only the overall trend of the electromyographic signal amplitude change over time.

[0046] In this embodiment, the correlation coefficient between the target EMG signal and the adjacent EMG signal is a quantification of the linear correlation between the target EMG signal and the adjacent EMG signal in the time domain. For example, this correlation coefficient can be calculated using the Pearson correlation coefficient algorithm, with a value range of [-1, 1]. The closer the value is to 1 or -1, the stronger the temporal synchronization between the target EMG signal and the adjacent EMG signal, and the more likely crosstalk or cooperative motion exists; the closer the value is to 0, the stronger the temporal independence between the target EMG signal and the adjacent EMG signal.

[0047] In this embodiment, the correlation coefficient between the target electromyographic signal and the envelope of adjacent electromyographic signals is a quantification of the linear correlation between the target muscle signal envelope and the envelope of adjacent muscle signals. For example, this correlation coefficient can be calculated using the Pearson correlation coefficient algorithm. If the target muscle and its adjacent muscles exhibit true coordinated movement (e.g., the biceps brachii and brachialis contract together during elbow flexion), the envelope will change synchronously, resulting in a higher correlation coefficient. If it is merely crosstalk between adjacent electromyographic signals, the envelope changes asynchronously, leading to a lower correlation coefficient.

[0048] For example, the determination of the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signals and their envelopes can be achieved as follows: calculate the Pearson coefficient between the target electromyographic signal and the adjacent electromyographic signal, denoted as the signal correlation coefficient; extract the first upper envelope of the target electromyographic signal and the second upper envelope of the adjacent electromyographic signal, and calculate the Pearson coefficient between the first upper envelope and the second upper envelope, denoted as the envelope correlation coefficient.

[0049] In one specific implementation of this application embodiment, the process of determining the signal correlation coefficient and envelope correlation coefficient can be achieved as follows: acquiring the target electromyographic signal acquired by the surface electromyographic signal acquisition device. and adjacent electromyographic signals The Pearson correlation coefficient algorithm is used to calculate the signal correlation coefficient. as follows: Target electromyographic signals were extracted by processing the upper envelope. First upper envelope and adjacent electromyographic signals The second upper envelope The envelope correlation coefficient was calculated using the Pearson correlation coefficient algorithm. as follows: .

[0050] When the target muscle is interfered with by adjacent muscles, the surface electrodes used to collect signals not only receive signals from below the target muscle fibers but may also pick up the activity of adjacent muscles. At this time, the target muscle electromyography (EMG) signal and the adjacent muscle EMG signal show synchronous fluctuations in the time series, with high correlation between the original signals but low correlation between the envelope. This reflects that the target muscle EMG signal and the adjacent muscle EMG signal share a common neural driving source or physical superposition effect, that is, the target muscle is currently being interfered with by the adjacent muscle signal. However, when the target muscle and the adjacent muscles participate in movement together functionally, their envelopes will show a high correlation, reflecting a real synergistic effect. Therefore, the embodiments of this application can determine whether the target muscle is interfered with by adjacent muscles by the above-mentioned signal correlation coefficient and envelope correlation coefficient. Specifically, the case where the target muscle and the adjacent muscles are functionally co-moving muscle blocks can be ruled out first, and then the existence of adjacent muscle interference can be further determined.

[0051] In this embodiment, the aforementioned preset conditions are threshold conditions set to distinguish between functional co-movement and interference requiring further judgment. For example, when the signal correlation coefficient is less than the first preset threshold and the envelope correlation coefficient is greater than the second preset threshold, the target muscle and its adjacent muscles are determined to be muscle groups of functional co-movement; otherwise, the target muscle and its adjacent muscles are determined to meet the preset conditions.

[0052] In one specific implementation of this application embodiment, the above threshold condition can be set as follows: setting the signal correlation coefficient. The threshold is 0.3, and the envelope correlation coefficient is... The threshold is 0.8; if the signal correlation coefficient Less than 0.3 and envelope correlation coefficient If the value is greater than 0.8, the target muscle and adjacent muscles are considered to be functionally co-moving muscle groups; other signals are used to proceed to the next step of determining whether there is interference from adjacent muscles.

[0053] For example, the determination of whether there is interference from adjacent muscles can be implemented as follows: when the correlation coefficient meets a preset condition, the first peak sequence of the target electromyography (EMG) signal and the second peak sequence of the adjacent EMG signal are obtained, and the degree of signal difference is determined based on the first peak sequence and the second peak sequence; the degree of interference from adjacent muscles is determined based on the signal correlation coefficient and the degree of signal difference, and when the degree of interference from adjacent muscles is greater than a third preset threshold, it is determined that there is interference from adjacent muscles in the target muscle. Specifically, the determination of the degree of signal difference based on the first peak sequence and the second peak sequence can be implemented as follows: for each first peak in the first peak sequence, the second peak in the second peak sequence that is temporally closest to the first peak is determined, and the peak difference between the first peak and the second peak and its corresponding time difference is calculated; the degree of signal difference is determined based on the peak difference and time difference between each first peak and the second peak.

[0054] In one specific implementation of this application embodiment, the above-mentioned process of determining adjacent muscle interference can be achieved as follows: extracting target muscle signals. First peak sequence and the corresponding first peak time series ;in, for The amplitude of the i-th peak, for The time of its appearance, The number of peak values ​​of the target muscle signal; extraction of adjacent electromyographic signals. The second peak sequence and the corresponding second peak time series ;in, for The amplitude of the i-th peak, for The time of its appearance, The number of peaks in the adjacent muscle signal; for the first peak sequence The first peak of each (k=1,2,...,m), in the second peak sequence The first peak value was determined in the middle. The second peak that is closest in time ; Calculate the peak difference between the first peak and the second peak. : and the time difference of their occurrence. : The degree of signal difference is specifically calculated using the following formula. : ,in, Let be the peak difference between the i-th first peak and the second peak. The time difference between the occurrence times of the first and second peaks represents the degree of signal difference. The higher the value, the stronger the target electromyographic signal. and adjacent electromyographic signals The more significant the difference in the time domain, the more the degree of interference between adjacent muscles is calculated using the following formula. : ,in, The correlation coefficient of the above signals, This is the normalization function. In this formula, when... A high value (strong synchronization, conforming to the characteristics of crosstalk signal superposition) and High values ​​(corresponding to signal differences) Small peak values ​​(consistent with crosstalk signal superposition characteristics) indicate a high probability of interference from neighboring muscles in the target muscle. The crosstalk probability threshold for the current neighboring muscle (i.e., the third preset threshold mentioned above) is set to 0.7. If the degree of neighboring muscle interference... A value greater than 0.7 indicates that adjacent muscles are interfering with the current target muscle.

[0055] In step S130, when there is interference from neighboring muscles in the target muscle, the independent frequency component sequences of the target electromyography signal and the neighboring electromyography signals are extracted. The interference components are removed based on the time-domain correlation coefficient and frequency overlap of each independent frequency component group to obtain the real target muscle signal.

[0056] The aforementioned adjacent muscle interference refers to the phenomenon where adjacent muscle electrical activity is conducted to the detection electrodes of the target muscle through media such as body fluids or skin, causing the target muscle electrical signal to deviate from its true activation state. This adjacent muscle interference prevents the target muscle electrical signal from accurately reflecting its own activation level, force output, and functional state. Therefore, in this embodiment, after determining the presence of adjacent muscle interference in the target muscle through the above process, this step is required to remove the interfering component.

[0057] Since different muscles have their own contraction patterns and frequency characteristics, the activities of the target muscle and its neighboring muscles are statistically approximately independent. The aforementioned independent frequency component sequence is a signal sequence with a single frequency characteristic and independent of each muscle obtained by decomposing the mixed signal using a frequency decomposition algorithm based on this independence. Each independent frequency component corresponds to only a specific frequency range and can reflect the electromyographic activity characteristics of a certain frequency band. This provides a frequency dimension basis for distinguishing the target muscle's own signal from the interference components of neighboring muscles, thereby decoupling the electrical activities of different muscles in space. For example, the Independent Component Analysis (ICA) algorithm can be used to decompose the mixed signal.

[0058] In the embodiments of this application, the aforementioned independent frequency component group is a pairwise combination obtained by matching the independent frequency component sequences of the target electromyographic signal and the independent frequency component sequences of adjacent electromyographic signals according to the principle of similar frequency ranges.

[0059] In this embodiment, the aforementioned time-domain correlation coefficient is used to reflect the time-domain synchronicity of two independent frequency components in an independent frequency component group. For example, the Pearson correlation coefficient algorithm can be used to determine the time-domain correlation coefficient. The closer its value is to 1, the more synchronized the fluctuations of the target muscle independent frequency component and the adjacent muscle independent frequency component are in the time series, and the target muscle independent frequency component is likely to be an interfering component of the adjacent muscle. The closer its value is to 0, the stronger the time-domain independence of the target muscle independent frequency component and the adjacent muscle independent frequency component is, and the target muscle independent frequency component is likely to be a signal component of the target muscle itself.

[0060] In this embodiment, the frequency overlap is used to reflect the overlap ratio of the frequency ranges of two independent frequency components in the independent frequency component group; the higher the frequency overlap between the independent frequency component of the target muscle and the independent frequency component of the adjacent muscle, the more it conforms to the characteristic of the adjacent muscle interference component being transmitted through the same frequency band, and the independent frequency component of the target muscle is likely to be the adjacent muscle interference component.

[0061] In this embodiment, the aforementioned interference component is an interfering component mixed into the target muscle detection electrode by the adjacent electromyographic signal transmitted to the target muscle through a medium such as body fluid or skin. It cannot reflect the activation state of the target muscle itself and needs to be removed from the target electromyographic signal to ensure the purity of the target muscle signal and its applicability to muscle quality assessment in fracture patients. For example, when the independent frequency component of the target muscle and the corresponding independent frequency component of the adjacent muscle highly overlap in frequency and show a high correlation in the time domain, the independent frequency component of the target muscle can be identified as an interference component.

[0062] For example, the above process of removing interference components can be implemented as follows: When there is interference from neighboring muscles in the target muscle, the first independent frequency component sequence of the target electromyography (EMG) signal and the second independent frequency component sequence of the neighboring EMG signals are extracted based on the independent component analysis algorithm; for any first independent frequency component in the first independent frequency component sequence and any second independent frequency component in the second independent frequency component sequence, the Pearson coefficient between any first independent frequency component and any second independent frequency component is calculated and denoted as the time-domain correlation coefficient; Fourier transform is performed on any first independent frequency component and any second independent frequency component to obtain the first spectral distribution corresponding to any first independent frequency component and the second spectral distribution corresponding to any second independent frequency component, and the overlap area between the first spectral distribution and the second spectral distribution is calculated and denoted as the frequency overlap degree; interference components are removed based on the time-domain correlation coefficient and the frequency overlap degree to obtain the true target muscle signal.

[0063] The above-mentioned method of removing interference components based on time-domain correlation coefficient and frequency overlap to obtain the real target muscle signal can be achieved as follows: determine the signal similarity between any first independent frequency component and any second independent frequency component based on time-domain correlation coefficient and frequency overlap; when the signal similarity is greater than a fourth preset threshold, determine any first independent frequency component as an interference component; remove the target muscle component that is determined to be an interference component from the first independent frequency component sequence to obtain the real target muscle signal.

[0064] In one specific implementation of this application embodiment, the process of removing interference components described above is explained in detail: ICA is used to analyze the target electromyographic signal. and adjacent electromyographic signals The sequence of the first independent frequency components of the target muscle is obtained by decomposition. Where m1 is the number of independent frequency components of the target muscle signal, and m1 is the sequence of the second independent frequency components of the adjacent muscle. Where m2 is the number of independent frequency components of the adjacent electromyographic signals; for the first independent frequency component sequence Any first independent frequency component in and the second independent frequency component sequence Any second independent frequency component in The first independent frequency component was calculated using Pearson coefficients. With the second independent frequency component The time-domain correlation coefficient between them : For the first independent frequency component sequence and the second independent frequency component sequence Perform Fourier transforms on each independent frequency component to obtain the corresponding spectral distribution, using the first independent frequency component as an example. With the second independent frequency component For example, obtain the first independent frequency component. First spectral distribution With the second independent frequency component Second spectral distribution The first independent frequency component is calculated using the following formula. With the second independent frequency component Frequency overlap : ,in, To find the minimum value function, used in each small frequency range The first spectral distribution is selected internally. With the second spectral distribution The smaller value in the value, that is, if the area of ​​the second spectral distribution of the adjacent muscle at a certain frequency point is much larger than the area of ​​the first spectral distribution of the target muscle, then the overlap energy at that frequency point is determined by the first spectral distribution of the target muscle, thus avoiding the miscalculation of high-energy regions of adjacent muscles into the overlap and ensuring... It only reflects the overlap of frequency energy shared by the two components; By each tiny frequency range The minimum spectral density accumulation on the surface yields the first independent frequency component. With the second independent frequency component Total spectral overlap area over the entire effective frequency range The value range is [0,1]. The closer the value is to 1, the stronger it reflects the first independent frequency component. With the second independent frequency component The higher the spectral overlap, the closer its value is to 0, reflecting the first independent frequency component. With the second independent frequency component The lower the spectral overlap, the more accurate the calculation of the first independent frequency component using the following formula. With the second independent frequency component Signal similarity between : ,in, The degree of frequency overlap mentioned above, The above-mentioned time-domain correlation coefficient; the signal similarity threshold (i.e., the above-mentioned fourth preset threshold) between the independent frequency component of the target muscle and the independent frequency component of neighboring muscles is determined to be 0.5. If the first independent frequency component With the second independent frequency component Signal similarity between If the value is greater than 0.5, then the second independent frequency component in the adjacent muscle is determined. The interference component is the target muscle.

[0065] Similarly, in the embodiments of this application, the first independent frequency component sequence of the target muscle can be determined. All interference components in the sequence are identified, and after identifying the interference components, the first independent frequency components that are identified as interference components in the first independent frequency component sequence are removed to obtain the final true target muscle signal C.

[0066] In step S140, the amplitude change trend is determined based on the signal amplitude sequence corresponding to the target sequence formed by the patient's real target muscle signals within the target time period.

[0067] In this embodiment, the target time period is a specific time interval set according to the muscle quality assessment needs of fracture patients for extracting and analyzing data; the target sequence is a sequence of real target muscle signals within the target time period; for example, the target sequence can be a sequence of real target muscle signals of a patient over N consecutive days.

[0068] In the embodiments of this application, the above-mentioned signal amplitude sequence refers to the numerical sequence extracted from the target sequence that reflects the magnitude of the signal amplitude at each sampling point, which can be used to reflect the strength changes of the patient's target electromyographic activity within the target time period.

[0069] For example, the above method of determining the amplitude change trend based on the signal amplitude sequence corresponding to the target sequence formed by the patient's real target muscle signals within the target time period can be achieved as follows: obtain the real target muscle signals of multiple consecutive days corresponding to the target time period to form a target sequence; calculate the average value of the real target muscle signals of each day in the target sequence and record it as the real signal amplitude of that day; calculate the difference between the real signal amplitudes of all adjacent two days and take the average to obtain the amplitude change trend.

[0070] In one specific implementation of this application, the process of determining the amplitude change trend described above can be achieved as follows: obtaining the patient's real target muscle signal sequence for N consecutive days (i.e., the target sequence described above) determined through the above steps. ,in, (i=1,2,…,N) represents the real target muscle signal groups for day i. Each group contains multiple real target muscle signals, the number of which is determined by the sampling frequency. For each day's real target muscle signal group, the average amplitude of each real target muscle signal in the group is calculated and recorded as the real signal amplitude for that day, thus obtaining a muscle signal amplitude sequence for N consecutive days. ,in, Let be the actual signal amplitude on day i; calculate the above amplitude change trend using the following formula. : The trend of this amplitude change This reflects the average daily change in the intensity of target electromyographic activity over N consecutive days. If the value of w is greater than 0, it indicates that the daily average amplitude is on the rise, reflecting that the average muscle activation intensity is stronger each day than the previous day, and the muscle function is recovering well. If the value of w is close to 0, it indicates that the daily average amplitude is basically stable, reflecting that the muscle activation intensity has entered a plateau period, which may be a stable state after recovery to a certain stage. If the value of w is less than 0, it indicates that the daily average amplitude is on the fall, reflecting that the average muscle activation intensity is weaker each day than the previous day, and the possibility of poor recovery or muscle function degeneration should be noted.

[0071] In step S150, the contraction state sequence segment and the relaxation state sequence segment are identified based on the sequence signal envelope of the target subsequence in the target sequence, and the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment is determined.

[0072] In this embodiment, the target subsequence is a subsequence of the actual target muscle signal extracted from the target sequence that is close to the current time. For example, the target subsequence can be the subsequence of the most recent three days from a target sequence of N consecutive days.

[0073] In this embodiment, the aforementioned contraction state sequence segment is a signal segment in the target subsequence whose sequence signal envelope amplitude is continuously higher than a preset contraction threshold, corresponding to the active activation state of the muscle.

[0074] In this embodiment, the aforementioned relaxed state sequence segment is a signal segment in the target subsequence whose sequence signal envelope amplitude is continuously lower than a preset relaxation threshold, corresponding to the inactive or low-activation state of the muscle.

[0075] Exemplarily, the recognition of the contraction state sequence segment and the relaxation state sequence segment of the sequence signal envelope based on the target subsequence in the target sequence and the determination of the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment can be achieved as follows: Determine the target subsequence in the target sequence, where the target subsequence is a subsequence composed of the real target muscle signals of the preset number of days closest to the current time in the target sequence; Obtain the sequence signal envelope of the target subsequence; Traverse the sequence signal envelope, identify the sequence segment with the amplitude continuously greater than the fifth preset threshold and the continuous duration reaching the preset duration as the contraction state sequence segment, and identify the sequence segment with the amplitude continuously less than the fifth preset threshold and the continuous duration reaching the preset duration as the relaxation state sequence segment; For adjacent contraction state sequence segments and relaxation state sequence segments, calculate the difference between the maximum contraction amplitude in the contraction state sequence segment and the minimum relaxation amplitude in the relaxation state sequence segment and take the average, which is denoted as the amplitude difference.

[0076] In a specific implementation manner of the embodiment of the present application, the process of determining the amplitude difference can be achieved as follows: Obtain the real target electromyogram signal sequence of the recent N1 (N1 < N, N1 = 3) days (i.e., the above-mentioned target subsequence) ; Extract of the sequence signal envelope, set the signal envelope threshold (i.e., the above-mentioned fifth preset threshold) to 10, and determine the sequence segment with the signal envelope amplitude exceeding 10 continuously for 50 ms (i.e., the above-mentioned preset duration) in as the contraction state sequence segment, denoted as , and determine the sequence segment with the signal envelope amplitude lower than 10 continuously for 50 ms in as the relaxation state sequence segment, denoted as ; Calculate the above-mentioned amplitude difference through the following formula : , where is the number of cycles from the contraction state sequence segment to the relaxation state sequence segment; is the maximum contraction amplitude of the contraction state sequence segment in the i-th group of cycles, is the minimum relaxation amplitude of the relaxation state sequence segment in the i-th group of cycles; reflects the average amplitude span of the muscle from the maximum contraction activation to the minimum relaxation inhibition in m groups of contraction-relaxation cycles. The larger its value, the greater the gap between the maximum activation intensity of the muscle during each contraction and the minimum activation intensity during relaxation, indicating that the muscle has stronger contraction force and better relaxation control, and the muscle function state is good; The smaller its value, the smaller the amplitude span between contraction and relaxation, which may be due to insufficient contraction activation or incomplete relaxation, indicating that there is room for improvement in muscle function.

[0077] In step S160, the muscle recovery state of the patient is determined based on the amplitude change trend and the amplitude difference, and a rehabilitation training plan is formulated based on the muscle recovery state.

[0078] In the embodiment of the present application, the above-mentioned muscle recovery state is a parameter for measuring the recovery level of the patient's target muscle function. Specifically, the muscle recovery state can be calculated by the following formula:

[0079]

[0080] Where, is the above-mentioned muscle recovery state; is the above-mentioned amplitude change trend, reflecting the long-term change rate of the amplitude of the target electromyogram signal; is the above-mentioned amplitude difference, reflecting the intensity span of the target muscle in the short-term contraction and relaxation states; is a normalization function; the above formula comprehensively quantifies the muscle recovery level based on the long-term trend and short-term function, The closer the value of is to 1, the closer the recovery level is to the healthy state.

[0081] Exemplarily, formulating a rehabilitation training plan based on the muscle recovery state can be achieved as follows: determining the patient's recovery level according to the muscle recovery state, and determining the rehabilitation training plan based on the recovery level.

[0082] Specifically, if , it is determined that the recovery level is a mild recovery state, and the rehabilitation training plan is mainly passive rehabilitation and mild active training. For example, passive joint activities and low-intensity isometric contraction exercises are supplemented with electrical stimulation or biofeedback to help the patient re-establish muscle activation ability; if 0.3 < g ≤ 0.7, it is determined that the recovery level is a moderate recovery state, and the rehabilitation training plan is to increase active participation in training, such as elastic band resistance training, medium-intensity isometric and isotonic exercises, combined with coordination and stability training, to gradually improve muscle strength and endurance; if 0.7 < g ≤ 1, it is determined that the recovery level is a healthy recovery state, and the rehabilitation training plan is to enter the functional rehabilitation and strengthening training stage, such as free weight training and high-intensity functional movement training (up and down stairs, gait training), to enhance the maximum muscle contraction ability and sports performance.

[0083] The foregoing mainly describes the solutions provided by the embodiments of this application from a methodological perspective. To achieve the above functions, it includes corresponding hardware structures and / or software modules for executing each function. Those skilled in the art should readily recognize that, based on the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein, this application can be implemented in hardware or a combination of hardware and computer software. Whether a function is executed in hardware or by computer software driving hardware depends on the specific application and design constraints of the technical solution. Those skilled in the art can use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0084] It should be noted that the order of the embodiments described above is merely for descriptive purposes and does not represent the superiority or inferiority of the embodiments. The processes depicted in the accompanying drawings do not necessarily require a specific or sequential order to achieve the desired result. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0085] The various embodiments in this specification are described in a progressive manner. The same or similar parts between the various embodiments can be referred to each other. Each embodiment focuses on describing the differences from other embodiments.

Claims

1. A method for detecting muscle mass in fracture patients based on electromyography signal analysis, characterized in that, The method includes: Acquire the target electromyographic signal of the target muscle and the adjacent electromyographic signals of the adjacent muscles of the patient; Determine the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signal and their envelope. When the correlation coefficient meets a preset condition, determine whether there is adjacent muscle interference in the target muscle based on the degree of signal difference between the target electromyographic signal and the adjacent electromyographic signal. When the target muscle is interfered with by neighboring muscles, the independent frequency component sequences of the target electromyography signal and the neighboring electromyography signal are extracted. The interference components are removed based on the time-domain correlation coefficient and frequency overlap of each independent frequency component group to obtain the real target muscle signal. The amplitude change trend is determined based on the signal amplitude sequence corresponding to the target sequence formed by the real target muscle signals of the patient within the target time period. Based on the sequence signal envelope of the target subsequence in the target sequence, identify the contraction state sequence segment and the relaxation state sequence segment, and determine the amplitude difference between the contraction state sequence segment and the relaxation state sequence segment; The patient's muscle recovery status is determined based on the amplitude change trend and the amplitude difference, and a rehabilitation training plan is formulated based on the muscle recovery status. The preset condition is that the target muscle and its adjacent muscles are not muscle groups that move together functionally. Specifically, when interference from neighboring muscles exists in the target muscle, the independent frequency component sequences of the target electromyography (EMG) signal and the neighboring EMG signals are extracted. Interference components are removed based on the time-domain correlation coefficient and frequency overlap of each independent frequency component group to obtain the true target muscle signal. This includes: when interference from neighboring muscles exists in the target muscle, extracting a first independent frequency component sequence of the target EMG signal and a second independent frequency component sequence of the neighboring EMG signals based on an independent component analysis algorithm; calculating the Pearson coefficient between any first independent frequency component in the first independent frequency component sequence and any second independent frequency component in the second independent frequency component sequence, denoted as the time-domain correlation coefficient; and for any... A Fourier transform is performed on the first independent frequency component and the arbitrary second independent frequency component to obtain the first spectral distribution corresponding to the arbitrary first independent frequency component and the second spectral distribution corresponding to the arbitrary second independent frequency component. The overlap area between the first spectral distribution and the second spectral distribution is calculated and denoted as the frequency overlap degree. The signal similarity between the arbitrary first independent frequency component and the arbitrary second independent frequency component is determined based on the time-domain correlation coefficient and the frequency overlap degree. When the signal similarity degree is greater than a fourth preset threshold, the arbitrary first independent frequency component is determined to be the interference component. The first independent frequency components that are determined to be the interference components in the first independent frequency component sequence are removed to obtain the real target muscle signal.

2. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 1, characterized in that, Determining the correlation coefficient between the target electromyographic signal and the adjacent electromyographic signals and their envelope includes: Calculate the Pearson coefficient between the target electromyographic signal and the adjacent electromyographic signal, and denote it as the signal correlation coefficient; The first upper envelope of the target electromyographic signal and the second upper envelope of the adjacent electromyographic signal are extracted. The Pearson coefficient between the first upper envelope and the second upper envelope is calculated and denoted as the envelope correlation coefficient.

3. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 2, characterized in that, When the correlation coefficient meets a preset condition, determining whether there is interference from neighboring muscles in the target muscle based on the degree of signal difference between the target electromyographic signal and the adjacent electromyographic signal includes: When the signal correlation coefficient is less than a first preset threshold and the envelope correlation coefficient is greater than a second preset threshold, the target muscle and its adjacent muscles are determined to be a muscle group that functions together; otherwise, the target muscle and its adjacent muscles are determined to meet the preset conditions. When the correlation coefficient meets the preset condition, the first peak sequence of the target electromyography signal and the second peak sequence of the adjacent electromyography signal are obtained, and the degree of signal difference is determined based on the first peak sequence and the second peak sequence. The degree of interference between adjacent muscles is determined based on the signal correlation coefficient and the degree of signal difference. When the degree of interference between adjacent muscles is greater than a third preset threshold, it is determined that the target muscle has interference between adjacent muscles.

4. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 3, characterized in that, Determining the degree of signal difference based on the first peak sequence and the second peak sequence includes: For each first peak in the first peak sequence, determine the second peak in the second peak sequence that is closest to the first peak in time, and calculate the peak difference between the first peak and the second peak and its corresponding time difference; The degree of signal difference is determined based on the peak difference and time difference between each of the first peak and the second peak.

5. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 1, characterized in that, The determination of amplitude change trends based on the signal amplitude sequence corresponding to the target sequence formed by the actual target muscle signals of the patient within the target time period includes: The target muscle signals for multiple consecutive days corresponding to the target time period are obtained to form the target sequence; Calculate the average value of the real target muscle signal for each day in the target sequence, and record it as the real signal amplitude for that day; The difference between the actual signal amplitudes of all two consecutive days is calculated and averaged to obtain the amplitude change trend.

6. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 5, characterized in that, The step of identifying contraction-state and relaxation-state sequence segments based on the sequence signal envelope of the target sub-sequence in the target sequence, and determining the amplitude difference between the contraction-state and relaxation-state sequence segments, includes: A target subsequence is determined in the target sequence; the target subsequence is a subsequence composed of the real target muscle signal that is closest to the current time within the target sequence by a preset number of days. Obtain the sequence signal envelope of the target subsequence; Traverse the envelope of the sequence signal, identify the sequence segment whose amplitude is continuously greater than the fifth preset threshold and whose duration reaches the preset duration as the contraction state sequence segment, and identify the sequence segment whose amplitude is continuously less than the fifth preset threshold and whose duration reaches the preset duration as the relaxation state sequence segment; For adjacent contraction state sequence segments and relaxation state sequence segments, the difference between the maximum contraction amplitude in the contraction state sequence segment and the minimum relaxation amplitude in the relaxation state sequence segment is calculated and averaged, and denoted as the amplitude difference.

7. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 1, characterized in that, Determining the patient's muscle recovery status based on the amplitude change trend and the amplitude difference includes: The product of the amplitude change trend and the amplitude difference is calculated and normalized to obtain the muscle recovery state.

8. The method for detecting muscle mass in fracture patients based on electromyography signal analysis according to claim 7, characterized in that, The rehabilitation training plan based on the muscle recovery status includes: The patient's recovery level is determined based on the muscle recovery status, and the rehabilitation training program is determined based on the recovery level.

Images